Transitions to better lives : offender readiness and rehabilitation [1 ed.] 9781843927181, 9781843927198, 9781843927204

Transitions to Better Lives aims to describe, collate, and summarize a body of recent research – both theoretical and em

348 72 7MB

English Pages 320 [332] Year 2010

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Transitions to better lives : offender readiness and rehabilitation [1 ed.]
 9781843927181, 9781843927198, 9781843927204

Citation preview

Transitions to Better Lives Offender readiness and

willan

publishing

Andrew Day, Sharon Casey, Tony Ward, Kevin Howells and James Vess

Transitions to B e tte r Lives

Transitions to Better Lives Offender readiness and rehabilitation

A n d re w Day, Sharon Casey, Tony W a rd , Kevin Howells and James Vess

WILLAN P U B L IS H IN G

Published by W illan Publishing Culm cott House M ill Street, U ffculm e Cullom pton, Devon EX15 3AT, UK Tel: +44(0)1884 840337 Fax: +44(0)1884 840251 e-m ail: info@w illanpublishing.co.uk w ebsite: www.w illanpublishing.co.uk Published sim ultaneously in the U SA and C anada by Will an Publishing c / o ISBS, 920 NF. 58th Ave, Suite 300, Portland, O regon 97213-3786, USA Tel: +001(0)503 287 3093 Fax: +001(0)503 280 8832 e-m ail: [email protected] w ebsite: www.isbs.com

© Andrew Day, Sharon Casey, Tony Ward, Kevin I low ells and Jam es Vess 2010 I h e rights of Andrew Day, Sharon Casey, Tony Ward, K evin H ow ells and Jam es Vess to be identified as the authors o f this book have been asserted by them in accordance w ith the Copyright, Designs and Patents Act of 1988. All rights reserved; no part o f this publication m ay b e reproduced, stored in a retrieval system , or transm itted in any form or by any m eans, electronic, m echanical, photocopying, recording or otherw ise w ithout the prior w ritten perm ission o f the Publishers or a licence perm itting copying in the UK issued by the Copyright Licensing A gency Ltd, Saffron House, 6 -1 0 Kirby Street, London EC1N 8TS.

First published 2010 ISBN 978-1-84392-718-1 paperback 978-1-84392-719-8 hardback British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library

FSC Mixed Sources Product 3'»up from virll-nvlnj^ed lo tttU « rd c tfie r« sn ira ile d M v fC M C trl m . SOS COC-2tt? www.lic.org

Project managed by Deer Park Productions, Tavistock, Devon Typeset by GCS, Leighton Buzzard, Bedfordshire Printed and bound by T.J. International, Padstow, Cornwall

Contents

Figures and tables Abbreviations Acknowledgements Authors and contributors

vii ix xi xiii

Part One What is Treatment Readiness? 1

The M ultifactor Offender Readiness M odel

2

The origins of treatm ent readiness Ralph C. Serin, Sharon M. Kennedy, Donna L. Mailloux and Laura ]. Hanbi/

12

3

W hat are readiness factors?

27

4

The Good Lives M odel of offender rehabilitation and treatm ent readiness

47

The assessm ent of treatm ent readiness

62

5

3

Part Two Readiness and Offenders 6

7

Interpersonal violence: the need for individualised services Sex offenders: understanding low readiness

93 104

Transitions to Better Lives

8

Substance use and readiness

125

9

Readiness and treatm ent engagem ent in personality disordered offenders: tow ards a clinical strategy

141

Part Three Clinical and Therapeutic Approaches to Working with Low Levels of Readiness 10

The m odification of low readiness

161

11

Goal-focused interventions with offenders Man/ McMurran

180

12

Treatment readiness and the therapeutic alliance Christina Kozar

195

13

Readiness and risk: a case illustration

214

14

Ways forward and conclusions

234

Appendix: Measuresof Treatment Readiness Selected journal articles References Index

236 252 254 289

vi

Figures and tables

Figures 1.1 2.1

O riginal m odel of offender treatm ent readiness Conceptual model o f treatment responsivity

8 16

Tables 1.1 1.2 3.1 3.2 3.3 3.4 5.1 5.2 5.3 5.4 5.5 9.1 10.1

Im pedim ents to readiness for anger m anagem ent D istinguishing betw een m otivation, responsivity and treatm ent readiness The potential im pact of cognitive distortions on treatm ent readiness A ffective readiness factors Potential m oderators of the effects of affective readiness factors Influences on perceptions o f coercion Psychom etric properties for EBA fram ew ork N orm s and reliability criteria Validity and utility criteria Ratings of instrum ents for case conceptualisation and treatm ent planning Ratings of instrum ents for treatm ent m onitoring and evaluation Engagem ent in therapy contexts N egotiated casew ork

7 10 30 31 35 44 70 72 74 86 89 154 169

Transitions to Better Lives

10.2 11.1 11.2

v iii

M otivational interview ing Goal positioning system A ction plan

174 186 189

Abbreviations

AMI ARCQ ASPD ATSA CBT CLA H RC CV TRQ DSM-IV-TR DSPD EssenCES FoSOD GLM GLP G PS MI M MPI M O RM N IH R O RS PCI PCI-OA PCL-R PDO

A daptations of m otivational interview ing A nger Readiness to Change Questionnaire A nti-social personality disorder A ssociation for the Treatment of Sexual Abusers Cognitive behavioural treatm ent Collaborations for Leadership in Applied H ealth Research and Care Corrections Victoria Treatment Readiness Questionnaire Diagnostic and Statistical Manual of Mental Disorders, 4th edn, text revised D angerous and Severe Personality Disorder Essen Clim ate Evaluation Schem e Facets of Sexual O ffender Denial Good Lives M odel Good Lives Plan Goal positioning system M otivational inverview ing M innesota M ultiphasic Personality Inventory M ultifactor O ffender Readiness M odel N ational Institute of H ealth Research O utcom e Rating Scale Personal Concerns Inventory Personal Concerns Inventory - O ffender Adaptation Psychopathy C hecklist - Revised (Hare) Personality disordered offender ix

Transitions to Better Lives

RCQ RCQ-TV RCT RN R SE SOCRATES SRS STRS TA TRCRS TTM URICA VRS VRS-SO VTRQ WAI

Readiness to Change Questionnaire Readiness to Change Q uestionnaire [Treatment Version] Random ised controlled trials R isk-need-responsivity Supportive-expressive (therapy) Stages of Change Readiness and Treatment Eagerness Scale Session Rating Scale Serin Treatment Readiness Scale Therapeutic alliance Treatment Readiness Clinical Rating Scale Transtheoretical model U niversity of Rhode Island Change A ssessm ent Violence Risk Scale Violence Risk Scale - Sex O ffender Violence Treatment Readiness Questionnaire W orking Alliance Inventory

Acknowledgements

This book is the product of a long-term collaboration betw een a group of clinical and forensic psychologists w ho are interested in the field of offender rehabilitation. O ver the last ten years the group has investigated different aspects o f treatm ent readiness, publishing a num ber of journal articles on the topic (for a full list of these articles, see page 252). Parts o f this book draw on these articles, but they have been rew ritten and collated for the purposes of this volum e. In addition, we have invited others with particular expertise in this field to prepare chapters. We are very grateful to them for their contributions. We would also like to extend our thanks to Brian Willan, Julia W illan and their colleagues at W illan Publishing as well as Bill and M ichelle A ntrobus and the rest of the team at Deer Park Productions for their support w ith this project.

xi

Authors and contributors

Sharon C asey com pleted her PhD in forensic psychology in 2004. H er areas of expertise include offender assessm ent, juvenile offending, program m e evaluation, and psychology and the law. Dr Casey has been a collaborator with other m em bers of the group on several publications and projects, and has been involved in the com pletion of a num ber of com petitively tendered research consultancy and consultancy projects. D r Casey has also worked w ith a num ber of governm ent agencies in both South A ustralia and Victoria and supervised a large num ber of Psychology H onours and M asters theses. A ndrew Day is a clinical and forensic psychologist w ho is interested in research and practice in the area of offender rehabilitation. He is currently an A ssociate Professor w ith the Centre for O ffender Reintegration at Deakin, in the School of Psychology, Deakin University, A ustralia. H is m ost recent books include Writing Court Reports (2007, A ustralian A cadem ic Press), Anger and Indigenous Men (2008, Federation Press) and Integrated Responses to Domestic Violence: Research and Practice Experiences in Working with M en (2009, Federation Press). Laura J. H anby, M A, is a PhD candidate in forensic psychology at Carleton University, O ttaw a, O ntario, Canada. Her research exam ines the relationship betw een offender com petencies, correctional program m e perform ance and offender change. She has recently been involved in a range of research projects, including the role of xiii

Transitions to Better Lives

psychological assessm ents in parole decision-m aking, the relationship betw een the therapeutic alliance and sex offender program m e perform ance, and the variability in treatm ent engagem ent by type of violent offender. K ev in H ow ells is Professor of Foren sic/C lin ical Psychology in the Institute of M ental Health at N ottingham University. He is also Head of the Peaks A cadem ic and Research Unit at R im p to n H ospital, with a responsibility for service evaluation and research relating to the Dangerous and Severe Personality Disorder initiative at the hospital. H e is a chartered forensic and clinical psychologist and has long-term clinical and research interests in cognitive behavioural treatments, anger, violence, treatm ent readiness and, latterly, therapeutic clim ates in forensic units. He is a Fellow of the British Psychological Society and has worked in the UK, A ustralia and the USA. Sharon M . K enned y received her PhD in clinical psychology from the U niversity of Ottawa in 1990 and currently w orks with the Correctional Service of Canada as an Area Psychologist for Greater O ntario and N unavut D istrict at the O ttaw a Parole Office. In addition to her responsibilities at the O ttaw a Parole O ffice, D r Kennedy is a part-tim e lecturer at the U niversity of O ttaw a in forensic psychology. She has worked as a consultant and trainer for the N ational Institute of Corrections, International Com m unity Corrections A ssociation, the Am erican Probation and Parole A ssociation, the Canadian Training Institute, the D epartm ent of Youth Services in O hio, and the Federal Bureau of Prisons. Her current research interests include ris k / need assessm ent, violent offenders, and treatm ent responsivity and readiness to change. C h ristina K ozar has worked as a forensic psychologist w ithin correctional environm ents since 1997, largely in the delivery of offending behaviour program m es. She is currently a PhD candidate at Deakin U niversity and is exam ining the relationship betw een the therapeutic alliance and outcom es in offending behaviour program m es for clients who are personality disordered. M ary M cM u rran is Professor in the U niversity of N ottingham 's Institute of M ental Health. She joined HM Prison Service in 1980 and worked up the ranks to becom e Head of Psychology Services at a young offenders' centre. A fter qualifying as a clinical psychologist, she becam e Head of Psychology Services at Ram pton H ospital and xiv

A u th o rs and c o n trib u to rs

then at the East M idlands Centre for Forensic M ental H ealth. In 1999, she was awarded a five-year Senior Baxter Research Fellow ship by the N H S's N ational Program m e on Forensic M ental H ealth Research and Developm ent. She has w ritten over 100 academ ic articles and book chapters. H er edited text, Motivating Offenders to Change (Wiley, 2002) was com m ended in the British M edical A ssociation's 2003 Book Com petition. She is a Fellow of the British Psychological Society and form er Chair of the Society's Division of Forensic Psychology. In 2005, she w as recipient of the D ivision of Forensic Psychology's Lifetim e A chievem ent Award. She w as founding co-editor of the British Psychological Society journal Legal & Criminological Psychology. D onna L. M aillou x, M A, is a PhD candidate in forensic psychology at Carleton University, O ttaw a, O ntario, Canada. H er research has included the areas of sex offender program m ing and assessm ent, learning disabilities, treatm ent readiness and responsivity, offender change, hostage-takings and psychopathy. She has published research relating to juvenile and adult populations for both m ales and females. R alph C. S erin was em ployed w ith Correctional Service C anada from 1975 to 2003 in various capacities. He is presently an Associate Professor at Carleton U niversity in the D epartm ent of Psychology where he teaches forensic and correctional psychology courses. He has published in the areas of parole, treatm ent readiness and responsivity, risk assessm ent, psychopathy, sexual offenders, and the assessm ent and treatment of violent offenders. Current research topics include correctional program m ing (evaluation, m easurem ent of readiness and gain, factors influencing program m e drop-out); violent offenders (pathw ays to desistance, typologies, specialised intervention and m anagem ent); and d ecision-m aking (strategies, com petencies, and factors influencing accuracy). Jam es Vess has over 25 years of clinical and research experience w ith forensic populations. H e received his PhD in clinical psychology from O hio State University in the United States, and then served in a variety of treatm ent, assessm ent, and supervisory roles at Atascadero State H ospital, the m axim um security forensic psychiatric facility in California. He was subsequently a Senior Lecturer at Victoria U niversity of W ellington, w here his research focus w as prim arily on risk assessm ent w ith violent and sexual offenders, as w ell as public policy dealing w ith high-risk offenders. H e is now a Senior Lecturer XV

Transitions to Better Lives

at Deakin U niversity and a m em ber of the Clinical Forensic Research group of the recently form ed Centre for O ffender Reintegration at Deakin. H e has also remained active in forensic practice, and has provided expert w itness evidence in a variety of H igh Court and Court of A ppeals cases for Preventive Detention and Extended Supervision in both N ew Zealand and Australia. Tony Ward, MA (H ons), PhD, DipClinPsyc, is a clinical psychologist by training and has worked in the forensic field since 1989. He has taught clinical and forensic psychology at the U niversities of Canterbury and M elbourne, and is currently Professor of Clinical Psychology at Victoria U niversity o f W ellington, New Zealand. H e has over 260 publications and his recent books include Rehabilitation: Beyond the Risk Paradigm (with Shadd M aruna, Routledge, 2007) and People Like Us: Desistance from Crime and Paths to the Good Life (with Richard Law s, Guilford, in press). H is research interests centre on rehabilitation, forensic ethics and cognition.

xvi

A M elodious Poem

It is our m otivation, good o r bad , that d eterm ines the qu ality o f o u r actions. W hen o u r m otivation is pure, Even the rough earth looks friend ly and the roads w e w alk safe. W h en o u r m o tiv atio n is poor, E ven a good hom e feels cold and the road w e w alk lonely. Since all is d ep en d en t on ou r intention, B eing consum ed by the m ean spirit o f envy, D riven alw ays by hatred and desire, D o you not think this is the cau se o f ou r suffering? O h you intelligent people, think about it seriously.

Lam a Tsong K hapa the G reat (thirteenth century, trans. S. Rigzin, in Fallon 2005: 243)

P art One

W hat is Treatment Readiness?

Chapter I

The M ultifactor Offender Readiness Model

These are particularly challenging times for researchers and practitioners who seek to work with offenders in ways that will assist them to live better lives. A range of different perspectives currently inform this w ork, from those that em phasise the rights of victim s and com m unities to those that em phasise the rights of individual offenders. In m any parts o f the w orld, m ore and m ore people are being im prisoned and for longer periods of time. Com m unities are becom ing more risk aversive and punitive in their attitudes towards offenders and there would appear to be a grow ing determ ination to m ake individuals pay severely for transgressions against the state. At the sam e time significant effort is put into rehabilitating offenders and helping them to plan for a successful reintegration back into society. Indeed, the last tw enty or so years have seen significant investm ent in the developm ent and delivery of offender rehabilitation program m es across the w estern world, in both prison and com m unity correctional (probation and parole) settings, and support for rehabilitative ideals is perhaps now more clearly enshrined in public policy than perhaps at any time in the past. That is not to say, however, that the value o f offender rehabilitation is universally recognised, and it is in this context that interest in issues such as hum an rights, offender dignity, and the values of offender rehabilitation has grow n (see Ward and Birgden 2007; Ward and M aruna 2007). The socio-political context in w hich any w ork w ith offenders takes place ensures that attem pts to reintegrate or rehabilitate offenders w ill alm ost certainly com e under a high level of scrutiny, both public and professional. It is now more im portant than ever that 3

Transitions to Better Lives

rehabilitation providers can dem onstrate that their efforts are effective in reducing rates of reoffending or, at the very least, consistent with those practices that have been show n to be effective in other settings. M ost correctional agencies have now developed accreditation and quality assurance system s designed specifically to ensure that the program m es offered m eet basic standards of good practice. There are thousands of controlled outcom e studies from which to determ ine the types of intervention that are likely to be effective (H ollin 2000), the results of which, w hen aggregated, offer consistent and persuasive evidence that offender rehabilitation program m es can, and do, have a positive effect on reducing recidivism . Furtherm ore, it is clear that these reductions are likely to be o f a m agnitude that is socially significant. It has also becom e apparent that program m es that adhere to certain principles are likely to be even more successful in reducing recidivism (Andrew s and Dow den 2007). It is this know ledge that has led to the developm ent of a model of offender m anagem ent com m only know n as the 'w hat w orks' or 'risk-n eed s-resp on siv ity' (or RNR) approach, based largely on the sem inal w ork of D on A ndrews and Jam es Bonta. The RN R approach centres around the application of a num ber of core principles to offender rehabilitation (prim arily the risk, needs, and responsivity principles), each of which seeks to identify the type of person w ho m ight be considered suitable for rehabilitation initiatives. Perhaps most progress here has been m ade in the area of risk assessm ent, with recent years seeing the developm ent and validation of a wide range o f specialist tools designed to help identify those w ho are m ost likely to reoffend. The logic is com pelling - if the goal of intervention is to reduce recidivism , then effort should be invested in w orking with those w ho are the m ost likely to reoffend, rather than those w ho probably will not. It is possible to m eaningfully categorise offenders into different risk brackets using a relatively small set of variables (such as the age at first offence or the num ber of previous offences). A focus o f current w ork in this area is on the identification and assessm ent of those risk factors that have the potential to change over time. These 'dynam ic' risk factors, or w hat have becom e known as 'crim inogenic need s' (see W ebster et al. 1997), are particularly im portant in determ ining treatm ent targets (that is those areas of functioning that m ight be addressed w ithin offender rehabilitation program m es). In com parison, the third major tenet of the RN R approach - responsivity - has been som ew hat neglected. This term is com m only used to refer to those characteristics of individual offenders (such as m otivation to change) that are likely 4

The Multifactor Offender Readiness Model

to influence how m uch they are able to benefit from a particular programm e. In m any respects, the RN R approach has revolutionised correctional practice. It has promoted the idea o f com m unity safety as the prim ary driver behind correctional case m anagem ent, and given offender rehabilitation program m es a central role in the sentence planning process. The approach has had a m ajor im pact on practice in relation to offender assessm ent and the selection of appropriate candidates for intervention around the w estern world. It has, however, had less influence on the actual practice of offender rehabilitation (see Andrews 2006; Bonta et al. 2008), and significant gaps in know ledge rem ain (Andrews and Dow den 2007). Critics of the RN R m odel have, in a range of different ways, draw n attention to how the m odel struggles to inform the process of program m e delivery, and how psychological and behaviour change takes place. This may be, in part, because the RN R model was developed as an approach to offender m anagem ent rather than psychological therapy. It may also perhaps relate to difficulties in the w ay in w hich som e of the key terms (notably risk and needs) have been conceptualised, and in particular how the overarching focus on risk can be experienced as dem otivating for individual participants in rehabilitation program m es, ultim ately contributing towards high rates of program m e attrition and a lack of rehabilitative success (Thom as-Peter 2006; Ward and Stew art 2003). W hile the notion that offender rehabilitation is som ething that can be done to som eone, possibly even w ithout their consent, has appeal, it is also therapeutically naive. The gains m ade in the area of offender assessm ent and selection have not, in our view, been m atched by progress in the area of offender treatm ent, where concerns are com m only expressed about issues of offender m otivation and engagem ent in behaviour change, therapist skill and training, program m e integrity, and the social clim ate of institutions in which interventions are delivered. Perhaps now here are these issues more apparent than in the areas of treatm ent readiness and responsivity. It is our contention that work in this area has been ham pered by a lack of conceptual clarity about the construct of responsivity, how it m ight be operationalised, and how it m ight be reliably assessed. In this book, we explore the idea that even greater reductions in recidivism than those dem onstrated in program m es that adhere to the evidence-based principles of risk and needs can be m ade w hen program m es are able to be responsive to individual needs. We discuss the m eaning and nature of the term 'treatm ent readiness' and how this m ight inform the rehabilitative 5

Transitions to Better Lives

process. Readiness is proposed as an overarching term that encom passes both the internal com ponents of responsivity (offender m otivation, problem aw areness, em otional capacity to engage with psychological treatm ent, goals, and personal identity), as well as those external com ponents that m ay be specific to the environm ent in w hich treatm ent is com m only offered. O ur interest in the notion of treatm ent readiness arose out of work in w hich we exam ined the effects of anger m anagem ent program m es offered to offenders (H ow ells et al. 2005; H eseltine et al. 2009). These evaluations suggested that anger m anagem ent training, at least of the type com m only offered in A ustralian prisons at the time, was unlikely to be particularly effective in bringing about behavioural change - in this context this referred to physical aggression and violent behaviour of a crim inal nature. At the time, prison adm inistrations across Australia dedicated considerable energy and resources to the developm ent and delivery of anger m anagem ent program m es to violent offenders, and so these apparently weak treatm ent effects required som e explanation. A num ber of hypotheses were proposed, including those relating to the selection of appropriate candidates, the m atching of the intensity o f the intervention to the level of risk and need, and the extent to w hich those who are im prisoned for violent offending m ight be considered to be ready for treatment. In a subsequent paper, H ow ells and D ay (2003) developed the notion of treatm ent readiness by identifying seven im pedim ents that potentially inhibited the effective treatment of offenders presenting with anger problem s (see Table 1.1). This w ork was subsequently elaborated into a more general model of readiness which w as then applied to all form s of offender rehabilitation program m ing (Ward et al. 2004b). The M ultifactor O ffender Readiness M odel (M O RM ) proposed that im pedim ents or barriers to offender treatm ent can reside w ithin the person, the context, or w ithin the therapy or therapeutic environm ent. The follow ing definition of treatm ent readiness was put forward: the presence of characteristics (states or dispositions) within either the client or the therapeutic situation, which are likely to promote engagement in therapy and which, thereby, are likely to enhance therapeutic change. A ccording to this definition, readiness to change persistent offending behaviour requires the existence of certain internal and external conditions w ithin a particular context (see Figure 1.1). Offenders w ho are ready to enter a specific treatm ent program m e are thus viewed as possessing a num ber of core psychological features that enable them to function w ell in a particular rehabilitation program m e at a particular time. 6

The M u ltifacto r O ffe nd e r Readiness Model

Table 1.1 Number 1

2 3

4 5

6 7

Impediments to readiness for anger management Description The complexity of the cases presenting w ith anger problems. This includes the coexistence of mental disorders with aggressive behaviour. The setting in which anger management is conducted. Existing client inferences about their anger problem. For exam ple, inferences indicating that the anger was viewed as appropriate and justified. The impact of coerced or mandatory treatment. The inadequate analysis of context of personal goals within which the anger problem occurs. It is possible that the expression of anger could increase the likelihood that important personal goals are achieved. Ethnic and cultural differences. Gender differences in the experience and expression of anger.

Source: Adapted from Howells and Day (2003).

Individual or person readiness factors are cognitive (beliefs, cognitive strategies), affective (em otions), volitional (goals, w ants, or desires), and behavioural (skills and com petencies). The contextual readiness factors relate to circum stances in w hich program m es are offered (mandated vs voluntary, offender type), their location (prison, com m unity), and the opportunity to participate (availability of program m es), as well as the level o f interpersonal support that exists (availability of individuals w ho w ish the offender w ell and would like to see him or her succeed), and the availability of adequate resources (quality of program m e, availability of trained and qualified therapist, appropriate culture). It is suggested that these personal and contextual factors com bine to determ ine the likelihood that a person will be ready to benefit from a treatm ent program m e. Those w ho are treatm ent ready will engage better in treatm ent, and this will be observably evident from their rates of attendance, participation, and program m e com pletion. A ssum ing that program m es are appropriately designed and delivered, and they target crim inogenic need, higher levels of engagem ent are considered likely to lead to reductions in levels of crim inogenic need and a consequent reduction in risk level. The m odel thus incorporates w hether or not a person is ready to change his or her behaviour (in the general sense); to elim inate a specific problem ; to elim inate a specific problem by virtue of a specific method (such as cognitive behavioural therapy); and, finally, 7

Transitions to Better Lives

Source: A d a p te d fro m W ard e.l ah (2004b).

Figure 1.1

O riginal m odel of offender treatm ent readiness

to elim in ate ¿a specific problem by virtue of a specific m ethod at a specific time. To be treatm ent ready, offend ers m ust n o t only recognise that their offend ing is problem atic, bu t also m ake a decision to seek help from others. T his im p lies a b elief that they are unable to d esist from offend in g unaided. O nce the offend er m akes a genuine com m itm ent not to reoffend, he or she m ay then be taught the relevan t skills and strategies in treatm ent to help achieve this goal. The decision to seek help m ay also be affected by factors such as w hich services are available, attitu d es or beliefs about those services, beliefs about the im portance o f privacy and autonom y, o r that problem s are likely to d im in ish over tim e anyw ay. T h e extent to w hich a behaviou r or a feeling is defined as a problem w ill, in part, be determ ined by cultural ru les and norm s relating to w hat is acceptable or approp riate (for exam p le w om en gen erally have m ore p ositive attitu d es tow ards help -seek in g than men: Boldero and Fallon 1995), and in environm en ts where certain types of offend in g are considered norm ative, it is unlikely th at the ind ividu al will see h is /h e r offend in g as problem atic. O ther con textual factors, such as poverty, m ay also influence the d ecision to recognise a particular behav iou r as a problem . O f course, an offend er m ay be ready to w ork on a particular problem , b u t not necessarily on e that the therapist view s as relevant and central to his or her offend in g; to be treatm ent ready, both the treatm ent provider and the offend er hav e to agree on both the goals and the tasks of the treatm ent. The M O R M was developed in a w ay that d istin gu ishes betw een three d istin ct although related constructs: treatm ent m otivation,

The Multifactor Offender Readiness Model

responsivity, and readiness. The constructs of m otivation and responsivity are conceptualised as som ew hat narrow er in scope than that of readiness (see Table 1.2). Furtherm ore, readiness directs us to ask w hat is required for successful entry into a program m e, while the concept of responsivity focuses attention on w hat it is that can prevent treatm ent engagem ent. Ward et al. (2004b) suggest that the responsivity concept has not really developed conceptual coherence and, as such, is often poorly operationalised as a list of relatively independent factors (see Serin 1998). We suggest that treatment readiness m ay be a better model because of its greater scope, coherence, testability, and utility (fertility). O ur aim in w riting this book is to describe, collate, and sum m arise a body of recent research, both theoretical and em pirical, that explores the issue of treatm ent readiness in offender program m ing. The book is divided into different sections. In the first, we unpack our model of treatment readiness and how it has been operationalised. Ralph Serin and colleagues also describe their understanding of the notion of treatment readiness (Chapter 2). We then discuss in Part Two how the construct has been applied to the treatment of different offender groups. In Part Three, we discuss som e of the practice approaches that have been identified as holding prom ise in addressing low levels of offender readiness. We have included contributions from a num ber of authors w hose w ork has stim ulated discussion and helped to inform practice in offender rehabilitation. Collectively we hope that these chapters offer a useful resource for researchers and academ ics alike, describing current thinking and know ledge in this area. We chose to call the book Transitions to Better Lwes to remind us of the ultim ate purpose behind any attem pt to rehabilitate offenders - that is, to help individuals learn how to m eet their needs in w ays that are both personally fulfilling and socially responsible. It is this possibility that m otivates practitioner and offender alike and, in our view, lies at the very heart of successful rehabilitation.

9

Transitions to B e tte r Lives

T ab le 1.2

D istinguishing betw een m otivation, responsivity and treatm ent

readiness C onstruct

Description

M otivation

M otivation is w idely recognised as im portant in that offenders are usually selected for treatm ent partly on the basis of being m otivated to participate. Professionals typically judge that offenders are m otivated w hen they express regret for their offences, express a desire to change, and sound enthu siastic about the treatm ents on offer. M otivation in this context relates to w hether som eone w ants to enter treatm ent; that is, ascertaining his or her level o f v olition with respect to changing particular target behaviours. There is, how ever, no consensus regarding w hat is m eant by offenders' m otivation and n o system atic exam ination of the factors that influence it (M cM urran and Ward 21X14), despite it b ein g w idely accepted lhat a m ajor task in treatm ent is lo nurture and enhance m otivation to change.

Responsivity

The term responsivity is used to refer to the use of a style and m ode of intervention that engages the client group (A ndrew s and Bonta 2003). R esponsivity can be further divided into internal and external responsivity w hereby attention to internal responsivity factors requires therapists to m atch the content and pace of sessions to specific client attributes, such as personality and cognitive m aturity, w hile external responsivity refers to a range of general and specific issues, such as the use of active and participatory m ethods. External responsivity has been divided further into staff and setting characteristics (K ennedy 2001; Serin and K ennedy 1997). A lthough responsivity as usually understood in the rehabilitation literature, is prim arily focused on therapist and therapy features and thus is essentially concerned w ith adjusting treatm ent delivery in a w ay that m axim ises learning. Table 1.2 continues opposite

10

Th e M u ltifa c to r O ffe n d e r Readiness M odel

Table 1.2 continued

C onstruct

Description

Readiness

The concept of readiness w as originally articulated in an offender context by Ralph Serin (Serin and Kennedy, 1997; Serin 1998), although it had previously been used in offender substance use treatm ent program m es (e.g. D eLeon and Jainchill 1986). It has been broadly defined as the presence of characteristics (states or dispositions) w ithin either the client or the therapeutic situation, w hich are likely to prom ote engagem ent in therapy and w hich, thereby, are likely to enhance therapeutic change (H ow ells and Day 2003). To be ready for treatm ent m eans that the p erson is m otivated (i.e. w ants to, has the will to), is able to respond appropriately (i.e. perceives he or she can), finds it relevant and m eaningful (i.e. can engage), and has the capacities (i.e. is able) to successfu lly en ter the treatm ent program m e.

Source: A dapted from Ward et al. (2004b).

Chapter 2

The origins o f treatm ent readiness Ralph C. Serin, Sharon N\. Kennedy, Donna L Mailloux and Laura J. Hanby

The field of forensic and correctional psychology has witnessed an explosion of interest in risk assessm ent over the past three decades. E ncouragingly m any of these risk assessm ent approaches have evolved from sim ple aggregation of crim inal history factors to the identification of dynam ic needs and treatm ent targets. In turn, this has led to improved clarity regarding program m ing m odels for offenders, both in prison and in the community. Concurrently, follow ing the pessim ism of M artinson's (1974) 'nothing w orks' conclusion, a sem inal paper by A ndrews et al. (1990) led the w ay for an explosion of a new era of correctional rehabilitation. This applied research began to be know n as the W hat Works literature and has been the pre-em inent perspective for offender program m ing internationally for the past tw o decades. Strengthened by em pirical evidence regarding their utility, correctional program m es have m ultiplied exponentially, now dotting the correctional landscape in m ost w estern countries. Indeed, since the late 1980s the zeal to develop and deliver correctional program m es has continued unabated. N otw ithstanding the enthusiasm of clinicians, it becam e apparent in the early days of correctional program m ing that not all offenders em braced change. Although unsurprising, this treatm ent resistance raised concerns regarding the potential for w asting treatm ent spots through program m e attrition as w ell as the potential attenuation of effectiveness in evaluation studies. O f note, since these early days of offender program m ing, evaluations typically now include program m e drop-outs w ithin the com parison group to ensure that effect sizes are not unduly inflated (Rice and H arris 1997b, 2005). H ence, the 12

The origins of treatment readiness

issue of program m e attrition has becom e both a m ethodological and a practical concern. For those of us (RS and SK) w orking as part of a large corrections agency com m itted to offender program m ing, it becam e apparent that w hat w as required was som e approach to understanding and m easuring offender readiness, such that program m ing efforts were m axim ised. We therefore saw treatm ent readiness as but one construct, albeit crucial, to be included in the conceptualisation of treatm ent response and offender change. The focus of this chapter is to describe a program m e of research relating to the conceptualisation and assessm ent of treatment responsivity, beginning in 1994. A t that time, we viewed treatm ent responsivity as an overarching term to reflect treatm ent readiness and interpersonal style factors that would influence treatm ent perform ance. Treatment readiness was therefore a requisite com ponent w ithin a model of offender change and any assessm ent protocol of offender program m ing. Accordingly, this chapter reviews the related constructs discussed in the literature at that time in order to distinguish the uniqueness of treatm ent readiness. We provide som e observations on recent research regarding offender change and im plications regarding our earlier efforts, and com m ent on the continued theoretical and practical utility of treatm ent readiness. At the onset of this research, correctional program m es were evolving from an eclectic m ix of program m es (substance use, anger, and sex offender program m ing) m ainly delivered by psychologists to the more current structured program m es now com m only delivered by para-professionals. It w as in this context that an appreciation for an understanding of factors that m ight enhance or m itigate program m e participation was conceived. We wanted to develop a m odel and set of m easures that were easy to use by non-psychologists and would inform program m e retention and perform ance. O ur logic w as that if we could distinguish am ong program m e referrals in terms of treatment readiness, then prim ing could be provided to those offenders w ho presented as less ready to participate. U nderlying this w ork was the recognition that program m ing, regardless of its scope, was intended to be a vehicle for offender change (although this does not mean to imply that change cannot occur in the absence of participation in form al correctional program m es). N onetheless, change would be more likely to occur if the offender fully participated in the program m e. Unfortunately, in these early days there were not the eloquent descriptions of program m es that now exist (M cGuire 2001), nor the detailed criteria defining correctional program m es com m on on m ost correctional agencies' w ebsites (Correctional Service of Canada 2009). 13

Transitions to Better Lives

M ore recent conceptualisations of offender program m ing include com m ents about responsivity, in particular offender m otivation, but such language was scarce in 1994 (M ichenbaum and Turk 1987). It should be apparent that advances in program m ing m odels and risk assessm ent were ju st beginning to take hold in clinical practice as w e em barked on our w ork on our conceptualisation and assessm ent of treatm ent readiness. Quinsey (1988) w as prescient in noting that im provem ents in our understanding of treatability were likely to be of greater im portance than im proved risk prediction. Prior to 1990 such term s as treatability, m otivation, and readiness for change had been used interchangeably. As w ell, the terms were prom inent in the area of m ental health in that some legal statutes required a consideration o f treatability in a range of crim inal justice decisions ranging from the granting o f bail to sentencing and discretionary release. Earlier w ork by Q uinsey and M aguire (1983) spoke to the am biguity of the construct and poor inter-clinician reliability in its assessm ent. Further, Rogers and W ebster (1989) noted that treatability referred to the clinical determ ination of which patients, under w hat treatm ent m odalities and circum stances, will respond m ost favourably. That few clinicians could agree on how' to assess treatability led H eilbrun and his colleagues (1992) to seek to develop a scale that reflected the m ultifaceted nature of the concept. Their effort w as substantial b ut in the end yielded m odest reliability and was incredibly labour-intensive. Further, w hile such efforts might be viable in a mental health setting, it w as highly unlikely that such resources would be allocated for assessing treatability in correctional settings. N onetheless, few clinicians disagreed w ith the im portance of such a construct and there was widespread appreciation that offenders differed with respect to their interest in program m ing, regardless of dem onstrated risk level or nature of identified needs. At the same time as the issue of treatability was being addressed, w ork principally in the area o f addictions was also germ ane. M iller and Rollnick's (1991) influential w ork on m otivational interview ing (M I) and the work of Prochaska, DiClem ente and N orcross (1992) on readiness for change, as m easured b y the Readiness for Change Q uestionnaire (U niversity of Rhode Island Change A ssessm ent), under­ scored the im portance of both m otivation and readiness in predicting program m e engagem ent and outcom e. The form er challenged the apparent interest by some in confrontational approaches to treatment, w hile the latter underscored the need for view ing change as a stagebased process. Both approaches offered strategies for attending to responsivity issues in offender treatm ent such that outcom es would 14

The origins of treatment readiness

be enhanced and both rem ain im portant in correctional program m ing today (A ndrew s and Bonta 1998). At the time that we began this w ork it seem ed that there were converging lines of research that underscored the need to consider contextual factors that might influence offender program m ing. O ur goal was to m erge these com plem entary view points of treatability, readiness for change, and m otivation, and overlay a corrections interpretation. We also wanted to consider a range of dom ains that m ight influence treatm ent engagem ent so that there were m ultiple avenues (targets) to address in order to enhance program m e engagem ent, w hen needed. We term ed this construct 'treatm ent readiness'. O ur expectation was that improved engagem ent would lead to improved program m e perform ance, w hich in turn would yield improved program m e outcom es. The balance of the chapter describes our conceptual m odel, the developm ent of an assessm ent of treatment readiness and related constructs, and com m ents regarding its application in corrections. Adm ittedly, m uch of this viewed the problem through the lens of w hat an individual offender needed to do to enhance outcom es, but supported by skilled clinicians. This is in contrast to the more m ultifaceted M O RM m odel described elsewhere in this book.

Conceptual model O ur initial efforts (RS and SK) were to consider the extant literature and to brainstorm regarding how w e perceived the different com ­ ponents that influence offender change and how they might be situated. This initial m odel, then, organised our w ork plan regarding a program m e of research, especially w ith respect to the developm ent of an instrum ent to assess treatm ent readiness. Next, we brainstorm ed to create an inclusive list of dom ains that could be obstacles to change, based on our observations during our respective careers as clinical psychologists in corrections. These dom ains were reviewed and refined to yield a relatively independent (we hoped) series of dom ains. Given our interest in a behavioural rating strategy rather than self-report questionnaire, w'e developed a series of anchors to guide assessors. Figure 2.1 presents our conceptual m odel in 1997. We hypothesised that treatm ent readiness w as more expansive and m ultidim ensional than sim ply an index of m otivation. Indeed, there had ju st been a clinical rating of offender m otivation developed as part of an offender intake assessm ent but it was not well incorporated 15

Transitions to Better Lives

into decisions about program m ing. We hypothesised that a variety of factors would contribute to an overall index of treatm ent readiness. Further, personality factors, w hich w e termed 'interpersonal style', w ere hypothesised to interact w ith treatm ent readiness, thereby influencing overall treatm ent response. Finally, we w anted to have an overall conceptual m odel that provided staff w ith indices of offender change. A t the tim e we referred to this as 'treatm ent perform ance', but now we would likely use the term 'offender com petency' (H anby et al. 2009). W hat should be apparent is that the operationalisation of these three dom ains (treatm ent readiness, interpersonal style, and treatm ent perform ance) was slightly disconnected from the original conceptual m odel. That is, interpersonal style is prom inent in the assessm ent protocol but less so in the m odel. In hindsight, improved clarity regarding the relationship betw een interpersonal style and treatm ent readiness would have been desirable.

Treaimeni motivation and readiness

'r 1reaimeni compliance and participation

r 1rea imeni gains

1realm on l. generalization

Figure 2.1 16

Conceptual model of treatment responsivily.

The origins of treatment readiness

As can be seen in Figure 2.1, treatm ent response was hypothesised to be com prised of two contributing dom ains: treatability and treat­ m ent effectiveness. Treatability is reflected in indices of treatment readiness and treatm ent com pliance. Specifically, it was hypothesised that treatm ent readiness w ould lead to improved com pliance and participation, hence m eeting the requirem ent for determ ining that an individual w as treatable, as implied in the earlier research on treatability. H aving met the criterion for treatability, it is then of interest to determ ine if the offender changes as a function of program m e involvem ent and if such change generalises post-program m e (upon release). In this manner, recidivism is seen as a failure to generalise gains (assum ing gains were real and sufficient to change crim inal behaviour). In conclusion, treatm ent readiness was conceived to be part of a broader conceptual model to understand offender change and treatm ent outcom e. The balance of the chapter, however, focuses on the definition and assessm ent o f treatm ent readiness, as w ell as its im plications in offender program m ing. W hen this w ork was initiated our intention was to identify offenders for w hom there w as concern regarding low readiness so that pre-treatm ent prim ers could be provided. In this manner, the treatm ent readiness assessm ent w ould com plem ent m otivational inter­ view ing efforts (A nstiss et a l, in press), but such efforts would not be routinely applied, rather they would be presum ably reserved for a sub-group of program m e referrals. The expectation was that such preventative efforts would attenuate program m e attrition and in­ group m isbehaviour. N ow adays m ost high-intensity program m es for higher-risk offenders utilise som e form of engagem ent or m otivational enhancem ent to address potential treatm ent readiness (that is obstacles to change) concerns (Polaschek 2009; Serin and Preston 2001). The initial brainstorm ing exercise and subsequent review by colleagues yielded an initial 11 factors to consider w ithin the broader dom ain of treatm ent readiness. These are listed below. In order to be inclusive and reflect the overall assessm ent m odel, the factors for interpersonal style and treatm ent perform ance are also provided. B ehaviou ral rating assessm en t protocol Treatment readiness ratings 1 Problem recognition 3 M otivation 5 Expectations 7 Views about treatment

2 4 6 8

Goal setting Self-appraisal Behavioural consistency Self-efficacy

Transitions to Better Lives

9 Dissonance 11 Affective Com ponent Interpersonal style ratings 1 Procrim inal view s 3 Grandiosity 5 N eutralisation 7 Procrastination 9 Pow er and control 11 Victim stance Treatment performance ratings 1 Know ledge of program m e content 3 Disclosure 5 Know ledge application 7 U nderstanding of crim inality 9 Insight 11 Disruptiveness 13 Depth of em otional understanding

10 External supports

2 Procrim inal associations 4 Callousness 6 Im pulsivity 8 M otivation for anger 10 Problem -solving

2 Skills acquisition 4 O ffender confidence 6 Skills application 8 M otivation 10 A ttendance 12 A ppropriateness 14 Participation

The treatm ent readiness and interpersonal style dom ains of the initial assessm ent protocol were developed to reflect 11 factors. Each factor (for instance problem recognition) had two representative indices (problem acknow ledgem ent and problem understanding) w ith behaviourally anchored ratings using a four-point Likert scale and exem plars for each rating. This yielded 22 ratings for each subject for each of the dom ains of treatm ent readiness and interpersonal style. N either prim ary author (RS and SK) were directly involved in the d eliv ery m anagem ent or evaluation of standardised correctional program m es at the tim e this work w as initiated. Staff directly involved in delivering such program m es were approached regarding participation in research to develop an assessm ent protocol relating to offender change, but this w ork was not part of the Correctional Service of C anada's annual research plan. For these reasons, data collection was som ew hat sporadic and sam ples were of convenience. N onetheless, response from practitioners was favourable and sufficient data were collected to perm it refinem ents of the treatm ent readiness assessm ent instrum ent and an evaluation of its psychom etric properties (Serin et al. 2007). Initial correlations am ong item s were reviewed and factor analysis w as em ployed to reduce the initial 22item version to a m ore useful eight-item version, using data collected 18

The origins o f treatm ent readiness

on a sam ple of 262 federally incarcerated offend ers w ho participated in various correctional program s. T his abbreviated version has acceptable reliability (alpha = .83) and accounted for 45 per cen t o f the variance in the treatm ent read iness factor. Initial norm ative data regarding m ale offend ers (n = 268), w om en offend ers (n = 29) and sexual offend ers (n = 39) are available. It is these eight factors that are listed below and are described in greater detail later in this chapter. 1 2 3 4 5 6 7 8

Problem recognition B enefits of treatm ent Treatm ent interest Treatm ent distress Treatm ent goals Treatm ent behav iou rs M otivational consistency Treatm ent support

Essentially, we posit that these final eight item s reflect our effort at o p erationalisin g the con struct o f treatm en t read iness in offenders. T he follow ing excerp t from the user m anual presents our view of treatm ent readiness: Treatm ent readiness is a d om ain that captures an ind iv id u al's w illingn ess to engage in the treatm ent process. For som e, they see them selves as h av ing v ery few problem s that require therapeu tic intervention and d o not have any desire to m ake changes. These ind ividu als tend to be forced into treatm ent and are reluctant to pu t forth any effort into changing. O thers m ay be aw are of the problem s in their lives but are h esitan t to m ake a com m itm ent to change. O n the other extrem e are those w ho are com m itted to chang in g and are enthu siastic abou t m od ifying their behaviou rs. The treatm ent read in ess dom ain is intended to op erationalize this con tin u u m in an effort to assist clinicians in d eterm ining treatm ent placem ent. T his d om ain has excellent internal con sisten cy prod ucing an alpha of .83 in a sam ple o f 265 m ale offend ers entering a cog nitiv e skills program . The item s p roduced factor loadings in the v ery good to excellent range (.60 to .77) w ith a m ean o f .67.

19

Transitions to Better Lives

Description of treatm ent readiness items Problem recognition

Treatment readiness begins with an appreciation by the offender of the need to change som ething (friends, substance use, attitudes, problem solving) in order to avoid future criminality. Problem recognition assesses an offend er's aw areness that specific crim inogenic problem s do exist and have contributed to their involvem ent in offending behaviour. In this manner, intent need not be deliberate and crim inal behaviour could be perceived to be accidental or circum stantial; nonetheless, the offender accepts that there is a problem . Problem recognition also addresses the offend er's understanding of the im pact of these problem s (short- and long-term consequences, relation to crim e and other lifestyle variables such as financial, em ploym ent, family, and interpersonal relationships). Hence, this is more than the offender sim ply stating they have a problem. W here the offender believes that circum stances or other people are the sole cause of their problem s, this reflects an absence of problem recognition. O ur inclusion of this item wanted to address both recognition and accountability (H anby et al. 2009). The expectation w as that offenders w ho score higher on problem recognition would be more ready for treatment. The operational definition from the user manual is: 'Problem recognition assesses an offend er's aw areness that specific crim inogenic problem s exist. The first item considers only recognition of specific difficulties. This item assesses the offend er's appraisal of their current situation. This is assessed in terms of their understanding and ow nership of their problem s.' Benefits o f treatment

A nother im portant aspect o f treatm ent readiness is the offend er's recognition that there are potential benefits to treatm ent involvem ent. The expectation w as that offenders w ho score higher on seeing benefits for treatm ent would be more ready for treatment. Now, alm ost 15 years later, we would describe this item in term s of outcom e expectancies (Serin and Lloyd 2009): that is, offenders w ho expect that involvem ent in treatm ent will reduce future crim inal behaviour and that crim e desistance is a valued outcom e. Phrased differently, offenders w ith stronger beliefs in their ability to overcom e obstacles involved in giving up crim e through program m ing will more readily

20

The origins of treatment readiness

attem pt to give up crim e, continue to strive against setbacks and ultim ately succeed. The operational definition from the user m anual is: 'This item is intended to tap into an offend er's view s regarding the overall benefits of participating in treatment. An offender w ho describes the long-term benefits (lifestyle stability such as em ploym ent, relationships, no crim e) and short-term benefits (earlier release, fewer release conditions) of treatm ent w ould be assessed as recognizing the benefits of treatm ent.' Treatment interest

The next treatm ent readiness item is very sim ilar to the benefits of treatm ent item . The intent was to capture w hether the interest in treatment reflects both internal and external reasons. Those offenders who can identify that treatm ent participation will be of value to them selves as well as to others would be considered m ost ready for treatment, thus heading towards a path of crim e desistance. Individuals tend to adopt a particular style of explaining life events, w hether internal (caused by features w ithin the person) or external (caused by others, the environm ent or chance). W hen desisting from crim e, it may be most adaptive for offenders to reject blam e for their crim inal history and current problem s (external) w hile still holding them selves responsible for finding m ethods for staying crim e-free (internal) (M aruna and M ann 2006). O thers have noted the need to seek resources and a requirem ent for active effort in the change process (M oulden and M arshall 2005). Treatment distress

A lthough negative affect can be a precursor to crim inal involvem ent (Pithers et al. 1989; Zam ble and Q uinsey 1997), our experience and some research regarding individuals suffering from anti-social personality disorder (A lterm an and C acciola 1991) suggested that some distress w as also im portant in influencing offender com m itm ent to change. More recently, this view point has been supported by physiological assessm ent where cortisol levels w ere related to treatm ent response (Fishbein et al. 2009). Essentially, offenders who present as indifferent and lacking in som e em otional distress were hypothesised to be less ready than offenders w hose distress regarding their current circum stance prom pted them to consider change.

21

Transitions to Better Lives

Treatment goals

A nother aspect of our conceptualisation of treatm ent readiness is setting realistic treatm ent goals. G oal-setting assesses an offend er's ability to identify and realistically create treatm ent goals. This item considers the know ledge and skills necessary for treatm ent gain. Again, now adays we m ight incorporate the issue of outcom e expectancies (how effortful is treatm ent, how viable is the treatm ent goal, is the treatm ent goal som ething developed in conjunction w ith the offender or proscribed?). More recent w ork describing approach and avoidance goals in sex offenders (M ann et al. 2004) highlights the im portance of integrating the identification o f goals into a treatm ent plan in order to enhance program m e effectiveness. Moreover, it seem ed to us that alerting offenders to the effort required in treatm ent and the high probability of lapses should am eliorate program m e attrition and post-program m e failures due to unrealistic expectations. Treatment behaviours

It is rare that an adult offender w ho presents for treatm ent has never had prior experience regarding change efforts. N oting that prior history is often a good predictor of future behaviour, we wanted to include an item regarding treatm ent behaviours. This can reflect prior treatm ent experience of the initial two or three sessions of the current program m ing efforts. This item assesses the offend er's m otivation for treatment. Behavioural indication of good m otivation should reflect, where applicable, tim ely attendance at interview s a n d / or groups, hom ew ork com pletion, com pliance with prior treatment, a n d /o r positive com m ents about treatm ent as a process not an outcom e. Anecdotally, offenders som etim es present for treatment noting a prior positive rehabilitative effort (previously com pleted a group, or previously seen a psychologist). Yet w hen enquiries are m ade to retrieve greater details regarding these experiences, there is a disconnect. That is, the offender cannot rem em ber key learning points. Som etim es they cannot even recall the nam e of the staff member, despite having ju st explained how great the program m e was and how m uch they benefited from the experience! A t best this is a problem of transfer, at w orst an indication that the offender recalls little of the experience and w as sim ply telling you w hat they thought you wanted to hear.

22

The origins of treatment readiness

Motivational consistency

Related to this issue of m otivation is one of consistency. This item highlights the im portance of an offend er's verbal statem ents and their actions regarding treatment. If an offender has not previously participated in treatm ent then this item refers to behavioural consistency outside of treatm ent (such as m eeting a caseworker). O ffenders w ho state that they are motivated towards treatm ent, but show incongruence (in poor attendance, being late or infrequent attendance w ithout a legitim ate reason, failure to com plete hom ew ork, a n d /o r stating low m otivation to other staff or offenders) would be seen as lacking m otivational consistency. O ur preliminary' research has flagged an issue w ith this item. It is possible for an individual to be consistently lacking in m otivation. In this case we have suggested a score of '0' as the intent is that higher scores reflect greater treatment readiness. Treatment support

At the time that we created this item as part of the dom ain of treatm ent readiness, our clinical w ork inform ed the need to include som ething relating to aftercare and support. Some w ork regarding the im portance of pro-social m odels w as available (Andrews and Kiessling 1980), but it is only m ore recently that its im portance in understanding offender success has been dem onstrated (LeBel et al. 2008; M aruna 2001; M assoglia and U ggen 2007). Given that we perceive offender change as a process, this item assesses the degree of support for change by others significant to the offender. It is im portant to allow the offender to determ ine w ho is im portant to them (preferably family, friends, employer, or clergy) and then probe for degree of support from them. Change w ithout support is unlikely to generalize. Of note is that enhancing com m unity support underscores the current re-entry initiative in the United States (Burke and Tonry 2006).

Using the scale Follow ing the initial interview, the item s are scored using a threepoint scale and aggregated. Self-reference questions, behavioural anchors, and questions to pose to the offender are all provided in the user m anual. For purposes o f com parison, norm ative pre- and

23

Transitions to Better Lives

post-treatm ent readiness scores are provided. For offenders w ho score below average (one standard deviation below the m ean), we suggest individual prim er sessions to overcom e obstacles to treatment engagem ent follow ing the principles o f m otivational interview ing (A nstiss et al., in press).

Summary Throughout the pilot research the feedback we received from program m e staff w as very positive. Perhaps the absence of an earlier model of treatm ent readiness augm ented its inform al use. A s well, program m e staff found the assessm ent protocol helpful in structuring their post-treatm ent reports in that they could system atically com m ent on changes w ith respect to treatm ent readiness, interpersonal style and program m e perform ance. As noted earlier, this m odest effort has yielded a user-friendly and short behavioural rating scale that appears to capture m eaningful aspects of treatm ent readiness. M ore recent research has supported the im portance of m any of the item s and underscores the utility of an assessm ent of treatm ent readiness prior to program m e involvem ent.

Empirical support The results from this pilot research w ere encouraging in that the revised eight-item treatm ent readiness rating scale appears reliable; it appears sensitive to change and prelim inary norm s are now available. However, the absence of inter-rater reliability and concurrent and predictive validity data are disconcerting. For now, sites choosing to use the treatment readiness scale clinically should provide sufficient training with the m anual to ensure inter-rater reliability. N onetheless, interest from other colleagues in other countries has led to some encouraging findings. Unpublished research suggests that the dom ains of treatm ent readiness and interpersonal style differentiate am ong offenders in term s of program m e attrition in the United K ingdom (Watson and Beech 2002) and Canada (Stew art 2005). There is also unpublished evidence that treatm ent readiness is correlated to recidivism in a sam ple of offenders in H ong Kong (Lee 2005). The eight-item version of the treatm ent readiness scale has also been used to distinguish am ong types of sexual offenders (M alcolm 2002), but was not predictive of program m e drop-out by sex offenders due to 24

The origins of treatment readiness

exceedingly low base rates (Latendresse 2006). H owever, this latter research confirm ed the strong relationship betw een total scores on the original and revised versions.

Implications and new directions Conceptually, we view treatm ent readiness as requisite for offender change and the strategies proposed for m anaging high-risk offenders seem appropriate for engaging offenders who lack readiness for treatm ent (A nstiss et al., in press; Polaschek 2009; Serin and Preston 2001). N onetheless, som e offenders arrive at treatm ent ready to change, so there will be heterogeneity am ong offenders regarding readiness. A review of current literature suggests the concept of treatment readiness rem ains popular and is particularly salient in understanding violent offenders (Day et al. 2009a). Indeed, through num erous papers (Cham bers et al. 2008; Casey et al. 2007; Day et al. 2007) these authors have single-handedly refined and underscored the construct as it applies to violent offenders. M oreover, they have applied the construct of treatm ent readiness to broader m odels of offender change (Day et al. 2006; Ward et al. 2004b), suggesting its utility over other m odels such as the transtheoretical model of change (Casey et al. 2005). With the exception of this recent research regarding treatm ent readiness in violent offenders, som e 15 years later, various terms continue to be used interchangeably - treatment m otivation, treat­ m ent engagem ent, and readiness for change. Indeed, in m any respects the field has m ade relatively little progress in understanding offender change and dem onstrating its ability to m easure such change, although its im portance is certainly underscored (M cM urran and Ward 2004). It is clear that clinicians continue to struggle with the assessm ent of these constructs around readiness and treatment engagem ent, and yet such constructs are critical to our understanding and predicting offender change. A s a brief behavioural assessm ent, we posit that our treatm ent readiness scale m ay have potential m erit in assisting clinicians to assess treatm ent engagem ent and the need for pre-treatm ent prim ing. We also believe that a full testing of our model and alternative explanations of offender change is required. As assessm ent efforts continue, it m ay be that such a scale and model could inform the developm ent o f an assessm ent protocol. Since our early research (Kennedy and Serin 1997; Serin and K ennedy 1997), this work has evolved such that we now conceptualise 25

Transitions to Better Lives

offender change, of which treatm ent readiness is an im portant com ponent, to be part of a broader transition by the offender (Lloyd and Serin 2009; Serin and Lloyd 2009). The focus on this transition is m arkers of interpersonal change in support of successful desistance from crim e. We also view treatm ent readiness as an im portant com petency, one that is requisite for offender change and crim e desistance (H anby et al. 2009). Research regarding these refinem ents to the role of treatm ent readiness is ongoing, as is research regarding im provem ents to the assessm ent of offender change.

26

Chapter 3

W hat are readiness factors?

In this chapter we describe in som e detail those factors that have been identified in the M ultifactoral Offender Readiness M odel (M O RM ), introduced in Chapter 1 of this book. We have adopted the distinction made in the model betw een internal and external readiness factors, and will begin w ith a review of internal factors, as these are most com m only associated with current understandings of treatment readiness and, indeed, attem pts to m odify low levels of treatm ent readiness (see Chapter 10).

Internal readiness factors It is those core psychological features that enable offenders to function in a therapeutic context that have attracted m ost of the attention of researchers and clinicians. These are referred to in the M ORM as internal readiness conditions and may be either cognitive (beliefs, cognitive strategies), affective (em otions), volitional (goals, w ants or desires), or behavioural (skills and com petencies). In this chapter we consider each of these factors in turn. In m any ways, it is these characteristics of individual offenders that have attracted the most interest and discussion in relation to both treatm ent readiness and responsivity. Ways of m easuring these readiness factors are discussed later in this book (see Chapter 5), although they are also discussed in relation to violent offenders, sexual offenders, offenders w ith personality disorders, and those w ith substance abuse disorders. It is also these internal readiness factors that are m ost relevant to the 27

Transitions to Better Lives

clinical and therapeutic approaches to w orking w ith low levels of readiness discussed in Part Three. Cognitive factors

Self-efficacy is one aspect of cognition that has attracted a great deal of attention in relation to treatm ent engagem ent. It can be understood as the self-appraisal about how w ell one can perform actions to deal w ith a situation (Bandura 1997), and is thought to be related to both enhanced m otivation and perform ance (Bandura and Locke 2003). Cham bers and colleagues describe it in the follow ing way: 'Low selfefficacy leaves the individual believing that h e /s h e is unable to learn new social skills, or alternate w ays of life, and is thus unlikely to attem pt new behaviors' (2008: 281-2). General beliefs about personal change m ay also influence the likelihood of an individual offender identifying a need for treatm ent and engaging with a rehabilitation programm e. For an individual to take steps to change, he or she must not only believe that the benefits of the action outw eigh the barriers, but also experience some trigger to take action, and (in relation to health prom otion behaviours), believe that he or she is susceptible to the condition and view the condition as serious (Chew et al. 2002). In terms of offender rehabilitation, this suggests that the offender would need to see his or her offending as likely to recur (that is that s /h e is susceptible), that offending is a serious problem , and that the costs of change do not outw eigh the benefits (for exam ple not associating w ith friends who are likely to offend) if he or she is likely to seek out referral to a rehabilitation programm e. More specifically, offenders often hold particular attitudes about treatm ent in the crim inal justice system that can influence the way in w hich they present to program m e providers (Baxter et al. 1995). Expectations about what will happen in program m es will also influence readiness. These can arise from previous experiences of treatm ent, the experience of the assessm ent process, or the reputation that program m es and even individual program m e staff m em bers have in a particular institution. This area has received surprisingly little attention in the literature, although studies that have been conducted suggest that, generally, offenders are likely to hold quite negative view s towards crim inal ju stice agencies (Lyon et a l 2000), and are unlikely to approach correctional staff for help, particularly for em otional support (Dear et al. 2002; H obbs and Dear 2000). O ffenders also exhibit a num ber of attitudes, beliefs and thinking styles that potentially either reduce or increase their w illingness to

28

W hat are readiness factors?

engage in a rehabilitation program m e. For exam ple, the existence of hostile attitudes and beliefs can result in a tendency to view the world and the actions of others in a negative or cynical fashion, m aking it difficult to accept that therapists will behave in a trustw orthy m anner and deliver the kind of interventions that were initially prom ised. Such beliefs can be understood not only as dynam ic risk factors, or crim inogenic needs (given the association betw een particular thinking styles and behaving in anti-social or aggressive w ays), but also as factors that will im pinge upon an individual's ability to engage w ith a rehabilitation programm e. Cham bers et a l (2008) have discussed how holding particular beliefs about an offence m ay m ake it less likely that the individual will regard treatm ent as necessary, as w ell as how more general, or trait, aspects of cognition potentially influence treatm ent readiness. For exam ple, w hat Gibbs, Potter and G oldstein (1995) have referred to as prim ary self-serving cognitive distortions (self-centred attitudes, thoughts and beliefs that are m anifested as a belief in one's own view s, needs and expectations to the extent that the view s of others are inconsequential or totally disregarded), and secondary selfserving cognitive distortions (such as blam ing others, m in im isin g/ m islabelling and assum ing the w orst o f others) m ay both reduce levels of treatm ent readiness. Cham bers and colleagues suggest that strong prim ary cognitive distortions can seriously im pede readiness because the offender believes that he or she already 'know s it all' and, as such, does not see the need for personal change. Secondary distortions support the prim ary distortions by rationalising and justifying the offending behaviour and serving to protect the offend er's self-im age follow ing their anti-social behaviour (see Table 3.1). Affective factors

H ow ells and D ay (2006) have suggested that it is necessary for an offender to have a basic level o f em otional control to successfully enter a treatm ent programm e. Serran and colleagues (2003) concept of em otional responsivity identifies three inter-related com ponents of em otion that are thought to im pact upon levels of readiness: the access the offender has to em otional states; his or her ability to express such states; and his or her w illingness to do this in the therapeutic session (see Table 3.2). Subjective distress is not only likely to be an indicator of problem severity, but from both a clinical and a theoretical perspective is likely to have m otivational properties (Frank 1974). W hen psychological

29

Transitions to B e tte r Lives

T ab le 3 .1

The potential im pact of cognitive distortions on treatm ent

readiness. C ognitive d istortion

Potential im pact on readiness

Self-centred attitudes and beliefs

M ay result in little value bein g assigned to the view s o f others and even their ow n long-term interest. This d istortion w ould im pact readiness for rehabilitation becau se these v iew s are likely to foster reactance to suggestions of change.

Blam ing others

M ay lead the offender to believe that the offending w as not their fault, and that they are not responsible for their actions. This distortion im pacts on readiness for rehabilitation because an offender w ould believe that their actions are not contingent on controlling their own behaviour. By assign ing an external locus of control, these offenders believe that they do not control their offending behaviou r; therefore they are unable to change it.

M inim ising and m islabelling offending behaviour

Can lead to the b elief that behaving anti-socially is ad m irable and even a service to the com munity. C hange to w hat is seen as a positive behaviour, w hich m ight depict their identity, w ould be difficult to achieve.

A ssum ing the w orst

M ay enable anti-social behaviour through ascribing hostile intent to others. Since this is the way a hostile individual view s people in general, it is highly likely that they will view the therapist w ith hostility, thus hindering engagem ent in the program m e.

Source: Adapted from Chambers et al. (2008).

distress is experienced, an individual may contemplate behaviour change or become resistant to change. In their review of treatments for depression, Beutler, Clarkin and Bongar (2000) concluded that treatments achieve their greatest effects among those patients who present with moderate to high initial levels of subjective distress. The level of distress with which an offender presents during the reception and screening assessment process may, therefore, be relevant in making decisions about program m e referral, although we are not

30

W h a t are readiness factors?

T ab le 3.2

A ffective readiness factors.

Factor

Description

A ccess to, and experience of, em otional states

A n offender w ho is unable to access em otional states m ay be unable to engage in treatm ent by virtue of bein g unable to acknow ledge and describe past em otional experiences. G iven the dem onstrated role of strong em otions in the offence chains for both v iolent and sexual offending, it is essential that such offenders are, or becom e, able to access and experience such em otions if they are to understand and m odify their pathw ays to offending. W hile there m ay be a num ber of reasons w hy som e offenders do not have access to past em otional states, som e level o f negative affective arousal m ay b e necessary to m otivate som e offenders to engage in treatm ent, given that the goal, a l least for som e offenders, in attend ing treatm ent is likely to be distress or anxiety reduction. O f course, attend ing a program m e m ay also be a strategy to achieve non-affective goals (such as consid eration for parole). A p athological lack o f affect w ould also constitute a barrier to engagem ent. Support for this suggestion com es from observations that psychopaths perform poorly in treatm ent (for exam ple Sm ith, 1999). Psychopathy, as m easured by the PC.L-R and as understood by clinicians, is prim arily a problem of affective deficit, given that shallow ness of affect and lack of guilt and rem orse are defining features of this condition.

Direct expression o f affect

The direct expression of affect refers, for exam ple, to w hether the offender w ho currently feels strong fear, anger or non-specific distress displays these em otions in the therapeutic situation (obvious apprehensiveness, verbal aggression, tearfulness, self-disclosure). For N ovaco et al. (2001), effective treatm ent for v iolent offenders entails the evocation of distressed em otion - it follow s that an ability to express em otion is necessary if treatm ent is to be effective. W h y this should, in itself, enhance engagem ent is not obvious, Table 3.2 continues overleaf

31

Transitions to Better Lives

Table 3.2 continued Factor

Description except insofar as expression m ay som etim es assist the individual to acknow ledge the extent and intensity of their em otions, or, as Serran et al. (2003) have suggested to directly activate latent offence-related schem as. It is also possible that the direct expression of em otion is functional not for the client, but for the therapist, w hose know ledge about the client's subjective states is improved by the opportunity to observe the client.

The w illingness of the client to adm it to, and reflect on, their experience and expression of em otions in the therapeutic session

To be unw illing, or unable, to acknow ledge em otional reactions in front of others in a therapeutic situation (for exam ple a rehabilitation group) is to preclude the possibility of w orking therapeutically to understand and modify these em otions. Em otions form one dom ain of potentially painful experiences w hich m ay be disclosed or not (another being details of the offence). In prison environm ents, self-disclosure is a particularly sensitive topic. Disclosure of personal inform ation or about offences may place the individual at risk from other prisoners and offenders m ay also be reluctant to disclose personal issues to prison staff.

Source: Adapted from H ow ells and Day (2006).

aw are o f any em pirical research that has directly addressed this question. A nother potentially im portan t (and related) com ponent of read iness w ill be the in d iv id u al's em otional reaction to both their offence and their status as 'o ffe n d e r'. T he relation ship betw een distress that is attributed directly to the offence (guilt, sham e, rem orse; see below ), and d istress that is attributed p rim arily to the con sequ ences o f im p risonm ent (cop ing, ad justm ent) is another area that w arrants m uch further investigation. It m ay be, for exam ple, that pre-treatm ent levels of d istress w ill only b e a significant eng agem ent factor w hen certain attributions are m ade about the causes of that distress. G iven that m ost offend er rehabilitation program m es target m edium to high-risk offend ers, and giv en the high w eighting o f previous 32

W hat are readiness factors?

offences in m ost m ethods o f assessing risk, the great m ajority of offenders being assessed for program m es will have established histories of offending. The conferral of the status 'offend er' for these participants will be long-standing. There will, however, be others w ho are referred follow ing one-off offences: those w ho have no significant history of offending, such as som e perpetrators of dom estic hom icides. Intuitively, it m ight be predicted that individuals in this group would have stronger negative, em otional reactions to their offence and to their offender status. A m ong such offenders, em otions of guilt, shame and rem orse may be particularly com m on. O nce again we know of no em pirical studies investigating potential differences in these em otions in one-off as opposed to repetitive offenders, though the high incidence of subsequent suicide in dom estic hom icides would suggest that there are stronger guilt, sham e and rem orse experiences in this group (West 1965). In order to assess the potential influence of offence-related em otions such as guilt, sham e and rem orse on readiness for treatment there are two requirem ents. First, these em otions need to be defined and differentiated. Second, possible m ediating m echanism s need to be identified. Proeve (2002) has suggested that the cognitions that characterise guilt and sham e have different foci. In guilt, the individual's focus is on the act, w hile in sham e it is on the self (Tangney 1999).1 It has been suggested (for exam ple Lew is 1995; Proeve 2002; Proeve and H ow ells 2002) that sham e can be distinguished from guilt in terms of the self-evaluative com ponents o f em otion. A lthough guilt m ay involve focusing on aspects of the self that lead to the transgression, typically the self is not negatively evaluated in a global way as is the case w ith sham e where the self is seen as inferior, incom petent, or otherw ise bad.2 Thus it would appear that shame and guilt can be distinguished to some degree in terms of their phenom enology and accom panying cognitive and behavioural processes. There are also indications that shame is more strongly associated with other psychological variables that m ight im pair readiness than is guilt. Sham e proneness is associated with

■Guilt is also thought to involve a focus on the negative consequences of the act for others and an accompanying belief that the individual has violated a personal, moral standard. A number of action tendencies have been described for guilt, including apologising, undoing damage and attempting to repair the damage done. 2Shame also involves an awareness of judgement of the self by others, of the defectiveness of the self in the gaze of the observing other (Taylor 1985). 33

Transitions to Better Lives

low empathy, anger, irritability, externalisation, blam ing of others for negative events, resentm ent, suicidal behaviour and psychopathology, w hereas guilt tends to show an inverted or no relationship to many of these variables (Bum by et al. 1999; Tangney 1991). The action tendency associated with sham e appears to be hiding oneself from others, w hereas confession and reparation are more salient for guilt. In this context, higher levels of guilt m ay be associated with increased levels of em otional self-disclosure. In the above discussion, it is assum ed that offenders identify their offending as the problem to be addressed. It is also assum ed that offenders are genuinely distressed about their offending: that is, that their offending is ego dystonie rather than syntonic. These are little m ore than assum ptions, and there is a clear need for em pirical investigations on this topic. There are also likely to be a num ber of client characteristics that m oderate the effects of each of the three em otional responsivity factors outlined above. Individuals differ in terms of the extent to w hich they experience em otions, express em otions and generally describe their experiences in em otional terms. Gender, gender role stereotypes, cultural values, and personality factors (such as a pervasive lack o f trust in others) are all relatively stable characteristics that w ill influence the level of treatm ent engage­ m ent (see Table 3.3). In summary, affective factors appear closely related to m otivational structures and consequently the desire to enter a behaviour change program m e. However, these factors rem ain poorly understood, difficult to assess reliably, and not w ell investigated empirically. There would appear to be a strong case for further research exploring the association betw een different affective states and styles and treatment readiness. Personal goals, volition, and identity

Volition refers to the form ation of an intention to pursue a certain goal and the developm ent of a plan to achieve the goal in question. In the crim inal justice context, the exercise of volition requires the ability to consent to treatment program m es, and to m ake informed decisions about participation (Birgden and Vincent 1999). There is also therapeutic value in choice; individuals are more likely to be m otivated if they are offered alternative courses of action from which to choose. Typically, in the offender rehabilitation literature, volitional factors have been understood in term s o f m otivation to change. Levels of

34

W h a t are readiness factors?

T ab le 3.3

Potential m od erators of Ihe effects of affective readiness factors.

Factor G ender and gender role stereotypes

C ultural values

Description M en and w om en m ay differ in display rules and in the influence of social contexts on em otional expression, such as the gender o f other people in the im m ediate environm ent. This m ay have im plications for em otional responsivity in m ale offenders in therapeutic groups. M ale violent offenders m ay b e low in em otional responsivity b y virtue, in part, of their m ale gender and en su in g gender-role expectations. This low em otional responsivity w ould be further exaggerated if m ale gender-role expectations were particularly extrem e and stereotypical in this client group. It is possible, of course, that gender differences in em otion are in part substantial and in part secondary or constructed. There are likely to b e cultural differences in the experience and expression of em otions. These m ay relate to either categories available w ithin language to label em otions, or to the secondary appraisal process. Averill (1983), for exam ple, has argued that culture not only effects w hen it is appropriate to be angry, bu t also the extent to w h ich violence is an understand able response to anger. In ad dition, there are likely to be significant cultural (and subcultural) influences on the extent to w hich em otional disclosure is regarded as acceptable. Status effects, for exam ple, are also likely to im pact upon levels o f disclosure in treatm ent. D isclosure can m ake the speaker vulnerable and dem onstrate both trust and subm ission to the other. Research in other settings has show n that disclosure is often asscxiated w ith social inferiority. M any prisoners, particularly those convicted of v iolent offences, will be occupied w ith issues of status and power, and m ay see disclosure as inconsistent w ith their self-identity.

Source: A dapted from H ow ells and D ay (2006).

35

Transitions to Better Lives

m otivation have been shown to be consistent predictors of retention in substance use program m es, and client m otivation has also been related to better treatm ent engagem ent in such program m es. Joe, Sim pson and Broom e (1998) concluded that indicators of intrinsic m otivation were more im portant predictors of engagem ent and retention in treatm ent than other socio-dem ographic variables. Client m otivation is thus only one of a broader range of responsivity variables considered in offender rehabilitation and in psychological treatm ents. N evertheless, m otivation is a vital construct to consider. H um an m otivation has a long history as a topic of theoretical debate w ithin psychology. Ford (1992) has attem pted to integrate a range of findings and concepts from the field of m otivation into a M otivational System s Theory fram ew ork. M otivation from this perspective is defined as the 'organised patterning of an individual's personal goals, em otions and personal agency beliefs' (Ford 1992: 78). Thus m otivation involves directedness (tow ards the goal), em otional/affective energising, and expectancies about being able to achieve the goal. The absence of m otivation could be due to any of these three com ponents being deficient. Thus a client m ight be unm otivated for therapy because the therapeutic goal is not im portant to them, because there are em otional or affective inhibitions to goal pursuit, or because there is a perceived low capability of achieving the relevant goal. Karoly (1993, 1999) has applied goal system constructs to clinical phenom ena and to treatment. He defined goals as 'im agined or envisaged states tow ards w hich people intentionally aspire and actively w ork to bring about (or to avoid, in the case of negative goals) (1999: 274).3 Personal goals can be conceptualised at three different levels: the latent, the phem onenological and the external observer (Austin and Vancouver 1996). Latent goals m ay be outside of phenom enal aw areness w hile phenom enological goals are experienced directly and are (presum ably) capable of self-reporting. External observers m ay infer goals from features of the individual's behaviour.

3Previous researchers have used different terms to describe goals of this sort. Emmons (1999), for example, investigated 'personal strivings' while Little (1983) described 'personal projects'. Ford (1992) distinguishes three types of goals, based on their level of prioritisation by the cognitive regulatory system: wishes, current concerns and intentions. Personal goals are thus cognitive representations and potential self-regulatory mechanisms by means of which behaviour is activated and coordinated. 36

W hat are readiness factors?

Personal goals need to be understood in terms of content. W hat does the individual want? The em phasis in goal system theory tends to be on tangible, task-specific incentives rather than on broad, higher-order incentives, such as control or com petence (Karoly 1999). Other dim ensions of goals in K aroly's fram ew ork include goal topography, structure, process representation, dynam ics, m odality of representation, procedural predispositions, m indset effects, social context effects and interface w ith em otion. Goals are w idely believed by researchers and theorists to be hierarchically organised, cascading from higher order goals to goals at the level of local and briefly experienced psychophysiological states. Ford (1992) has attem pted to define the necessary conditions for achievem ent and com petence from a goal system s perspective. For the purposes of the present discussion, we can substitute 'achieving rehabilitation goals' for ach ievem en t/com p eten ce, the latter terms having been used because o f the edu cational/d evelopm ental context of much goal system research. Ford's analysis, w hen extended to the field of readiness for treatm ent, would suggest the follow ing as necessary conditions for achieving therapeutic goals: 1 Personal goals are constituted by, supportive of, or consistent w ith the therapeutic goal. 2 Em otional states are congruent w ith therapeutic goal pursuit and achievem ent. 3 Capability beliefs are present (regarding the achievability of therapeutic goals). 4 Positive context beliefs are present (perceived supportiveness of the environm ent in achieving therapeutic goals). 5 Actual cap ab ility /sk ill exists. 6 Actual environm en tal/contextu al support exists. Issues of treatm ent m otivation have been directly addressed by goal theorists, albeit briefly. Karoly, for exam ple, stresses the need to exam ine treatm ent targets in the context of broader client goals and the m otivational salience of change. H e also suggests that 'therapeutic failures of various kinds (prem ature term ination, resistance, relapse etc) can result from the therapist-assessor's failure to appreciate the structural relation betw een tim e-lim ited treatm ent goals and life goals in general' (1993: 279). Karoly (1999) m akes the point: A ssum ing that a 'therapy goal' represents a to-be-achieved destination, it m ust be borne in mind that the instantiation of 37

Transitions to Better Lives

any new trajectory or pathw ay is alw ays accom plished in the context of existing and projected pathw ays and hierarchically distant aspirations. Therapy goals that help achieve, or are consistent w ith, m eaningful higher order goals stand a better chance of long-term success than do therapy goals that are at odds w ith higher order goals or values. (1999: 264-5) In summary, goal system s theory would suggest that the offender who has goals that are incom patible w ith im plicit or explicit rehabilitation goals will be low in readiness. Additionally, the absence of effective self-regulative strategies and processes in relation to goal attainm ent would also be a determ inant of the occurrence of the presenting clinical problem (for exam ple anger difficulties) and would, in turn, form an im pedim ent to effective change in therapy. The goal system perspective indicates that determ ining the client's goal structure and associated self-regulative skills are key com ponents of pre-treatm ent readiness assessm ent. There have been two bodies o f w ork in the field of offender rehabilitation that consider the role that the personal goals of offenders m ight play in behaviour change program m es. The first focuses on the personal concerns of offenders and is discussed by M ary M cM urran in Chapter 11. The second is a broader rehabilitation theory, the Good Lives M odel, w hich suggests that a focus on the prim ary goods sought by an individual offender is likely to improve both treatm ent engagem ent and program m e outcom es. This approach is discussed in Chapter 4. A num ber of other constructs overlap w ith the notion of personal goals. The subject factor of personal identity, for exam ple, is one that is likely to be of particular im portance. In seeking to achieve goals, we construct personal identities (Em m ons 1999), a term that refers to the kind of life sought, and relatedly, the kind of person that an individual would like to be. Ward et al. (2004b) suggest that, for exam ple, if an offender decides to pursue a life characterised by service to the com munity, a core aspect of his or her identity will revolve around the prim ary goods o f relatedness and social life. The offend er's sense of mastery, self-esteem , perception of autonom y and control will all reflect this overarching good and its associated sub­ clusters of goods (such as intimacy, caring, honesty). They argue that the im portant issue for readiness is that an individual's personal identity m ust allow for the possibility of an offence-free lifestyle (and includes the possibility of change) and is not based too strongly on

38

W hat are readiness factors?

being an 'offend er' (also see the reference to practical identities and values in Chapter 4). Of course, personal, cultural and social needs all im pact on personal identity and, as such, should also inform program m e developm ent and delivery. That said, rehabilitation program m es are often developed in w ays that do not cater well for the needs of particular groups such as culturally and linguistically diverse clients, and those w ith low er incom es and less education. Behavioural factors

Behavioural factors include possession of the basic com m unication and social skills necessary to participate successfully in treatment. These are assumed to be necessary conditions for entering a rehabilitation program m e and are not to be confused w ith the skills that are expected to be acquired during treatment. In other words, if an individual is unable to initiate and m aintain basic conversations with others, then he or she is not 'read y' for treatment. The nature of the required skills or com petencies will depend on the m ethods of delivery and the content of the particular treatm ent program m e. A program m e with a large educational com ponent w ill require different com petencies (such as literacy) than will a program m e based prim arily on roleplaying and rehearsal of core behavioural skills (confidence in group setting) or a program m e w ith a significant com ponent of intellectual analysis of the antecedents for the person's offences (verbal ability, capacity to discuss thoughts, feelings and behaviour in front of other group m em bers). Given the prevalence of mental disorders (H odgins and M ullerIsberner 2000) and intellectual disability (Day 2000) in offender populations, it is also im portant to note that the existence of mental illness or intellectual disability may functionally disable these im portant core skills and prevent the individual concerned from successfully functioning in groups or having the necessary attention and concentration abilities to acquire new skills. A lthough these may be regarded as both affective and cognitive readiness factors, the negative sym ptom s associated w ith conditions such as schizophrenia and mood disorders (even when the disorder is in rem ission) may m ean that the individual is unable to arrive at program m e sessions on time, sit for an extended period w ith concentration on group activities, engage and em pathise w ith the problem s expressed by other group participants, or organise and carry out hom ew ork tasks.

39

Transitions to Better Lives

External readiness factors The focus of the above discussion has been on internal readiness factors - those characteristics o f offenders, such as motivation, beliefs about treatm ent, em otional regulation styles, that are likely to influence perform ance in a rehabilitation program m e. We now turn our attention to those readiness factors that lie outside of the individual offender. By external factors w e are referring to those characteristics of the environm ent, or the context, in w hich rehabilitation is offered that im pact on an individual's ability to engage w ith a particular program m e. A lthough som e of these factors m ay appear obvious, they are often overlooked in term s of their potential im pact on program m e outcom es. For exam ple, rehabilitation program m es need to be available w ithin the agency or institution in which the offender is located. Sentence planning requirem ents som etim es mean that prisoners are moved to locations in which program m es are not offered, perhaps as a result of limited bed space. W hether a program m e is offered in an institutional or a com m unity setting will have im plications for w hether certain skills can be m eaningfully taught. Location will influence the degree to which fam ily m em bers can visit regularly and support the rehabilitation process. The encouragem ent of other prisoners, prison officers, com m unity corrections officers, and clinicians for the offender to enter a specific program m e m ay also be critical. If offenders are placed in prison units characterised by guarded ness and suspicion, they m ay be less likely to volunteer to participate in program m es that they know will require self-disclosure and openness. In addition, the existence of rewards for successfully com pleting a rehabilitation program m e is also potentially significant. In som e jurisdictions prisoners are only considered eligible to apply for parole if they can dem onstrate that they have addressed the causes of their offending by participating in and com pleting a program m e. Program m es also need to be adequately resourced. D espite the considerable need for treatment, there is evidence to suggest that existing offender rehabilitation program m es tend to be over-burdened, and that m any offenders receive either very lim ited treatm ent or none at all. The ability to provide m eaningful treatm ent program m es requires the presence of skilled and trained staff and the physical resources necessary to run the program m e in a given setting. In our view, these all represent im portant readiness factors. Factors such as when program m es are offered, for exam ple in relation to stage of sentence, may also play an im portant role in 40

W hat are readiness factors?

w hether offenders see an opportunity for treatm ent participation. Initial contact with the crim inal ju stice system (conviction and sentencing) will for som e (but by no m eans all) be a critical event that leads the offender to reflect on the need for change. Those w ho are approaching release from prison m ay not have sufficient time to com plete a program m e and consequently not be offered one; conversely, they may be more ready to participate as the possibility of early release on parole draw s closer. Alternatively, offenders m ight prefer to w ait so they can access com m unity treatm ent program m es on a voluntary basis. Those beginning long sentences may feel that they will have m ultiple opportunities over the course of their sentence and as such feel no im m ediate pressure to attend. In summary, then, it is suggested that the extent to w hich a person is ready for treatm ent will extend beyond his or her psychological characteristics and those of the treatm ent itself. The focus should be ju st as m uch on w hether the environm ent is right for successful rehabilitation to take place as on individual factors such as m otivation to change. Too often, in our experience, rehabilitation program m es are offered w ithout adequate planning, resourcing, or adequate consideration of the context in w hich they take place. We would suggest, however, that by far the m ost significant external com ponent of treatm ent readiness lies in relation to the level of coercion placed upon the individual to attend a particular program m e. This is included here as an external readiness condition as m any offenders are pressured, if not legally obliged, to participate in rehabilitation program m es independently of their desires to do so.

Level of coercion The use of the crim inal ju stice system to force offenders to receive psychological treatm ent is one of the m ost controversial aspects of service provision. As Day, Tucker and H ow ells (2004) have observed: W hilst many practitioners are uncom fortable w ith the perceived infringem ent of civil liberties associated w ith enforced treatm ent, coercion is not inherently unethical. Paying taxes and obeying laws related to dangerous driving are also coerced, but these form s of coercion are accepted by m ost as prom oting the general good. Indeed, the legal system and the act of im prisonm ent itself are based on coercion. (2004: 259) 41

Transitions to Better Lives

M arlow e and colleagues (2007), in their discussion of com pulsory substance use treatm ent, further suggest that the intrusion of a judge into the treatm ent process can be disruptive or harm ful to the developm ent of a therapeutic relationship. They suggest that 'clients may be hesitant to confide im portant inform ation to their counsellors for fear the inform ation would be disclosed to the court and used against their legal interests', and that 'being "treated like a crim inal" by being brought into court on a regular basis might elicit counterproductive feelings of resistance or reactance' (2007: S5). Their conclusion is that judicial m onitoring can elicit iatrogenic effects. N evertheless, coercing offenders to attend rehabilitation program m es is increasingly accepted as an appropriate course of action, particularly for those w ho are regarded as posing a continuing threat to public safety. It is som etim es seen as the only effective m eans to ensure that offenders attend program m es (Burdon and Gallagher

2002 ). Coerced treatm ent needs to be distinguished from pressured treat­ ment, although both have objective and subjective dim ensions. O ne of the difficulties w ith any consideration of the effects of coercion is the lack of consistency w ith w hich the term is defined. For exam ple, it is com m only assum ed that because a program m e is m andated, there is a high level of coercion, and yet when asked participants may report that they do not feel coerced. Coerced or com pulsory treatm ent can be defined from three different perspectives: legal social controls (civil com m itm ent, court-ordered treatment, diversion-to-treatm ent program m e; see W ells-Parker 1995; Wild 1999, 2006); form al social controls (m andatory referral to em ployee assistance program m es providing addiction treatment, usually follow ing em ployer drug testing; see Law ental et al. 1996); and inform al social controls (threats, ultim atum s initiated by friends and fam ily m em bers; see Room et al. 1991; H asin 1994). D ay et al. (2004) suggest that: Coercion and pressure are not sim ple objective facts. The person may feel coerced and pressured into treatm ent even when there is no objective requirem ent to engage in treatment. Equally, the person may be objectively coerced (for exam ple by a court) but have little subjective sense of being coerced (for exam ple, when treatm ent is congruent w ith their ow n goals and aspirations). The term coercion often im plies being forced to do som ething against one's will, and includes an im plicit evaluative com ponent that com pliance will in som e w ay be unpleasant or aversive. Being pressured into a course of action is a sim ilar concept, although 42

W hat are readiness factors?

w ith pressure the person will be able to exercise a higher degree of choice about com pliance. (2004: 260) Typically offender rehabilitation program m es are coercive in the sense that there are negative consequences for non-participation in treatm ent, although the nature of these consequences m ay vary significantly across jurisdictions. O ffenders w ho refuse treatment may, for exam ple, find it to be im possible to reduce their security classification, which in turn m ay have an im pact on the living conditions and freedom s they will experience during their im prisonm ent. Grubin and Thornton (1994) reported that nearly half (41 per cent) of sexual offenders in the UK said that they would participate in treatm ent only in order to gain parole. H utchins (2003) has described the process as follows: 'if the offender does not com plete the program satisfactorily, parole is virtually routinely refused w hen the offender is first eligible. If refused, a date is fixed by the Board w hen parole will be considered again and it is suggested to the inm ate that h e /s h e do a ... program whilst in custody and before the next hearing' (2003: 1). An im portant consideration here is w hether coercing offenders into attending program m es (or placing legal pressure on them to attend) is likely by itself to lead to low er levels of treatm ent readiness. In their review, Day et a l (2004) suggest that it is the individual's perception of coercion that is m ore likely to determ ine treatm ent readiness, and even when clients do perceive that they are being coerced, then pre­ treatment anti-therapeutic attitudes can change over the course of a program m e. D ay et al. (2004) have argued that given that perceptions of coercion will not alw ays reflect the objective situation, it becom es im portant to identify when and how legal pressure is perceived as coercive. They suggest five factors that m ight influence the extent to w hich legal pressure is perceived as coercive (see Table 3.4), which m ight also be refram ed as aspects of treatm ent readiness. It is proposed then that the legal context in w hich offender rehabilitation is offered, and in particular the extent to w hich legal pressure is placed on the offender to participate in a program m e, will be an im portant determ inant of treatm ent readiness. O f course, the im pact of such external readiness factors can be understood in terms of their influence on internal factors, such as beliefs and attitudes about the appropriateness or value o f program m es. W ild, N ew tonTaylor and A lletto (1998) have suggested that the m ost appropriate w ay to understand coercion is in relation to the concept of m otivation. W hen offenders perceive that they are being coerced into treatm ent, they are more likely to see their participation as controlled by 43

Transitions to Better Lives

Table 3.4

Influences on perceptions of coercion

Factor

Description

A greem ent on the need for treatm ent

In order for an offender to accept coercion into treatm ent, he or she w ill also need to see his or her offending as likely to recur (that s / h e is susceptible), believe that offending is a serious problem , and that the costs of change do not outw eigh the benefits (such as not associating w ith friends w ho are likely to offend). H e or she w ill also need to have som e confidence that the treatm ent can be effective. It m ay be that an ind iv id u al's perceived level of coercion is linked with their personal treatm ent goals.

The aversiveness of the treatm ent

The m ore u n pleasant or distressing that treatm ent is thought to be, the m ore likely it is that pressure to attend will be perceived as coercive. For som e offenders the thought of bein g asked to disclose personal inform ation in a group setting is acutely distressing.

Inform ation about the treatm ent

L evels of perceived coercion may be low er w hen offenders are clear abou t w h at they are being coerced to do. Providing offenders with inform ation about the treatm ent and convincing them that rules w ould be enforced are am ong the m ost effective form s of coercion. Including offenders in the decision-m aking process m ay also reduce perceptions of coercion.

R elationship with the source o f the pressure

Views about the legitim acy of the courts to m ake decisions, and confidence in the legal process m ay also affect the extent to w hich legal pressure is perceived as coercive. Inform al pressure from fam ily and friends will also be perceived differently according to the im portance of these relationships to the offender.

Table 3.4 continues opposite

44

W h a t are readiness factors?

Table 3.4 continued Factor

Description

Personality factors

Individual differences in perceptions of coercion exist over and above that of referral source. Psychological reactance, for exam ple, can be understood as a m otivational state or as a personality trait, and has been defined as the degree to w hich an individual feels com pelled to regain lost or threatened freedom s. As coercion im plies a loss or threat to an ind ividu al's freedom o f choice to attend treatm ent, it m ay b e that those offenders w ho score highly on trait reactance w ould be m ore likely to perceive coercion. Reactance as a characterological factor is also thought to m ediate the effectiveness o f various therapeutic interventions, w ith highly reactant individuals generally having poorer treatm ent outcom es.

Source: A dapted from Day et al. (2004).

external contingencies (such as the requirement to fulfil obligations, to gain parole, or to look good). Deci and Ryan's (2000) model of motivation as occurring on a continuum from extrinsic to intrinsic motivation is useful here, as it suggests that extrinsic motivation can change to intrinsic motivation via a process whereby external motivators (socially sanctioned mores or requests) are internalised. For offenders, then, who feel coerced into program m es and who initially show little internal motivation to change their behaviour, it is possible that internal motivation will develop over the course of the programme as they come to personally endorse the values and self-regulations (Deci et al. 1999) identified in the treatment.

Conclusion In summary, there are a number of readiness factors that might best considered as external to the individual offender. Generally, these are features of the environment or context in which programmes are offered that will determine how the individual understands and evaluates the proposed intervention. These are practical issues such 45

Transitions to Better Lives

as the availability of program m es and the presence of qualified staff to deliver them, through to the pressure to participate, either legally, from those w ithin the correctional environm ent, or from significant others. The w ay in which an individual reacts to such pressure is likely to be an im portant determ inant of treatm ent readiness. H owever, it is characteristics of individual offenders that are m ost likely to influence the capacity or ability to engage with a particular rehabilitation program m e. This may relate to particular beliefs or attitudes about program m es in general, or the specific program m e under consideration, to affective or em otional factors that will influence m otivation to participate, and to broader issues related to the individual's goals and extent to which participation is regarded as in his or her best interests. M any o f the internal readiness factors described here are also discussed in other parts of the book. These are the aspects of readiness that can be readily assessed, and modified through interventions designed to im prove levels of readiness such that offenders are able to participate m eaningfully in a programm e.

46

Chapter 4

The Good Lives Model of offender rehabilitation and treatm ent readiness

The Good Lives M odel (GLM ) of offender rehabilitation is a strengthbased approach by virtue of its responsiveness to offenders' core aspirations and interests, and its aim of providing them with the internal and external resources to live rew arding and offence-free lives. It is closely aligned w ith positive psychology (Linley and Joseph 2004) because of its stress on prom oting offender w ell-being and its overall positive orientation to treatm ent, although it was developed independently of this perspective. We propose that rehabilitation theories are com posed of three levels of ideas: (1) a set of general assum ptions concerning the ethical values guiding rehabilitation, the nature of hum an beings, conception of risk, and the aim s and purpose of rehabilitation practice; (2) a set of general aetiological (causal) assum ptions that account for the onset and m aintenance of offending; and (3) the practice im plications of both of the above. In our view, it is helpful to think of the three levels as ordered in terms of their degree of abstractness, with the general aim s and values providing a conceptual foundation for the subsequent levels (aetiology and practice). Each level o f the GLM is discussed in greater detail below. A s a rehabilitation theory the GLM com prises a num ber of ethical, m etaphysical, epistem ological, m ethodological, aetiological, and treatment assum ptions that are intended to guide practitioners in their w ork w ith offenders (Ward and M aruna 2007). In this chapter we will outline the fundam ental assum ptions of the GLM . Our intention is to provide a reasonably detailed sum m ary of this recent rehabilitation theory and consider its relationship to the M O RM m odel of offender 47

Transitions to Better Lives

readiness outlined in detail earlier in the book. Because a primary em phasis of the GLM is on offenders' values and their associated good lives plans (GLP) it is m uch easier to m otivate them to engage in treatm ent (see also Chapter 11). The Good Lives Model w as form ulated as an alternative approach to correctional treatm ent that has the conceptual resources to integrate aspects of treatm ent not well addressed by the risk-need-responsivity (RNR) model (A ndrew s and Bonta 2006), such as the form ation of a therapeutic alliance, agency concerns, and m otivating individuals to com m it them selves to treatm ent and ongoing desistance from offending (Ward et al. 2007; Ward and M aruna 2007; Ward and Stewart 2003). The GLM has been m ost extensively applied to rehabilitation w ork with sex offenders and therefore the assessm ent process and interventions consistent with the GLM have been developed in the m ost detail w ith this particular population. It im portant to note, however, that the GLM is a general rehabilitation theory that is applicable to a wide range of problem s, including other types of crim inal behaviour, and is not restricted to use w ith sex offenders. It has recently been used effectively in w orking with individuals convicted of violent, non-sex-related crim es (Langlands et al. 2009; W hitehead et a l 2007) and also applied to individuals w ith m edical disabilities (Siegert et al. 2007).

Principles, aims and values of the GLM Embodiment, plasticity and cognitive extension

The first m ajor set of theoretical assum ptions of the GLM revolve around recent research and theory in cognitive science relating to the nature of hum an agency. M ore specifically, this research suggests that: (a) hum an agents' physical em bodim ent has a profound im pact on their cognitive functioning and interface w ith the world; (b) human agents are characterised by plasticity o f cognitive functioning; and (c) hum an agents have cognitive system s that incorporate both internal and external com ponents (Ward, in press). The above claim s converge on a picture of organism s who are (naturally) designed to act in pursuit of biological, psychological and social goals (Clark 2008). We briefly discuss each o f these assum ptions in turn. The claim that human beings are em bodied is based on a unified conception of the mind and body and a rejection of dualism . That is, m ental properties are thought to be causally dependent upon the

48

The Good Lives Model of offender rehabilitation and treatment readiness

body and their form determined in part by the experience of physical em bodim ent (Johnson 2007; Ward and N ee 2009). Furtherm ore, the body also plays an im portant part in altering the environm ent in w ays that facilitate problem clarification and effective action. It is the interface betw een inner and outer resources that m akes it possible for individuals to bring about goal-directed changes in the environm ent and ultim ately w ithin them selves. The dependence of goal-directed action and psychological functioning upon the body creates a source of vulnerability for hum an agents and underlines the need to ensure that threats to physical integrity are effectively m anaged. The provision of adequate food and water, safe and hygienic environm ents, freedom from physical danger, and accom m odation are necessary ingredients of a good life. Typically, this m eans that individuals need educational and vocational skills to be able to w ork in order to pay for these essential m aterials. The fact of being physically vulnerable agents points to our ultimate interdependence and reliance on each other for access to vital goods or at least to the m eans o f providing them for ourselves. Offenders as em bodied hum an agents require the m aterials needed to protect their physical integrity and subsequent ability to act in pursuit of their goals. The second assum ption concerning the nature of hum an beings and their capacity for agency trades on the view that they are cognitively versatile anim als w ho are able to quickly adapt to novel situations and acquire new cognitive repertoires and tools with relative ease (Clark 2008). H um an beings' sense of self is derived from the ability to effectively change the world and them selves in accordance w ith their personal com m itm ents (Clark 2008; Korsgaard 2009). From a rehabilitation standpoint, the 'soft' nature of hum an agency rem inds correctional practitioners that enhancing offenders' abilities to achieve better life plans is likely to alter their sense of them selves in ways that are socially beneficial as w ell as personally fulfilling (Ward, in press). The third agency-related assum ption builds on the fact of human beings' cognitive plasticity and claim s that external cognitive resources such as language, com puters, other m inds, and social and cultural institutions under som e circum stances can be viewed as part of people's (extended) m inds. In other w ords, we are not cognitively limited by the biological boundaries of skin and skull and are able to intentionally incorporate internal and external elem ents w hen engaged in cognitive tasks. We do not have the space to fully explain this com plex and novel idea but point out that it is logically connected 49

Transitions to Better Lives

to the previous two assum ptions (see Ward, in press). It is because hum an beings are physically em bodied that they are able to use tools of various kinds to change them selves and their world. Furtherm ore, it is their cognitive plasticity and soft agency that enables people to actively incorporate internal and external cognitive resources w hen engaged in problem -solving activates. The im plications of this assum ption for offender rehabilitation is that it m akes sense to focus our efforts on what m atters to people and to realize that external social and cognitive resources m ay well be actively recruited in offenders' problem -solving routines and strategies. If offenders are quarantined in environm ents that contain others like them and few pro-social m odels, the chances are that their beliefs, values and actions will continue to be anti-social in nature. Primary human goods

The above set of three presuppositions of the GLM centred on human em bodim ent and agency are the m ost fundam ental ones and the follow ing assum ptions are really derived from them. The biological nature of hum an beings and the supervening of psychological properties on physical processes and structures m eans that in order for individuals to function effectively their basic needs have to be met (Deci and Ryan 2000). Furtherm ore, the biological and psychological evidence suggests that all people, including offenders, are naturally inclined to seek certain goals, or w hat we have called primary human goods (such as relatedness, creativity, physical health, and m astery; see Ward and M aruna 2007; Ward and Stew art 2003). In essence, prim ary goods are states of affairs, states of mind, personal characteristics, activities or experiences that are sought for their ow n sake and are likely to increase psychological w ell-being if achieved (Kekes 1989; Ward and Stew art 2003). In addition to these prim ary goods, instrumental or secondary goods provide particular w ays (that is the m eans) of achieving prim ary goods: for exam ple, certain types of w ork or relationships. For instance, it is possible to secure the prim ary good of relatedness by the w ay of rom antic, parental or personal relationships. The notion of instrum ental goods or m eans is particularly im portant w hen it com es to applying the GLM to offending behaviour as it is assum ed that a prim ary reason w hy individuals com m it offences is that they are seeking prim ary goods in socially and often personally destructive ways. The psychological, social, biological and anthropological research evidence provides support for the existence of at least ten groups of

50

The Good Lives Model of offender rehabilitation and treatment readiness

prim ary hum an goods (see A spinw all and Staudinger 2003; Deci and Ryan 2000; Em m ons 1999, 2003; Linley and Joseph 2004; N ussbaum 2006; Ward and M aruna 2007). These include the following: 1 Life: The prim ary good of life incorporates physical needs and factors that are im portant for healthy living and physical functioning, such as food, water, a physically healthy body, and so on. 2 Knowledge: This prim ary good is based on the notion that human beings are inherently curious and possess the desire to understand aspects of them selves, their natural environm ents, and other people. 3 Excellence in play and work: This prim ary good refers to the desire to engage in leisure or fun activities for their ow n sake and to strive for m astery at work-related and leisure or recreational activities. 4 Autonomy: The prim ary good of autonom y refers to the desire to form ulate one's ow n goals and to seek w ays to realise these through actions and activities of one's choice w ithout facing undue interference from others (m oderated by cultural and social norm s). 5 Inner peace: The prim ary good of inner peace refers to em otional self-regulation and the ability to achieve a state of dynam ic em otional equilibrium and com petence. 6 Relatedness: The good of relatedness refers to the natural desire of hum an beings to establish w arm , affectionate bonds with other people. It is noted that these relationships range from intim ate, rom antic relationships to close fam ily relationships to platonic relationships and friendships. 7 Community: The prim ary good o f com m unity refers to the desire hum an beings have to belong to social groups and to feel connected to groups that reflect their interests, concerns and values. 8 Spirituality: The prim ary good of spirituality refers to the desire to discover and attain a sense of m eaning and purpose in life. 9 Happiness: The prim ary good of happiness refers to a hedonic (pleasure) state or the overall experience of being content and 51

Transitions to Better Lives

satisfied with on e's life, and includes the sub-good of sexual pleasure. 10 Creativity. The prim ary good of creativity refers to the desire for novelty and innovation in on e's life, the experience of doing things differently, or engaging in a specific activity that results in an artistic output or other novel or creative product. A n especially significant characteristic o f the GLM is that the goods are plural rather than singular and, therefore, a fulfilling life will m ost probably require access to all the prim ary goods even though individuals can legitim ately vary in the w ay they value or rank them. This m eans that there are m ultiple sources of m otivation and that each has their origin in the evolved nature of hum an beings. Values and practical identities

The plural nature of the goods sought is likely to result in their differential w eightings or endorsem ent by individuals. W hile all the prim ary goods need to be present to some degree (that is, m eet a threshold requirem ent), if persons are to achieve good lives there could be significant differences in the experiences, objects and activities they consider m ost im portant. A ccording to Korsgaard (1996), conceptions of practical identity provide 'a description under which you value yourself and find your life w orth living and your actions to be w orth undertaking' (1996: 101). Thus individuals' sense of identity em erges from their basic value com m itm ents: the goods they pursue in search of better lives. Interestingly, Korsgaard argues that when there are conflicts betw een different practical identities people have to w ork hard to establish som e degree of unity in their lives, and she suggests that a way of assisting this process is by focusing on our com m on hum anity and our (shared) inherent dignity. The existence of a num ber of practical identities also m eans that each of us will draw from a variety of d istinct value sources w hen faced w ith decisions about how best to act (Korsgaard 2009). For exam ple, a person may value being a father, psychologist, scientist, citizen and m em ber of a political party, and each of these practical identities will exert som e norm ative pressure on his actions and life. A t tim es the aim s and subsequent actions arising from the value com m itm ents of each of these practical identities could even conflict. The relevance of variation in value endorsem ents is that if offenders' sense of them selves and w hat really m atters depends upon the things they m ost value, then correctional practitioners ought to identify what 52

The Good Lives Model of offender rehabilitation and treatment readiness

prim ary goods are m ost heavily endorsed and in particular how they are expressed in their lives (A rcher 2000; Clark 2007; Em m ons 2003). Because hum an beings are thinking anim als there is a reflective gap betw een the experience of a desire to act in pursuit of a natural good or incentive, and actually doing so (Korsgaard 2009). This reflective gap allow s individuals space to critically evaluate desires and to decide w hether or not they are w orthy of fulfilm ent; w hether they are really of value. Arguably, problem atic actions such as sexual offending partly arise from individuals m aking faulty judgem ents and reveal a lack of forethought or know ledge concerning the relevant facts and the real value of the proposed actions. Thus, the process of rehabilitation requires not ju st the targeting of isolated 'factors' but also the holistic reconstruction of the 'self'. Goods and risks

According to the GLM , correctional interventions should aim to (a) promote offenders' aspirations and plans for better lives, as well as (b) m an age/red uce their risk to the community. This assum ption has both norm ative and pragm atic strands to it. N ormatively, the assertion that interventions should prom ote w ell-being alongside reduce risk reflects the ethical foundation of the GLM in human rights theory and practices (Laws and Ward, in press). Pragm atically, it is assumed that because crim inogenic needs and hum an needs are causally related (see below), the prom otion o f adaptive approach goals should also reduce dynam ic risk factors. Thus a m ajor aim of correctional reintegration w ork is to help individuals to construct a life plan that has the basic prim ary goods, and w ays of effectively securing them, built into it and does not involve inflicting harm on others. Ecological selves

As discussed above, according to the GLM people are multifaceted beings com posed of a variety of interconnected biological, social, cultural and psychological system s, and are interdependent to a significant degree. W hat this entails is that com plex anim als such as hum an beings can only flourish w ithin a com m unity that provides em otional support, m aterial resources, education, and even the m eans of survival. The com plexity of hum an functioning m eans that an adequate explanation of som ething as im portant as crim e will require m ultiple levels of analysis and theoretical perspectives. In particular, the interdependency of hum an behaviour points to the necessity of adopting an ecological fram ew ork. 53

Transitions to Better Lives

The fact that hum an beings are interdependent and that, therefore, a satisfactory understanding of behaviour will alw ays involve an appreciation of the contexts in w hich they exist, has im portant im plications for therapists w hen designing reintegration programm es. Thus, according to the G LM , any assessm ent and intervention should take into account the m atch betw een the characteristics of the individual and the likely environment where he or she will be functioning. Rather than view ing the offender as essentially a selfcontained deviancy m achine (or bearer o f risk - see below ) who therefore requires treatment designed to restore or repair or, more frequently, to m anage a faulty system , the aim is to locate him or her w ithin a social network. Treatment consistent with the GLM is viewed as furnishing individuals w ith som e of the agency scaffolding and resources required to establish im portant social bonds and to engage m eaningfully with the world. The nature o f risk

Because people are conceptualised to be constituted from , and to be em bedded w ithin, com plex system s, risk is viewed as m ultifaceted rather than purely individualistic (D enny 2005). In our view, risk is best viewed in contextual term s rather than conceptualised purely as constituted by individual deviancy. Thus it is to be expected that an adequate risk m anagem ent plan would need to take into account individuals' particular lifestyles and environm ents. Even those dynam ic risk factors that can be said to be located 'insid e' individuals (impulsivity, aggressiveness) are only m eaningful in their specific, cultural and situational contexts. The trouble with psychom etric approaches to risk assessm ent and m anagem ent is that they have a tendency to identify risk prim arily in terms of individuals' deviancy and to view offenders as essentially bearers of risk (Ward and M aruna 2007; Ward and Stew art 2003). By 'bearers of risk' we mean that in som e sense risk is seen as inhering w ithin individual offenders, and to a lesser extent their environm ents. A difficulty w ith such a static conceptualisation is that it fails to appreciate how risk can be created by correctional interventions and policies that effectively isolate offenders, such as com m unity notification or geographical restrictions (Vess 2009). The nature o f intervention

Finally, according to the G LM , a treatment plan should be explicitly constructed in the form o f a good lives conceptualisation or plan. 54

The Good Lives Model of offender rehabilitation and treatment readiness

In other w ords, it should take into account individuals' strengths, prim ary goods and relevant environm ents, and specify exactly w hat com petencies and resources are required to achieve these goods. An im portant aspect of this process is respecting the individual's capacity to m ake certain decisions them selves, and in this sense accepting their status as an autonom ous individual. This is in direct contrast to previous recom m ended practice in the treatm ent of offending behaviours, where therapists were cautioned not to allow offenders to participate in decision-m aking (see Salter 1988). U sing the GLM , we believe that each individual's preference for certain prim ary goods should be noted and translated into his or her daily routine (for exam ple the kind of w orks, education and further training, and types of relationships identified and selected to achieve prim ary goods).

Aetiological assumptions of the GLM As stated earlier, the aetiological com ponent of a rehabilitation theory flow s logically from a theory's basic assum ptions, is general in nature, and functions to give correctional w orkers a cognitive m ap or general overview of the broad causes o f anti-social behaviour. A ccording to the GLM , goals are usefully construed as prim ary hum an goods translated into m ore concrete form s, and as such are typically the objects of intentions and actions. Goals are the ultimate and interm ediate ends of any actions and collectively give shape to people's lives insofar as they create a structure of daily activities that represent w hat is of fundam ental im portance to them. In terms of practical identities, goals are typically them atically linked to concrete identities and the various roles and tasks they imply. For exam ple, as a psychologist a person has responsibility for the assessm ent and treatm ent of psychological disorders. Each of these dom ains of professional practice is linked to actions, guided by particular goals, such as conducting an interview com petently, interpreting psychological tests, or assisting an individual to overcom e his or her fears of intimacy. Alternatively, the practical identity of being som eone's rom antic partner generates a variety of tasks such as providing em otional support, spending tim e together, and m aintaining a household. In other w ords, goals are typically clustered together under specific descriptions; these descriptions are ultim ately anchored in practical identities (Em m ons 1999; Korsgaard 2009). A ccording to the GLM , there may be a num ber of distinct problem s w ithin the various dom ains o f hum an functioning that 55

Transitions to Better Lives

can result in offending behaviour: em otional regulation difficulties, social difficulties, offence supportive beliefs, em pathy problem s, and problem -solving deficits. Yet, such individuals' general underlying personal m otiv ation s/go als are rarely inherently bad. Instead, it is the m eans used to achieve these goods that are deviant. The value of this understanding is that it helps to focus clinical attention on prim ary goods, the ultim ate underlying m otivating factors, and aw ay from an exclusive focus on the psychosocial difficulties with w hich individual clients are struggling. That is, there are likely to be distortions in the internal and external conditions required to achieve the prim ary goods in socially acceptable and personally satisfying ways. The GLM guided analysis goes beyond deficit-based etiological theories (theories that focus on w hat individuals lack) by encouraging clinicians to think clearly about ju st what it is that the person is seeking w hen com m itting the offence. This inform ation has direct treatm ent im plications and can provide a pow erful way of m otivating individuals to engage in therapy; the aim is to help them to secure hum an goods that are im portant to them, but to do so in w ays that are socially acceptable and also more personally satisfying. The latter point is especially im portant, as m ost of the causal factors involve self-defeating attem pts to seek personally valued goals and consequences. The GLM can explain w hy this is so and provide clinicians w ith a clear understanding o f where the problem s reside in an individual's life plan. From the perspective of the GLM there are two routes to the onset of offending, each reflecting ind ividu als' agency: direct and indirect (Ward and G annon 2006; Ward and M aruna 2007). The direct pathway is im plicated w hen offending is a prim ary focus of the (typically im plicit) cluster of goals and strategies associated with an individual's life plan. This m eans that the individual intentionally seeks certain types of goods directly through crim inal activity. For exam ple, an individual m ay lack the relevant com petencies and understanding to obtain the good of intim acy with an adult, and furtherm ore may live in an environm ent where there are few realistic opportunities for establishing such relationships. Thus, the actions constituting offending are a m eans to the achievem ent of a fundam ental good. The indirect route to offending occurs w hen the pursuit of a good or set of goods creates a ripple effect in the person's personal circum stances and these unanticipated effects increase the pressure to offend. For exam ple, conflict betw een the goods of relatedness and autonom y m ight cause the break-up of a valued relationship and subsequent feelings of loneliness and distress. The use of alcohol 56

The Good Lives Model of offender rehabilitation and treatment readiness

to alleviate the em otional turm oil could lead to loss of control in specific circum stances and this m ight increase the risk of offending. These indirect or ripple effects are particularly evident when two practical identities a person is invested in conflict and cause him or her uncertainty about how best to act. An exam ple of this conflict of identities is w hen an offender values both his roles as a w orker and as a husband. The two identities can on occasions clash and in som e circum stances the pressure to w ork longer hours in order to get a job done m ight interfere w ith his responsibilities as a partner. From the standpoint of the GLM , crim inogenic needs are con­ ceptualised as internal or external obstacles that frustrate and block the acquisition of prim ary hum an goods. We suggest that there are four m ajor types of difficulties often evident in individuals' life plans. In our view these types of problem s are overlapping but conceptually distinct. It is also im portant to note that the real problem resides in the secondary goods rather than the prim ary ones. In other w ords, it is the activities or strategies used to obtain certain prim ary goods that create problem s, not the prim ary goods them selves (that is prim ary goods are sought by all humans). First, an individual w ho has problem s w ith the means he or she uses to secure goods m ay be using inappropriate strategies to achieve the necessary prim ary goods needed for a good life. Second, an individual's life plan m ight also suffer from a lack of scope w ith a num ber of im portant goods left out of his or her plan for living. Third, some people may also have conflict (and a lack of coherence) am ong the goods being sought and their associated practical identities and therefore experience acute psychological stress and unhappiness (Em m ons 1999). Fourth, a final problem is when a person lacks the capabilities (know ledge, or skills) to form or effectively im plem ent a life plan in the environm ent in w hich he or she lives, or to adjust his or her goals to changing circum stances (for exam ple im pulsive decision-m aking). The problem of capability deficits has both internal and external dim ensions. The internal dim ension refers to factors such as skill deficits w hile external dim ension points to a lack of environm ental opportunities, resources and supports. In summary, the aetiological com m itm ents of the GLM are general in form and stem from a view o f hum an beings as creatures capable of reflective agency, usually acting under the conceptual constraints of a range of practical identities. That is, w e propose that hum an beings are goal-seeking, culturally em bedded anim als w ho utilise a range of strategies to secure im portant goods from their environm ents w hen occupying personally valued social or cultural roles (partners, workers, 57

Transitions to Better Lives

citizens, playm ates, artists, helpers and so on). W hen the internal or external conditions necessary to achieve valued outcom es associated w ith practical identities are incom plete or absent, individuals tend to becom e frustrated and may engage in anti-social behaviour. The etiological com m itm ents serve to orient correctional w orkers and require supplem entation from specific theories to supply more fine­ grained explanations of anti-social behaviour and particular types of offences.

Implications of the GLM for practice A GLM -oriented treatment program m e seeks to tailor an intervention plan around an offend er's core values and associated practical identities. The good lives plan unfolds from this value centre and incorporates all of the various goods required to function as a reflective and effective agent w ithin specific environm ents. W here possible, local com m unities and resources are recruited and the objective is to assist in the building of a better life rather than sim ply trying to contain risk. For exam ple, an individual's treatm ent plan could be based on his or her desire to learn a trade (becom e a m echanic, for exam ple) and establish a rom antic relationship. The skills required to becom e a m echanic, such as m echanical know ledge of engines, effective w ork habits, at least a reasonable degree of social and com m unication skills, affective and self-control com petencies, m ay reduce risk w hile consolidating the offender w ithin a social netw ork. A ccess to w ork­ m ates and hobbies that cohere w ith his or her interests m ight further open up opportunities to m eet potential partners who are law -abiding and supportive. The result of such a plan will hopefully be a life that is fulfilling, m eaningful, ethically acceptable and socially productive (Burnett 2002; M aruna 2001). We will now briefly describe each o f the five phases of a GLM rehabilitation fram ew ork (for more detail on GLM -oriented treatment see Laws and Ward, in press; Ward and M aruna 2007; Ward and Stew art 2003). The first phase when intervening w ith offenders from the standpoint of the GLM involves the detection of the social, psychological and m aterial phenom ena im plicated in individuals' offending. This requires a careful analysis of offenders' level of risk, their living circum stances, physical and social problem s and psychological capabilities around the tim e o f their offending and stretching into their past as well. O ffenders are likely to have m ultiple problems, 58

The Good Lives Model of offender rehabilitation and treatment readiness

such as poverty, substance abuse, lack of accom m odation, high levels of im pulsiveness and aggressive behaviour, and so on. In the second phase of the GLM the function of offending (what the individual expected to achieve iria his offending) is established through the identification of prim ary goods that are directly or indirectly linked to the crim inal actions. In addition, the identification of the overarching good or value around which the other goods are oriented should also be ascertained. This step requires that practitioners identify the practical identities endorsed by offenders and clarify how they are causally related to their offending actions. It is anticipated that the core goods (for exam ple, m astery or caring) will be translated into more concrete values and tasks that directly connect w ith offenders' general life circum stances and their offencerelated actions. In the third phase of the GLM rehabilitation process, the selection of the practical identities and their overarching good(s) or value(s) is undertaken and m ade a focus of a plan. As discussed earlier, frequently practical identities are aligned w ith the prim ary goods and in a sense sim ply flesh out the abstractness of the good in question. In effect, practical identities and their goals, strategies and practices provide the detail needed to effectively w ork w ith an offender. For exam ple, an individual m ight nom inate know ledge and relatedness as the two m ost im portant goods and decide that going to university and establishing a relationship w ith a w om an are m eans to these ends. In the fourth phase, a greater level of detail is added to the above developing plan and the selection of secondary goods or values that specify how the prim ary goods will be translated into ways of living and functioning, is undertaken. In this step identification of the contexts or environm ents in w hich the person is likely to be living while in the com m unity during or follow ing treatm ent is conducted. For exam ple, the practical identity o f being a university student (and partner in a relationship) is now exam ined w ith respect to a possible environm ent and the educational, social, psychological and material resources required to m ake this possible are noted. The GLM is a regulatory and pragm atic model so it is im perative that the probable environm ents a person will be living in are identified and their potential to provide the required resources to realise the good lives plan ascertained. In the fifth phase, the practitioner constructs a detailed inter­ vention plan for the offender based on the above considerations and inform ation. The plan will be holistic, specify the internal and external 59

Transitions to Better Lives

conditions required to successfully im plem ent it, revolving around offenders' core values and their associated practical identities, and the various tasks for correctional practitioners will be carefully detailed. Dynam ic risk factors or crim inogenic needs are indirectly targeted when cognitive behavioural techniques and social interventions are utilised in the acquisition of offender com petencies. Thus, taking into account the kind of life that would be fulfilling and m eaningful to the individual (prim ary goods, secondary goods, and their relationship to w ays of living and possible environm ents), the evaluator notes the capabilities or com petencies the individual requires in order to have a reasonable chance of applying the plan. Practical steps are then taken to organise the various actors involved and to put the good lives plan into action. The offender is consulted in all the various phases and in a robust sense he or she drives the content of the plan, if not its form. Furtherm ore, the practitioner seeks to balance the ethical entitlem ents of the offender w ith those of victim s and m em bers of the community.

The GLM and the MORM In this section we briefly discuss the relationship betw een the GLM and the M ORM . The GLM is a broad rehabilitation fram ew ork that has the potential to organise and guide all aspects of offender rehabilitation, while the M O RM is a specific theory of treatm ent readiness. The focus of the M O RM is on the factors associated w ith engaging an offender in the treatm ent process and in this role it can be incorporated w ithin the GLM fram ew ork as a useful conceptual model of intervention preparation. The assum ptions of the GLM concerning the nature of hum an beings and the im plications of these facts for the onset and treatm ent of offending related problem s arguably underpin the M ORM . That is, the basic assum ptions of this readiness model about the required person and contextual factors for effective engagem ent in treatm ent point to problem s in the scope, capacity, m eans and conflict w ithin offenders' good lives plans (GLP). To be ready for treatm ent m eans accepting that it can provide a w ay of achieving im portant goals, valued outcom es and their reflection in practical identities that are salient for an offender (see Chapters 1, 2 and 3). The critical task for practitioners is to provide a bridge betw een w hat an offender values and needs, his current circum stances (im prisonm ent, or serving a com m unity sentence), and a possible future life. Providing that bridge is the task of readiness 60

The Good Lives Model of offender rehabilitation and treatment readiness

interventions: creating a sense of efficacy, linking past and future lives to valued goals, highlighting the way treatm ent program m es can be helpful in securing these personal goals, creating a context that scaffolds and supports treatm ent, and m ost importantly, treating the offender with respect and listening carefully to his concerns and hopes. W hat we are saying is that the GLM contains the necessary general conceptual resources to help practitioners locate the relevant readiness factors in specific individuals. It can do this because of its view that people are inherently value-seeking and, as such, actively pursue and seek to im plem ent im portant goals b y w ay of their associated GLPs. Individuals' ability to construct and im plem ent adaptive and fulfilling GLPs depends on their possession of the necessary capabilities and an environm ent that is receptive to their efforts. Readiness interventions depend upon the identification of these (often) im plicit GLPs and their associated practical identities, and represent attem pts to directly connect them to the process of treatment. Readiness interventions can be viewed as conduits betw een past and future lives by way of em phasising the things that really m atter to offenders and show ing how participation in correctional program m es can help them to achieve the things they value (see Chapter 11).

Conclusion In this chapter we have outlined the basic assum ptions, aetiological com m itm ents and practice im plications of the GLM . O ur aim has been to stress the focus of the GLM on the possibility of better lives for offenders and therefore underline the im portance of agency considerations rather than sim ply reduction of risk factors. In our view, the GLM has the theoretical resources to provide practitioners w ith a conceptual m ap to guide all aspects of their clinical w ork with offenders and also help steer the professional activities of other correctional w orkers and com m unity volunteers. There is natural resonance betw een the GLM and readiness concepts because of their com m on assum ptions about the im portance of intervening with offenders holistically w ithin their social ecology. Furtherm ore, the fact that the GLM is based upon identifying offenders' core values and establishing ways of realising them m akes it easier to motivate them to engage in treatm ent program m es. It is easier to persuade individuals to w ork tow ards the creation of fulfilling lives rather than sim ply trying to be less harm ful to others.

61

Chapter 5

The assessment o f treatm ent readiness

Treatment readiness is now w idely acknow ledged as playing an im portant role in the process of offender rehabilitation. U nderstanding the extent to w hich offenders are treatm ent ready can help to im prove program m e selection processes, reduce program m e attrition, use program m e resources m ore efficiently, and assist in the developm ent of interventions for those who are considered to not be ready for treatm ent (Burrow es and N eeds 2009; Casey et al. 2005; H ow ells and Day 2003). M ost importantly, given that the prim ary goal of rehabilitation program m es is reduced recidivism , the better the fit betw een the needs of the individual and the program m e(s) to w hich an offender m ight be referred, the more likely it is that risk of recidivism can be effectively m anaged (Ward et al. 2004b). Yet despite first being articulated in an offender context by Serin and K ennedy (1997; Serin 1998) more than a decade ago, there have been few attem pts to clearly operationalise the readiness construct and develop appropriate m easurem ent tools. A ccording to Ward et al. (2004b), the m ajor obstacle to such developm ent has been a failure to distinguish betw een the three distinct but related concepts o f treatm ent m otivation, responsivity, and readiness (see Chapter 1 for a m ore detailed discussion of these constructs). A s a consequence, researchers have worked w ithin each of these fram ew orks to develop instrum ents that in their view assess w hether an individual is 'treatm ent ready'. For exam ple, M cM urran and colleagues (M cM urran et al. 2006, 2008; Sellen et al. 2006, 2009) have utilised a m otivational fram ew ork in adapting the Personal Concerns Inventory (Cox and K linger 2004a) for use with an offender 62

The assessment of treatment readiness

population. Serin (Serin 1998; Serin and K ennedy 1997; Serin et al. 2007) has also adopted a m otivational approach, but advocates that in addition to a generic model of readiness, there is a need to develop instrum ents for unique offender groups (for exam ple, sex offenders for whom denial rates are high even after conviction). Perhaps the m ost frequently used m otivational m easures, however, are those derived from the transtheoretical model - the Readiness to Change Q uestionnaire (RCQ) (Rollnick et al. 1992); University of Rhode Island Change A ssessm ent (URICA) (M cC onnaughy et al. 1983, 1989); Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES) (M iller and Tonigan 1996); Violence Risk Scale (VRS) (Wong and Gordon 1999-2003); Violence Risk Scale-Sexual O ffender (W RS-SO) (Wong et al. 2003) - although there is considerable debate about the suitability of such m easures for use w ith offender populations (see Casey et al. 2005; Littell and Girvin 2002). More recently, Casey and colleagues (2007) drew on the m ultifactor offender readiness model (M ORM ; Ward et al. 2004b) to develop a general screening measure that assesses the internal characteristics of treatm ent readiness as described in the m odel. This m easure, as well as an adaptation for use w ith offenders entering violence program m es (see Day et al. 2009a), has been found to successfully predict offender engagem ent in group rehabilitation program m es. The dearth of theoretical developm ent that surrounds the readiness construct and concom itant lack of clarity regarding definition and m easurem ent poses a dilem m a for practitioners. On the one hand, there is an expectation of evidence-based practice w hich necessitates a clear understanding of 'w hat w orks when, w here, and for w hom ' (M cGuire 2004: 339), while on the other, little is available in terms of psychom etrically sound assessm ent tools to m easure treatment readiness. A ssessm ent is a critical elem ent in identifying an individual's strengths and w eaknesses. In fact, H unsley and M ash (2008) have argued that assessm ent is the foundation that underpins evidencebased treatm ent, claim ing that 'as the identification of evidence-based treatm ents rests entirely on the data provided by assessm ent tools, ignoring the quality of these tools places the w hole evidence-based enterprise in jeop ard y' (2008: 3). Thus, like treatm ent, assessm ents should also be evidence-based, w ith the selection of assessm ent tools guided by scientific evidence for the clinical utility of the selected instrum ents. This can be a difficult task. Both Kazdin (2005) and Sechrest (2005) have pointed out that it is not possible to identify a finite set of studies that establishes the psychom etric properties of an instrum ent. This chapter considers published m easures that purport 63

Transitions to Better Lives

to assess treatm ent readiness w ith a particular focus on the extent to which these tools m eet the criteria for evidence-based assessm ent. W hat follow s is a brief overview of evidence-based assessm ent and the 'good enough' principle developed by H unsley and M ash (2005, 2007, 2008; M ash and H unsley 2005) for clinical assessm ent. This principle is the basis upon w hich the authors have developed a rating criteria that can be used to assess the psychom etric properties of any particular assessm ent instrum ent. W hile their fram ew ork has, to date, been applied predom inantly to assessm ents used in the identification of psychological disorders and the m easurem ent of psychosocial deficits that m ay accom pany such disorders, it is equally valid in a forensic context where there is a heavy reliance on assessm ent in terms of identifying client needs, focusing interventions and, in an environm ent where there is considerable political and social pressure for change, evaluating treatm ent outcom es. This is followed by a review and evaluation of instrum ents identified from the literature as being suitable for assessing an individual's readiness to enter into offender rehabilitation program m es, w ith a particular focus on use for the purposes of (a) case conceptualisation and treatm ent planning and (b) treatm ent m onitoring and evaluation.

Evidence-based assessment (EBA) The ability to conduct assessm ents has been identified as a core com petency of professional psychologists (H unsley and M ash 2008) and becom e a unique and defining feature of their professional expertise (K rishnam urthy et al. 2004). Yet despite this critical im portance, m uch less attention has been paid to evidence-based assessm ents than to statem ents about evidence-based practice and best practice guidelines. W here attention has been paid, the focus has prim arily been on ensuring that practitioners use measures with established reliability and validity (see C ham bless and H ollon 1998; Kazdin et al. 1986), although these criteria are limited in term s of establishing the full nom othetic or idiographic utility of a m easure (M cFall 2005). M ore recently, greater attention has been paid to the developm ent of evidence-based assessm ent (EBA) protocols and using specific criteria to evaluate som e of the tools used m ore frequently in both research and practice (see, inter alia, special issues of Psychological Assessment, 2005, 17(2) and Journal of Clinical Child and Adolescent Psychology, 2007, 36(1) for more detailed discussions of specific assessm ent types and measures). One of the m ore w ell-developed rating criteria is that 64

The assessment of treatment readiness

of H unsley and M ash (2005, 2007, 2008), w hich is outlined below and is used here to evaluate treatm ent readiness assessm ents. It is first useful, however, to consider w hat is m eant by evidence-based assessm ent. An evidence-based approach to the provision of professional services in m ost health and hum an service system s (health, m ental health, social w ork, crim inal justice) is now a fundam ental principle of best practice (Barlow 2004). W hat this m eans from a practical standpoint is that providers need to draw on inform ation from a broad range of sources (such as research data, clinical experience, client preferences) w hen m aking decisions about service options (H unsley and M ash 2005). Given that tim e constraints and resource lim itations are frequently key factors in this decision m aking-process, the role of assessm ent becom es critical. Practitioners rely heavily on the accuracy of m easurem ent tools for diagnostic and case conceptualisation purposes, as w ell as decisions about the course and efficacy of treatment. In fact, Weisz, Chu and Polo (2004) have argued that evidence-based practice is an assessm ent-intervention dialectic w hich involves (a) accurate identification of initial treatment targets, (b) selection of the m ost appropriate evidence-based treatm ent targets, and (c) periodic assessm ent of treatm ent to ascertain w hether treatment adjustm ents are required (that is, assess-treat-reassessadjust treatment). Although the term 'evidence-based assessm ent' has been used in the scientific literature in a num ber of different ways, the definition used in this chapter is one that includes the standard psychometric indices of reliability and validity but goes beyond these criteria to include a num ber of utility considerations (see Cohen and Parkm an 1998). U tility encom passes such things as treatment utility (the extent to which clinical assessm ent data contributes to positive treatment outcom es; N elson-Gray 2003) and diagnostic utility (the extent to which assessm ent data contribute to the form ulation of an accurate and com plete diagnosis; H unsley 2003). It also includes such factors as (a) assessm ent-related costs, including im provem ents in clinical decision-m aking resulting from the assessm ent; (b) any changes in false positive and false negative rates directly associated with the assessm ent (based on sensitivity and specificity indices); and (c) the econom ic and psychological costs associated w ith such errors (H unsley and M ash 2005, 2007). In addition to the broad range of population-specific and problem-specific psychom etric qualities to be considered for each assessm ent, establishing an EBA fram ew ork is further challenged by other factors, including the vast num ber of assessm ent m easures and 65

Transitions to Better Lives

procedures available relative to the num ber of available treatment options; the m any purposes of assessm ent as com pared to treatment; and the iterative nature o f the decision-m aking process (H unsley et al. 2004). U sing these considerations as a guide, H unsley and M ash (2005, 2007; M ash and H unsley 2005) have identified three critical aspects that define EBA. First, both the selection of constructs to be assessed and the assessm ent process should be guided by scientifically supported theories and em pirical evidence that establish im portant facets of a particular disorder or problem , identify the key sym ptom s or elem ents to be assessed, and w here necessary identify com m on co-m orbid conditions. A ssessm ents should, therefore, be disorder or problem -specific and include em otional and relational problem s that may also be experienced by the individual (such as loneliness or anger). Given the focus of assessm ent is to identify the precise nature of an individual's problem (s), it may also be necessary to adopt a m ulti-stage or iterative approach w hereby the assessm ent process shifts from general or non-specific for initial assessm ents and then becom es more problem specific as the assessm ent focus is refined. That is, EBA s need to be em bedded in the purpose of assessm ent (that is, screening, diagnosis, prognosis, treatm ent planning, treatment m onitoring, and treatm ent evaluation). Second, practitioners should opt for instrum ents that are psychom etrically strong. In addition to evidence of reliability, validity and clinical utility, m easures should also have appropriate norm s for norm -referenced interpretation a n d /o r replicated supporting evidence regarding the accuracy (sensitivity, specificity, predictive pow er and so on) of cut-off scores used for criterion-referenced interpretation (H unsley and M ash 2005, 2007; M ash and H unsley 2005). This also extends to individual characteristics, w ith a need for EBA s to be sensitive to an individual's age, gender, race and ethnicity, as w ell as specific cultural factors. Both the psychom etric properties and individual characteristics should be evident irrespective of the purpose for which an assessm ent tool is used (screening, diagnosis, prognosis, case conceptualization, treatm ent form ulation, treatm ent m onitoring, treatm ent evaluation). That said, not all psychom etric properties apply to all assessm ent purposes. W hile group validity statistics (for exam ple, sensitivity, specificity, positive and negative predictive pow er) are relevant to diagnosis and prognosis (see H su 2002), these statistics are m uch less relevant w hen assessm ent is used for treatm ent m onitoring or evaluation purposes. Finally, assessm ent is inherently a decision-m aking task, irrespective of the psychom etric 66

The assessment of treatment readiness

strength of any assessm ent tool (H unsley and M ash 2005, 2007; M ash and H unsley 2005). It is the practitioner who iteratively form ulates and tests hypotheses, a process frequently undertaken w ith in­ com plete or inconsistent data. As w ith evidence-based treatment, an evidence-based approach to assessm ent should, therefore, evaluate the accuracy and utility of the decision-m aking process. This would include a review of (potential) data synthesis errors or errors in test interpretation, the financial costs associated with the total assessm ent process, and the im pact of assessm ent on clinical outcom es (see Doss 2005 for a guide to conducting such an evaluation). In order to provide an operational definition of these criteria, H unsley and M ash (2005, 2007, 2008; M ash and H unsley 2005) have taken into account a range of issues that im pact upon the decisions m ade by practitioners in their choice of m easures and how best to integrate assessm ent outcom es into service provision as a whole. W hile psychom etric strength is, of course, an im portant factor in the choice of m easure, system ic considerations - m ost notably time constraints and resource lim itations - highlight the need for assessm ents that are brief, clear, clinically feasible and user-friendly (H unsley and M ash 2008); that is, m easures that are 'good enough to get the job done' (2008: 5). The rationale for adopting a 'good enough' position is the absence o f any com m only accepted guidelines that stipulate a sufficient level of scientific evidence to w arrant use of a particular m easure (see Kazdin 2005; Sechrest 2005). In an effort to find a balance between setting criteria that is either too stringent (and rendering EBA a clinically w orthless exercise) or too lenient (and thereby underm ining the notion o f EBA), the good enough principle has been operationalised using rating criteria that can assess specific categories of psychom etric properties w ith obvious clinical relevance, w ith each category rated as being less than adequate (m easure did not meet the m inim um set criteria), adequate (m easure meets m inim um level of scientific vigour), good (m easure generally possesses solid scientific support), excellent (m easure has extensive, high-quality support evidence), unavailable (research on measure unavailable or not yet published), or not applicable (particular psychom etric property not relevant to measure under consideration).

Assessment purposes Although there are many reasons w hy assessm ents are conducted, H unsley and M ash (2008: 6) have identified seven purposes that 67

Transitions to Better Lives

underpin m ost assessm ents: (a) diagnosis (determ ining the nature a n d / or cause(s) of presenting problems, which may or m ay not involve the use of a form al diagnostic and categorisation system); (b) screening (identifying those w ho have or are at risk for a particular problem and who m ight be helped by further assessm ent or intervention); (c) prognosis and other predictions (generating predictions about the course of the problem (s) if left untreated, recom m endations for possible course of action to be considered, and the likely im pact on that course of the problem ); (d) case conceptualisation/form ulation (developing a com prehensive and clinically relevant understanding of the patient, generating hypotheses regarding critical aspects of the patient's psychosocial functioning and contexts that are likely to influence the patient's adjustm ent); (e) treatment design and planning (selecting/develop ing and im plem enting interventions designed to address the patient's problem s by focusing on elem ents identified in the diagnostic evaluation and the case conceptualisation); (f) treatment m onitoring (tracking changes in sym ptom s, functioning, psychological characteristics, interm ediate treatment goals, a n d /o r variables determined to cause or m aintain the problem ); and (g) treatment evaluation (determ ining the effectiveness, social validity, consum er satisfaction, a n d /o r cost-effectiveness of the intervention). In the view of Hunsley and Mash, these can be further summ arised into three dom ains for the purposes of undertaking EBAs: (a) diagnosis; (b) case conceptualisation and treatment planning (as these tend to rely on the sam e data); and (c) treatment m onitoring and treatment evaluation (as these tend to use the same assessm ent data). Of these, (b) and (c) are the most relevant in terms of understanding the readiness construct.

Psychometric properties and rating criteria A ssessm ent tools can be either idiographic (assess unique aspects of an individual's experience) or nom othetic (assess constructs assumed relevant to all individuals). W hile it is som etim es difficult or irrelevant to apply psychom etric properties to idiographic m easures designed to assess individual change (such as self-m onitoring forms) or treatm ent outcom es (such as treatm ent attainm ent scales), H unsley and M ash (2007, 2008) have argued that where this is the case, EBA should exam ine the extent to which m easurem ent item s and instructions are consistent across assessm ent occasions. W ith respect to nom othetic instrum ents, the authors have argued that the most im portant psychom etric properties in terms of establishing an EBA 68

The assessment of treatment readiness

fram ew ork are as follows: norm s, internal consistency, inter-rater reliability, test-retest reliability, content validity, construct validity, validity generalisation, and clinical utility (see Table 5.1 for a brief description of each category). These categories are applied to both the specific assessm ent purpose (for exam ple case conceptualisation and treatm ent planning) and in the context of a specific disorder or condition (such as depression). H unsley and M ash's (2007, 2008) rating criteria for each of the psychom etric categories is provided in Tables 5.2 and 5.3. The authors point out that w hile it m ight be preferable for practitioners to consider only those m easures that m eet their criteria for 'good ' as a m inim um standard this m ay not alw ays be possible, principally because m easure developm ent is an ongoing process. Inclusion of the 'adequate' criteria allow s for an evaluation of (a) m ore recently developed m easures and (b) m easures where the research evidence across all psychom etric categories is not yet available. Both these factors are relevant in establishing an EBA fram ew ork for treatment readiness given how few m easurem ent tools are available and how little previous w ork has been undertaken in the area. A s noted above, the rating criteria will be used to assess the psychom etric properties of published m easures that purport to assess treatment readiness in two dom ains: first, case conceptualisation and treatm ent planning, where the em phasis is generally identifying an individual's level of m otivation to change, beliefs about w hether the individual believes change is possible, and any barriers to change (w hat has been referred to as 'the w ill' and 'the w ay'; see Ward et al. 2004b); and second, treatm ent m onitoring and evaluation that typically involves an assessm ent of the extent to w hich clients engaged in the treatm ent process, shifts in m otivational states, changes to a n d /o r replacem ent of m aladaptive behaviours, and goal attainm ent. Previous attem pts at establishing EBA s have done so using m easures that have been developed specific for one or both of these purposes (see H unsley and M ash 2008). H owever, given that the m easurem ent of treatm ent readiness in an offender context is a relatively new endeavour, the evaluation process will differ a little from that previously used. Rather than exam ining an array of tools in the context of each assessm ent purposes and offering an evaluation, w hat follow s is a description o f the m ost com m only used measure and an evaluation of each in terms of the assessm ent purpose. The rationale for taking this approach is that com pared to the assessm ent of psychological disorders and psychosocial deficits, there are far fewer measures that purport to assess treatment readiness. Moreover, 69

Transitions to B etter Lives

Table 5.1

P sy ch o m etric p ro p erties for EBA fram ew ork.

C ateg ory

C riteria for u se

N orm s

For stan d ard ised , n o m oth etically -b ased in stru ­ m en ts, n o rm s and criterio n -b ased c u t-o ff scores are n ecessary to en ab le the accu rate interp retatio n o f in d iv id u al test scores. S am p les shou ld be tru ly rep resen tativ e o f the sa m p le p op u latio n from w hich the in d iv id u al is d raw n in term s of d em o grap h ic (e.g. ag e, sex) an d o th er im p o rtan t ch aracteristics (e.g. clin ical v ersu s n o n -clin ical or o ffend er v ersu s n o n -o ffen d er sam p les). U sed to d eterm in e p re- and p o st-treatm en t fu n ctio n in g and ev alu ate w h eth er any ch an g e is clinically m ean in gfu l. R atin g s o f 'a d e q u a te ' requ ire d ata from a sin g le large clin ical sam p le; 'g o o d ' requ ires n o rm ativ e d ata from m u ltip le sam p le s (in clu d in g p o p u latio n sp ecific sam p les); w h ile 'e x ce lle n t' requ ires data from large rep resen tativ e sam p les.

Internal co n sisten cy

A ll item s that pu rp o rt to m easu re a sin gle con stru ct (e.g. treatm en t read in ess) should co n trib u te in a co n sisten t w ay to the d ata o btain ed for that m easu re (item s th at reflect the sam e c o n stru ct should yield sim ila r resu lts). W h ile internal co n sisten cy can b e rep o rted as the av erag e in ter-item correlation , av erag e item total correlatio n , sp lit-h alf reliability, th e m ost co m m o n ly used m easu re is C ro n b a ch 's alp ha (a ).

Inter-rater re liab ility

S im ilar resu lts sh o u ld be obtain ed w h e n a m easu re is used or scored b y a clin ician or researcher. In ter-rater (or in ter-o b serv er) reliability shou ld b e e stab lish ed o u tsid e o f th e stu d y for w hich the resu lts are rep orted (e.g. in a pilot stud y).

T est-retest reliab ility

T h e sam e resu lts should b e ob tain ed if th e test is ad m in istered to Ihe sam e sam p le on tw o d ifferent o ccasion s (i.e. assu m es n o su b stan tial ch an g e in the co n stru ct u n d er in v estig atio n b etw een the tw o occasions). Tw o im p o rtan t cav eats w hen co n sid erin g test-retest reliability: (1) th e tim e betw een m easu rem en t need s to be su fficien t to en su re the o u tcom e is not in flu en ced by tem p oral facto rs (e.g. too s h o rt a period m ay resu lt in p ractice effects); (2) so m e con stru cts are not exp ected to sh o w tem poral stability (e.g. m easu res of state-lik e variables). Table 5.1 continues opposite

70

T he assessment o f trea tm e nt readiness

Table. 5.1 continued C ateg ory

C riteria for u se

C on ten t v alid ity

Item s shou ld reflect the co n ten t d o m ain o f the con stru ct p u rp o rted ly m easured by an instrum en t (item s should rep resen t th e v ario u s asp ects or facets o f the con stru ct an in stru m en t w as d esigned to m easu re) and th e d eg ree to w h ich a test is a rep resen tativ e sam p le of the con ten t of w h atev er o bjectiv es or sp ecificatio n s the test w as originally d esigned to m easu re. To in v estig ate the degree o f m atch , test d ev elo p ers often en list w elltrained co lleag u es to m ak e ju d g em en ts ab o u t the d egree to w h ich the test item s m atched the test objectiv es or sp ecification s.

C o n stru ct v alid ity

A relation sh ip shou ld exist b etw een a theoretical con stru ct and any in stru m en t that p u rp orts to m easu re that con stru ct. A m easu re has strong con stru ct valid ity if it has b o th con v erg en t and d iscrim in an t validity. C o n v erg en t v a lid ity show s an accep tab le level o f ag reem en t b e tw e e n d ifferen t in stru m en ts th at th at p u rp o rt to m easu re the sam e co n stru ct (e.g. scores on tw o in stru m en ts that p u rp o rt to m easu re d ep ression are show n to be h ig h ly correlated ). D iscrim in an t (or d ivergen t) v alid ity tests w h e th e r con stru cts that shou ld n o t be related are, in fact, un related .

Validity g en eralisatio n

E vid en ce for v alid ity gen eralisatio n is d ep en d en t up on a b o d y o f accu m u lated research su p p o rtin g the u se o f a p articu lar in stru m en t across both situ atio n s and p o p u latio n s (the p red icto r or criterio n g en eralises across stu d ies and w ill con tin u e to sh o w sim ilar p aram eters w h e n the situ atio n ch an g es).

C linical u tility

R efers to the ease and efficien cy o f u sin g an assessm en t tool and the (clin ical) relev an ce and m e an in g fu ln ess o f the in fo rm atio n it provides. U tility gen erally co m p rises: av ailab ility and ease o f use; ad m in istratio n tim e; 'le a rn a b ility ' and clin icia n 's q u alification s; fo rm at; sco rin g and in fo rm ation d eriv ed ; m ean in gfu l an d relevant in fo rm ation obtain ed .

71

Transitions to Better Lives

Table 5.2 Norm s Adequate

N orm s and reliability criteria.

M easures o f central tendency and distributions for the total score (and subscores if relevant) based on a large relevant clinical sam ple are available.

Good

M easures of central tendency and distributions for the total score (and subscores if relevant) based on several large relevant sam ples (m ust include data from clinical and nonclinical sam ples) are available.

Excellent

M easures of central tendency and distributions for the total score (and subscores if relevant) based on one or morerepresentative sam ples (m ust include data from both clinical and non-clinical sam ples) are available.

Internal consistency Adequate Preponderance of evid ence indicates a values of .7 0 - 79. Good

Preponderance of evid ence ind icates a values of .8 0 -8 9 .

Excellent

Preponderance of evid ence indicates a values of > .90.

Inter-rater reliability Adequate Preponderance of evid ence indicates k v alu es of .60-.74; preponderance of evidence ind icates Pearson correlation or interclass correlation valu es of .70-.79. Good

Preponderance of evid ence indicates k v alu es of .7 5 - 84; preponderance of evidence indicates Pearson correlation or interclass correlation values of .80-.89.

Excellent

Preponderance of evidence ind icates k. values > .85; preponderance of evidence indicates Pearson correlation or interclass correlation valu es of > .90.

Test-retest reliability Adequate Preponderance of evidence ind icates test-retest correlations of at least .70 over a period of several days to several weeks. Good

Preponderance of evidence indicates test-retest correlations of at least .70 over a period of several m onths.

Excellent

Preponderance of evidence ind icates test-retest correlations of at least .70 over a period of a year or longer.

Source: H unsley and M arsh (2008: 8).

72

The assessment of treatment readiness

no clear distinction has been m ade in the literature regarding the purpose of each m easure (other than to assess readiness). For these reasons, it is more useful to review the elem ents of each m easure and then evaluate each in terms o f the two assessm ent categories. Copies of all m easures reviewed are contained in the Appendix.

Treatm ent readiness assessment tools Stages o f change measures

Given the centrality of m otivation to the treatm ent readiness construct, and in the absence of any offender-specific assessm ent tools, it is not surprising that there has been a heavy reliance of m easures based on the transtheoretical model (TTM ) of change (Prochaska and DiClem ente 1984, 1986). According to the TTM , behaviour change is thought to occur in a series of identifiable stages, the num ber of which differs betw een measures but generally includes som e variation of pre-contem plation (no w ish to ch a n g e/n o recognition of a problem ), contem plation (intention to change problem behaviour w ithin the next six m onths); preparation (intention to take im m ediate action, usually m easured as w ithin the next m onth); action (characterised by specific, overt m odifications w ithin the past six months); m aintenance (relapse prevention); and term ination (change process is co m p lete/n o further need to prevent relapse). An intrinsic com ponent of this stage construct is its developm ental, recursive nature (Begun et al. 2001), w hich typically involves betw een three and seven 'cycles' before long-term m aintenance of the desired change is achieved (Prochaska et al. 1992). M oves towards m aintenance are periodically interrupted by spiralling back to previous stages, w hich is subsequently followed by forward progress. Relapse is not seen as failure, but a predictable pattern in the change process. This allow s any relapse to be refram ed, viewed as a learning opportunity, and m ade available for refining future change and maintenance. O ne of the m ost w idely used o f the stages of change m easures is the U niversity of Rhode Island Change A ssessm ent (URICA) (M cConnaughy et al. 1983, 1989). Com prising of four stages (precon­ tem plation, contem plation, action, m aintenance), this 32-item measure w as originally designed to assess changes in sm oking behaviour but has since been used for a broad range of health-related and addictive behaviours (such as excessive alcohol and drug use; see Brown et al. 2000; Davidson et al. 1991) as w ell as mental health and

73

Transitions to Better Lives

Table 5.3

Validity and utility criteria.

C ontent validity Adequate The test developers clearly defined the dom ain of the construct bein g assessed and assured that selected item s w ere clearly representative of the entire set of facets included in the dom ain. Good

In ad dition to the criteria for an ad equ ate rating, all elem ents of the instrum ent (e.g. instructions, item s) w ere evaluated by ju d g es (experts or pilot research participants).

Excellent

In ad dition to the criteria used for a good rating, m ultiple groups o f ju d g es w ere em ployed and quantitative ratings w ere used by all the judges.

C onstruct validity Adequate Som e independently replicated evidence o f construct validity (e.g. pred ictive validity, concurrent validity, and convergent and discrim inant validity). Good

Preponderance of independently replicated evid ence across m ultiple types o f v alid ity (e.g. predictive validity, concurrent validity, and convergent and d iscrim inant validity) is indicative of construct validity.

Excellent

In ad dition to the criteria for a good rating, evid ence of increm ental validity w ith respect to other clinical data.

V alidity generalisation Adequate Som e evid ence supports the use of this instrum ent with either (a) m ore than one specific group (based on socio ­ dem ographic characteristics such as age, gender and ethnicity) or (b) in m ultiple contexts (hom e, school, prim ary care settings, in-patient sellin gs). Good

Preponderance of evid ence supports the use of this instrum ent w ith either (a) m ore than one specific group (based on socio-dem ographic characteristics such as age, gender and ethnicity) or (b) in m ultiple contexts (hom e, school, prim ary care settings, in-patient settings).

Excellent

Preponderance of evid ence supports the use of this instrum ent w ith (a) m ore than one specific group (based on socio-dem ographic characteristics such as age, gender and ethnicity) and (b) across m ultiple contexts (hom e, school, prim ary care settings, in-patient settings). Table 5.3 continues opposite

74

The assessment o f treatm ent readiness

Table 5.3 continued Treatm ent sensitivity Adequate Som e evidence of sensitivity to change over the cou rse of treatm ent. Good

Preponderance of ind ep endently replicated evidence indicates sensitivity to change over the course of treatm ent.

Excellent

In ad dition to the criteria for a good rating, evid ence of sensitivity to change across different types of treatm ent.

C lin ica l u tility

Adequate

Taking into account practical considerations (e.g. costs, ease o f ad m inistration, availability of ad m inistration and scoring instructions, d uration o f assessm ent, availability of relevant cut-off scores, acceptability to patients) the resulting assessm ent data are likely to be clinically useful.

Good

In ad dition to the criteria for an ad equ ate rating, there is som e published evid ence that the use of the resulting assessm en t data confers a dem onstrable clinical benefit (e.g. better treatm ent outcom e, low er treatm ent attrition rates, greater patient satisfaction w ith service).

Excellent

In ad dition to the criteria for an ad equ ate rating, there is ind ep endently replicated published evidence that the use of the resulting assessm ent data confers a dem onstrable clinical benefit.

Source: H unsley and M ash (2008: 9)

psychotherapy (see McConnaughy et al. 1989; Petrocelli 2002). This breadth has been possible because responses are made to general questions about the individual's 'problem' (for example, 'It might be worthwhile to work on some of my problems') rather than specific issues. Attitudinal differences that characterise each of the four stages of change are assessed, with items used (a) to ascertain a stage 'profile' of the individual (as originally proposed by the developers); (b) a continuous score of readiness (Project MATCH Research Group 1997, 1998); or (c) to identify motivational subtypes (DiClemente and Hughes 1990; M cConnaughy et al. 1983). In the offender context, the URICA has been used with specific populations, for example with offenders with drug use problems (see El-Bassel et al. 1998) and adapted to assess problem-specific stages of change (such as intimate partner/dom estic violence; see Levesque et al. 2000). 75

Transitions to Better Lives

A nother of the stage m easures, the Readiness to Change Q uestionnaire (RCQ) (Rollnick et al. 1992) is a 12-item measure consisting of three stages (precontem plation, contem plation, action). Developed along the sam e lines as the URICA (in fact, som e item s are direct adaptations, with the word 'problem ' changed to 'drinking'), its prim ary purpose is a brief m easure for excessive alcohol use in those w ho may not identify as problem drinkers or those with low levels of dependence, w ho are identified opportunistically in m edical or other settings (that is, it is intended for harm ful and hazardous drinkers not seeking treatm ent from specialist facilities). The Readiness to Change Q uestionnaire [Treatment Version] (RCQTV) (H eather et al. 1999) is a revision of the original m easure, which can be used to assess problem drinkers prior to treatm ent entry. A 15-item instrum ent, it differs from the RCQ in that specific reference is made to total abstinence in addition to reduced drinking. The authors argued that this distinction is appropriate as total abstention is more likely to be a selected goal in treatment populations (whereas it can be seen as a disincentive in those referring only to cut down on alcohol use). Like the U RIC A , the RC Q has been used to assess treatm ent readiness in offender populations. For exam ple, W illiam son and colleagues (2003) used a m odified version of the RCQ to assess violent offenders referred to an anger m anagem ent program m e (by changing the word 'drinking to 'an g er'; for exam ple, 'Som etim es I think I should try and cut dow n on my drinking' w as rephrased as 'Som etim es I think I should try and control my an g er'; 2003: 297). Tests of predictive validity revealed that the instrum ent correctly identified those offenders for w hom the program m e was successful in reducing scores on m easures of anger experience and anger control. The authors suggested that the m easure could help optim ise program m e outcom es by identifying those most suitable to participate as well as helping to facilitate staged-m atched interventions. The Stages of Change Readiness and Treatment Eagerness Scale (SOCRATES) (M iller and Tonigan 1996) is another problem -specific (alcohol) measure of behaviour change. Developed in parallel with other stages of change m easures (such as U RIC A ), the 19-item instrum ent is used to allocate clients presenting for alcohol-related treatm ent to one of three stages: recognition, taking steps, and asm bivalence. M iller and Tonigan claim that unlike m easures such as URICA, the SO CRA TES subscales represent 'continuously distributed m otivational processes that m ay underlie stages of change' (1996: 84) but note that it does not assess all possible m otivational vectors. A recent adaptation by M itchell et al. (2005) produced a 14-item 76

The assessment of treatment readiness

measure for m ilitary personnel seeking treatm ent for alcohol and drug problem s. W hile the factor structure was shown to replicate the original m easure, five of the 19 item s were om itted. The authors explained this in terms of population sample differences; that is, the m ilitary sam ple were younger, more ethnically diverse, more likely to be fully em ployed, presented w ith substance problem s m uch earlier, and therefore experienced m any few er physical sym ptom s associated w ith substance abuse, and the study included participants with com orbid drug and alcohol problems. The first m easure to incorporate treatm ent readiness into an assessm ent of offender risk, the Violence Risk Scale (VRS) (W ong and Gordon 1999-2003), and more recently the Violence Risk Scale-Sex O ffender version (VRS-SO ) (W ong et al. 2003), is used to identify w ho to treat (high -risk/need offenders), what to treat (identify dynam ic risk factors), how to treat (m atching therapeutic approaches to specific stage of change), and w hether there have been post-treatm ent im provem ents (see Wong and Gordon 2006; W ong et al. 2007). In other w ords, the VRS and V R S-SO assess risk, crim inogenic need, client responsivity, and treatm ent change using a single assessm ent process. The V R S com prises six static and 20 dynam ic variables while the VRS-SO has seven static and 17 dynam ic variables. The form er is derived from and theoretically underpinned by the psychology of crim inal conduct (Andrews and Bonta 2003) while the latter is m odelled closely on the VRS, also relying on other sources related specifically to sexual offending (for exam ple H anson and H arris 2000; Proulx et al. 1997; Ward and H udson 1998). For the purposes of risk assessm ent, all variables are rated on a four-point scale (0 ,1 , 2 ,3 ) using file inform ation and a sem i-structured interview. D ynam ic variables identified as treatm ent targets (rated 2 or 3) are then assessed for treatment readiness using a qualitative guide (see Wong and Gordon 1999-2003 for detailed stage descriptors for each dynam ic variable). A llocation to one of the five stages o f change (precontem plation, contem plation, preparation, action, m aintenance) is based on inform ation draw n from the file review and interview regarding the offend er's attitudes, behaviour and affect. A forward progression through the stages (except precontem plation to contem plation) is considered im provem ent (the VRS translates forward progress from one stage to the next as a quantitative risk reduction of 0.5). Personal concerns inventory-offender adaptation (PCI-OA)

The Personal Concerns Inventory-O ffender Adaptation

(PCI-OA)

77

Transitions to Better Lives

(M cM urran et al. 2006, 2008; Sellen et al. 2006, 2009) takes a different approach to m otivation from that described in the transtheoretical model by adopting a theory o f m otivation in w hich goal-striving plays a central role (see K linger and Cox 2004a). An adaptation of Cox and K lin ger's (2002) Personal C oncerns Inventory (PCI), the PCI-OA exam ines the latent processes or ('current concerns') that underpin any active goal pursuit. This exploration o f current concerns involves both the conscious and subconscious cognitive and affective processes that are in operation up to the point w here an individual either attains his or her goal(s) or gives up the goal pursuit. By w ay of exam ple, Sellen et al. (2009) explain how alcohol m ay be a valued goal (and problem drinking understood w ithin an overall goal fram ew ork). First, alcohol use is selected as a goal because of its rew arding properties (for exam ple the pharm acological effect on em otions). The incentive value for continued use is influenced by various factors including individual difference characteristics (the biochem ical responsivity to alcohol, personality traits), social factors (parental m odels, social norm s), and context (the availability of sim ilarly rew arding, non­ drug alternatives). Over time, this incentive value m ay change. For exam ple, excessive use m ay lead to addiction a n d /o r changes in individual circum stances (job loss or fam ily breakdow n). The PCI-OA requires offenders to describe their current concerns in the 12 life areas from the PCI (self changes; em ploym ent and finance; partner, fam ily and relatives; education and training; hom e and household m atters; substance use; friends and acquaintances; health and m edical m atters; hobbies, pastim es and recreation; love, intim acy and sexual; spiritual m atters; other areas) as w ell as two additional areas that are offender-specific (my offending behaviour; current living arrangem ents). As with the PCI, each area is rated in terms of value, attainability, im m inence and controllability (Cox and Klinger 2002). The PCI-OA also requires that offenders to rate how (a) offending and (b) being in prison helps or interferes with the attainm ent of their goals. Responses have been shown to fall along three dim ensions. The first two, adaptive motivation and maladaptive motivation, are consistent w ith the PCI. An adaptive m otivational profile is characterised by high levels o f perceived likelihood regarding goal attainm ents, expectations o f happiness when goals are attained, and com m itm ent to goal strivings. A m aladaptive m otivational profile, on the other hand, is characterised by having goals that are seen as unim portant, expecting little happiness at goal achievem ents, and having low com m itm ent to goals. The PCI-O A has also been found to have a third factor - lack of direction - w hich appears to reflect 78

The assessment of treatment readiness

unhappiness in the face of goal achievem ent, difficulty in know ing how to achieve goals, and a perception that prison and offending may be helpful in terms of goal achievem ents. Scores on each of the indices can then be calculated. An adaptive m otivation (AM ) index is calculated by subtracting the sum of variables with negative loadings from the sum of variables w ith positive loadings, and the m ean then calculated. The m aladaptive m otivation (MM ) index is calculated by subtracting the sum of variables with positive loadings from the sum of variables w ith negative loadings, and calculating the m ean for that subscale. Finally, lack of direction (LoD) is calculated by subtracting the sum of the rating scales w ith positive loadings from the sum of the scale with a negative loading. Treatment readiness measures

A lthough Serin and Kennedy (1997; Serin 1998) developed w hat was perhaps the first offender-specific assessm ent of treatm ent readiness, there has been little reference in the literature to their sem i-structured interview outside that published by Canadian Corrections (see Chapter 2). W illiam son et al. (2003) adapted the interview schedule to be used as a brief (11-item) questionnaire - the Serin Treatment Readiness Scale (STRS) - and w hile m oderately high correlations were noted betw een this m easure and their A nger Readiness to Change Q uestionnaire (A RCQ, adapted from Rollnick et al. 1992), no other psychom etric properties were reported. More recently, Serin, M ailloux and K ennedy (2007) have developed a clinical rating scale for offender readiness: the Treatment Readiness Clinical Rating Scale (TRCRS). The general prem ise that underpins this measure is sim ilar to that espoused in the M O RM (Ward et al. 2004); that is, treatm ent readiness can be influenced by a range of internal and external factors (such as treatm ent setting, individual offender characteristics, treatment intensity, m otivational issues). Serin et al. have also taken the position that any treatm ent response is increm ental in nature, a reflection of general treatability issues (that is, readiness and participation) w hich in turn provides an indication of the overall treatment effect (the treatm ent gains and the generalisation of new know ledge and skills to new situations). G iven the proposition that treatm ent readiness and interpersonal style are related and serve to influence program m e perform ance (see Serin and K ennedy 1997), the 16-item m easure developed by Serin et al. (2007) com prises eight item s that reflect these two dom ains. The reported factor analysis confirmed the existence of each dom ain as

79

Transitions to Better Lives

an independent, albeit correlated, subscale. Item s are representative of each dom ain (for exam ple, problem recognition and treatment goals reflect the treatm ent readiness dom ain; pro-crim inal view s and denial reflect interpersonal style), w ith responses m ade using a fourpoint Likert scale. Unfortunately, the authors have not provided any further inform ation relating to item content or how scores on the measure m ight be calculated and subsequently used to assess levels of treatm ent readiness. They do note, however, that unpublished research has shown that the dom ains differentiate offenders w ith respect to program m e attrition, recidivism and types of sexual offenders. The m easure has also be included in the evaluation fram ew ork o f a m otivational prim er for violent offenders (see Blanchette and M oser 2006). The Corrections Victoria Treatment Readiness Questionnaire (CVTRQ) (Casey et al. 2007) is a generic screening m easure that can be used to assess readiness for treatm ent prior to program m e entry. W hile not a test of the m odel per se, item content is derived from the internal factors identified in Ward et al.'s (2004b) m ultifactor offender readiness m odel on treatm ent readiness (for exam ple, beliefs about treatm ent, past experiences of program m es, offender goals). A n attem pt to distinguish treatm ent readiness from treatment m otivation and responsivity, the M ORM subscribes to the definition of readiness proposed by H ow ells and D ay (2003) as 'the presence of characteristics (states or dispositions) w ithin either the client or the therapeutic situation, w hich are likely to prom ote engagem ent in therapy and which, thereby, are likely to enhance therapeutic change' (Ward et al. 2004b: 650). To be ready for treatm ent, the individual needs to be m otivated (w ants to, has the will to), has the ability to respond appropriately (perceives he or she can), finds it relevant and m eaningful (can engage), and has the capacities (i.e., is able) to successfully enter the rehabilitation program m e. Readiness is, therefore, perceived as m ore inclusive than either m otivation or responsivity and is, in fact, thought to encom pass both these constructs. The 20-item CV TRQ consists o f four subscales: attitudes and m otivation (attitudes and beliefs about program m es and the desire to change); em otional reactions (em otional responses to the individual's offending behaviour); offending beliefs (beliefs about personal responsibility for offending behaviour); and efficacy (perceived ability to participate in treatm ent program m es). Responses are m ade using a five-point scale, w ith higher scores indicating higher levels of readiness (scores can range from 20 to 100). Initial RO C analysis 80

The assessment of treatment readiness

suggested a cut-off score o f > 7 2 , w hich provides a sensitivity of 69.49 and specificity of 59.38 (which represents a positive predictive value of 61.2 per cent and negative predictive value of 67.9 per cent). In the absence of large-scale predictive validity studies, the authors suggest that cut-off scores m ight reasonably fall betw een 68 and 74. The choice of cut-off w ill ultim ately depend on a cost-benefit analysis. If the main aim is to m axim ise the num ber of clients ready for treatm ent program m es, then choosing a cut-off that m axim ises sensitivity is generally preferable (but will lead to the inclusion into the program m e of more clients w ho are not treatm ent ready). Conversely, m axim ising specificity w ill ensure that m ost of those included in the program m e are, in fact, treatm ent ready (although this may result in lack of program m e access for those individuals who are indeed ready to change). Validation of the m easure also revealed that the attitudes and m otivation subscale was m ost strongly related to overall treatm ent engagem ent (the outcom e variable in the study). W hile suggesting that this aspect of readiness is likely to be particularly im portant, the profile of scores across subscales is likely to reveal those aspects of readiness that are salient for an individual. Interventions that target low readiness can therefore be targeted towards specific deficits identified by the measure. Finally, the Violence Treatment Readiness Q uestionnaire (VTRQ) (Day et al. 2009a) is an adaptation of the CV TRQ that can be used to screen violent offenders prior to program m e allocation. The only m odification to the original m easure was to change item w ording to reflect violence rather offending in general (for exam ple, 'I have not offended for some tim e now ' w as replaced w ith 'I have not acted violently for som e tim e now')- As with the general m easure, a m oderately high correlation was found betw een readiness as measured by the VTRQ and self-reported treatment engagem ent. In fact, this relationship w as stronger than that found for other measures of readiness used in the study to assess convergent validity, nam ely the RCQ (Rollnick et al. 1992) and STRS (W illiam son et al. 2003). The finding that post-treatm ent readiness scores were significantly higher than pre-treatm ent scores illustrates the dynam ic nature of the construct, increasing over the course o f program m e participation. W hile it may not be surprising that by the end of treatm ent participants can dem onstrate changes in their attitudes and m otivation, em otional reactions to their offences, offending beliefs and efficacy, it does point to an additional use of the VTRQ as a m easure of change in interventions that are designed to increase problem aw areness and m otivation prior to entry in structured treatm ent program m es. 81

Transitions to Better Lives

Overall evaluation: case conceptualisation and tre a tm e n t planning P erhaps the m ost im portan t point to be m ade w hen con sid erin g the suitability of readiness m easures for the pu rposes of case form ulation and treatm ent planning is the need to utilise m ore than a single m easure in order to obtain accu rate and m ean in gful inform ation (see H unsley and M ash 2008; Serin et al. 2007). For exam ple, m easures based on the transtheoretical m od el (Prochaska and D iC lem ente 1984, 1986), w hile suitable for d eterm ining an o ffen d er's m otivational state, are less suitable for estab lish in g treatm ent readiness (see Casey et al. 2005). Sim ilarly, the PC I-O A (M cM u rran et al. 2008a; Sellen et al. 2006, 2009) enables the p ractitioner to assess m o tivation and d eterm ine w hether there is a fit betw een an o ffen d er's goals and their crim inogenic need (s), but there is no provision in the instrum en t to d irectly assess read in ess for treatm ent. T his latter issue is addressed in the C V TR Q (C asey et al. 2007) and V T R Q (D ay et al. 2009a), b u t even these m easures need to be used in con ju n ction w ith, for exam ple, a m easure of m otivation, to ensure that the o ffen d er's read in ess is not a function of external pressure to enter treatm ent (such as for parole purposes). The biggest d ifficu lty in ad optin g an E BA ap proach is the lim ited nu m ber o f m easures from w hich to choose and the paucity o f evid ence regarding their psychom etric properties. This highlights the need for p ractitioners to ensure that the m easures selected are both relevant to the task at hand and psychom etrically sound. That said, the m ajority o f m easures described herein are av ailable at no cost and require little or no training for ad m inistration (the PC I-O A is perhaps the m ost difficult); the stages o f change m easures, V C TR Q and V TR Q , take only a few m inu tes to com plete. W hile the U R IC A (M cC onnau ghy et al. 1983) is one of the m ost w idely used m easures o f m otivational read in ess and d isplays ad equ ate levels o f clinical utility, it has poor v alid ity generalisation w ith resp ect to an offend in g context. The m ajor lim itation is a failure to clearly define 'p ro blem ' behaviou rs; first, som e offend ers m ay not perceive their beh av iou r as problem atic, and second , even w hen b ehav iou rs are acknow ledged as b ein g cause for concern, if the offend er presents w ith m ore than one problem (for exam ple poly-substance abuse, co-m orbid d iagnosis), d ifficu lties m ay arise in term s o f question interpretation and subsequ ent responses. A nother frequently used stage o f chang e m easure is the RC Q (Rollnick et al. 1992) and, m ore recently, the R C Q -T V (H eather et al. 1999). A lthou gh the p sychom etric properties are slightly stronger than the U R IC A (see 82

The assessment of treatment readiness

Table 3.4) and the m easure has been adapted for offender populations (see M cM urran et a l 1998; W illiam son et al. 2003), there appears to be som e confusion about w hich of the RCQ m easures is most appropriate. The original RCQ instrum ent was developed to assess m otivation to change am ong excessive alcohol consum ers identified opportunistically in m edical settings and is the one m ost com m only adapted for offender populations (M cM urran et a l 1998; W illiam son et al. 2003). A decision to use this m easure at the case conceptualisation and treatm ent planning stage, w hether to assess for alcohol or drug use or adapting it for use as an offence-specific m easure, needs to be inform ed by the offend er's beliefs about the problem behaviour. In other words, it is im portant to determ ine w hether it is the offender w ho has self-identified the behaviour as problem atic or w hether it is the practitioner's assessm ent of that behaviour. For exam ple, if an offender presents w ith a crim inogenic need relating to anger a n d /o r violence and recognises that his or her behaviour is problem atic, it would be m ore useful to adapt the treatm ent version of the RCQ, as it includes reference to abstaining from the problem atic behaviour. On the other hand, if in the process of case conceptualisation the practitioner identifies substance use as problem atic but the offender does not, it m av be more appropriate to use (or adapt) the original RCQ. O n face value, the integration o f stages of change, risk /n eed assessm ent, and treatm ent outcom es in the VRS (W ong and Gordon 1999-2003, 2006) and VRS-SO (W ong et al. 2003) would seem to provide a broad (m ulti-m ethod) approach to assessm ent. In term s of case conceptualisation and treatm ent planning, static variables can provide an em pirical-actuarial assessm ent of risk w hile the dynam ic variables identify areas of crim inogenic need that should be the target for intervention(s). Readiness can also be (qualitatively) assessed prior to treatment. The lim itation of the instrum ent in terms of assessing treatment readiness, from an EBA perspective at least, is the absence of any psychom etric evaluation o f the readiness com ponent of the measure. Thus w hile both the VRS and VRS-SO have been shown to be psychom etrically sound (see O lver et al. 2007; Wong et al. 2007; Wong and Gordon 2006), it is not possible to consider readiness as an independent property of the m easure other than to note that it has adequate content validity and validity generalisation. The qualitative nature of interpretation may also be problem atic, particularly when relying on responses to a sem i-structured interview. A s the lim itations of using stages of change m easures to assess treatment readiness have been discussed in m ore detail elsewhere 83

Transitions to Better Lives

(Casey et al. 2005; Littell and Girvin 2002; Tierney and M cCabe 2001), an extensive critique will not be given in this chapter. In brief, however, the m ain issues to consider are as follows. First, factors that m ight influence an individual's m otivation to change m ay differ m arkedly betw een a prison and non-prison environm ent. A n incarcerated offender m ay have little internal m otivation to change their behaviour, but the external m otivators m ight be such that the advantages of participating in a treatment program m e far outw eigh the disadvantages. Second, there is considerable difficulty in m onitoring any change in crim inal behaviour where the offender is incarcerated. Crim inal offending, particularly in a prison environm ent, will not only occur less frequently, b ut when it does occur, is typically less likely to be observed. M oreover, the artificial environm ent in which offenders live may contribute to any reduction in frequency, even a tem porary rem ission of offending behaviour. Third, there is a tendency for offender rehabilitation program m es to adopt a 'one size fits all' approach, which is inconsistent with evidence from the clinical arena that suggests stage-m atched treatm ents result in more successful outcom es. Finally, and perhaps m ost im portant from an EBA perspective, allocation can differ as a function of the scoring method used. The sim plest interpretation is placing the respondent in the highest scoring stage (which can be problem atic w hen the individual scores are equally high in more than one stage). It is possible to calculate readiness to change com posite scores by sum m ing the contem plation, action and m aintenance scores and then subtracting the precontem plation score. The Italian Com posite scores are calculated by subtracting the sum o f the precontem plation and contem plation scores from the sum of the action and m aintenance scores, w hile the com m itted action com posite scores are calculated by subtracting contem plation scores from the action scores (see M cM urran et al. 2006). Despite problem s with stages of change m easures, if the purpose of assessm ent is confined to m otivation to change problem atic substance use, the SOCRATES (M iller and Tonigan 1996; M itchell et al. 2005) m easure is psychom etrically m uch stronger than either the U RICA or RCQ. It provides inform ation on three dim ensions of m otivation, nam ely problem recognition (acknow ledgem ent that that there is a problem related to drugs or alcohol); am bivalence (un certainty/ openness to reflection as to w hether substance use is a problem ); and taking steps (w hether steps have already been taken to change drinking or drug-related behaviours). The narrow er focus on m otivation rather than on stages of change also serves to increase its utility; that is, 84

The assessment of treatment readiness

rather than m easuring the stage constructs as conceived by Prochaska and DiClem ente (1983, 1986), it explains the 'continuously distributed m otivational processes that m ay underlie stages of change' (M iller and Tonigan 1996: 84). In an offending context, the SOCRATES measure has been shown to reliably predict later substance use (see Prendergast et al. 2009) but not substance-related offending (N ochajsk and Stasiewic 2005; Prendergast et al. 2009). Looking next at the PCI-OA (Sellen et al. 2009), its utility would seem to be in understanding the concerns offenders m ay have and identifying positive goals rather than being a direct assessm ent of w hether to enter either offence-specific or generalist program m es. Moreover, while the authors claim that the sim ilarity in factor structure betw een this and Klinger and C ox's (2004a) original instrum ent (adaptive and m aladaptive m otivation subscales) and this adaptation supports the structural validity, the poor psychom etric properties of the lack of direction (LoD) subscale (o. = .36) underm ines this claim. In order to m eet EBA criteria one would expect at least adequate levels of internal consistency - only the adaptive m otivation subscale meets this criterion - for specific populations or offence types (H unsley and M ash 2008). A further lim iting factor (in term s of EBA) is that while the authors provide a method for calculating scores for each of the indices, no cut-off scores are provided. N or is there provision in the scoring to accom m odate high or low scores on both the adaptive m otivation and m aladaptive m otivation. An offender w ho scores high on both m otivational scales can have the same score as som eone w ho has scored low on both. Interpretation is left to the practitioner, w hich could in turn reduce its clinical utility. The strongest argum ent for using the m easure is at the case conceptualisation stage where inform ation about the offend er's goals can be considered in the context of overall risk assessm ent and identified crim inogenic and non-crim inogenic need. For the purposes of screening offenders prior to treatm ent entry, the CV TRQ (Casey et al. 2007) and VTRQ (Day et al. 2009a) both display adequate levels of reliability, validity (including generalisation validity) and clinical utility. The inclusion of cut-off scores means that w hen used in conjunction w ith an offend er's risk /n e e d s profile, practitioners can m ake a decision about treatm ent referral that is based on em pirical evidence of the relationship betw een readiness and engagem ent in the treatm ent process (that is, predictive validity). W here an offender is not treatm ent ready, scores on any of the four com ponents (attitudes and m otivation; em otional reactions; offending beliefs; efficacy) provide specific targets for im proving readiness. W hat 85

CTn

Table 5.4

Ratings of instruments for case conceptualisa lion and treatment planning.

Instrument URTCA RCQ/RCQ-'TV SOCRATES PCT-OA TRCRS V R S/V R S-SO C V TR Q /V TR Q

Internal consistency

Inter-rater reliability

+

++

NA

+

++



+++ -

+++

+

++

NA U U

U +

U ++

u u

Norms

+

lest-re lest reliability

Content validity J

Construct validity *

Validity generalisation

Clinical utility

U +

U

++

_

+

+++

+

-

+

+++

+++

+



u u u u

+

+

U

+ -

++ + +

u u

u

-

+ ++

U ++

+

URICA = University of Rhode Island Change Assessment; RCQ = Readiness to Change Questionnaire; SOCRATES = Stages of Change Readiness and Treatment Fagerness Scale; PCI-OA = Personal Concerns Inventory - Offender Adaptation; TRCRS = Treatment Readiness Clinical Raling Scale; VRS = Violence Risk Scale; CVTRQ = Corrections Victoria Treatment Readiness Questionnaire; VTRQ = Violence Treatment Readiness Questionnaire. + = Adequate; + + = Good; + + + = Excellent; NA = Not applicable; U = Unavailable; - = Less than adequate.

Transitions to Better Lives

00

The assessment of treatm ent readiness

this m eans at the case conceptu alisation and treatm ent planning stages is that practitioners can adopt the approach advocated b y W eisz et al. (2004) o f tre at-re asse ss-ad ju st treatm ent. For exam ple, offenders low in treatm ent read iness can receive an intervention designed to increase readiness (such as m otivational interview ing), have their level of read iness reassessed , and w hen a sufficient level of change is noted be m oved to a crim inogenic needs program m e. C on sistent w ith the m ulti-m ethod approach o f E BA , using this m easure to determ ine treatm ent referral m igh t be d one in con jun ction w ith (a) an assessm ent o f the in d iv id u al's m otivation to change (such as SO CRA TES, for substance use) to ensure that there is a correlation betw een m otivation and treatm ent readiness; and (b) an assessm ent o f the ind iv id u al's needs or goals using, for exam ple, the PCI (K linger and C ox 2004a), the PC I-O A (Sellen et al. 2006, 2009), or the G ood L ives M odel (Ward and Stew art 2003).

Overall evaluation: tre a tm e n t m onitoring and evaluation A ssessm ents d u ring treatm ent and follow -u p are n ecessary to determ ine the effectiven ess o f any intervention and w hether the changes noted im m ed iately follow ing treatm ent are sustained ov er tim e. A s noted above, the provision o f program m es to offend ers w ho are treatm ent ready is not only a better use of scarce resources, it is also likely to result in sustained treatm en t outcom es (as evidenced by red uctions in recidivism ). T h e choice o f instrum en t used for the purposes o f treatm ent m onitoring and evalu ation will depend on w hether the p ractitioner w ants to assess im provem ents in treatm ent readiness (follow ing specific in terv ention s to increase readiness) or w hether the goal is to evalu ate the effectiven ess o f an offence-specific program m e (for exam ple the relationship betw een treatm ent readiness, therapeutic alliance and behaviou ral change). A nother im portant con sid eration for assessm en t is that like m otivation, read in ess is d ynam ic in nature. A n offend er w ho is treatm ent ready at the com m encem ent of a program m e m ay w ax and w ane over the course o f the program m e (as a result, for exam ple, of shifts in m otivation, changin g attitu d es tow ards treatm ent, levels of eng agem ent w ith program m e content, beliefs about self-efficacy, external pressure, therapeutic alliance). T he stages of change m easures are m o st useful for m on itoring progress in treatm ent d esigned to im prove read iness (such as m otivational interview ing). O f the m easures available, the tw o w ith 87

Transitions to Better Lives

m ost evidence of treatm ent sensitivity are SO CRA TES (M iller and Tonigan 1996) and U RICA (M cC onnaughy et al. 1983, 1989). W hile SOCRATES would be more strongly recom m ended for offenders w ith substance abuse disorders (given its stronger psychom etric properties), the URICA m ay be m ore appropriate for other offence types (particularly if the 'problem beh av iou r' is clearly identifiable). Both the VRS (W ong and Gordon 1999-2003) and VRS-SO (Wong et al. 2003) enable the practitioner to m onitor the extent to which any new ly acquired positive attitudes and coping skills are stable, sustainable, and generalisable. Forward progression in treatm ent is subsequently used to redefine risk by at the com pletion of treatm ent (using an algorithm ). D espite the lack of psychom etric evidence for the states of change com ponent of this m easure, outcom e studies have shown that offenders assigned to treatm ent on the basis o f their VRS or V R S-O S scores do have better outcom es in terms of reductions in recidivism rates (see, for exam ple, Wong et al. 2003, 2007, 2005). A lthough the psychom etric properties of the PCI-O A (Sellen et al. 2009) are not strong, Sellen and her colleagues have noted m odest changes over the course of treatm ent in adaptive m otivation (with corresponding reductions in the m aladaptive m otivation and lack of direction subscales). G iven the higher level of sensitivity to change noted in the PCI (Klinger and Cox 2004a), this measure is perhaps preferable to the PCI-O A for assessing w hether treatm ent has resulted in a downward shift in personal concerns, at least until further validation of the PCI-O A is undertaken. Finally, in addition to screening offenders for program m e suitability, the CV TRQ (Casey et al. 2007) and VTRQ (Day et al. 2009a) can be used for both treatment m onitoring and outcom e evaluations. For exam ple, cut-off scores can be used to assess changes follow ing treatm ent designed to improve readiness (such as experiential and dram a therapy, m otivational interview ing). In terms of treatm ent m onitoring, Ward et al. (2004b) have argued that engagem ent in treatm ent is an interm ediate goal before change in crim inogenic need takes place. Therefore, an assess­ m ent of the relationship betw een treatm ent readiness and treatment engagem ent m idw ay through a program m e would enable practitioners to m onitor treatm ent efficacy.

Conclusions and future directions As noted at the beginning o f this chapter, despite the im portance of treatm ent readiness to the rehabilitative process, there rem ains a need

Table 5.5

Ratings of instruments for treatment monitoring and evaluation.

Instrument

+ +++ U -

+

++ +++ U + ++

NA NA U U

u

Test-retest reliability

Content validity

U +++

U +++ + + +

u u u

Construct Validity Treatment Clinical validity generalisation sensitivity ii lililv td

++ + U + +

_ -

+ U

+ ++ U

+ + U

-

-

+

-i-

URICA = University of Rhode Island Change Assessment; SOCRATFS = Stages of Change Readiness and Treatment Eagerness Scale; VRS = Violence Risk Scale; PCI-OA = Personal Concerns Inventory-Offender Adaptation; CVTRQ = Corrections Victoria Treatment Readiness Questionnaire; VTRQ = Violence Treatment Readiness Questionnaire. + = Adequale; ++ = Good; +++ = Excellent; NA = Not applicable; U = Unavailable; - = Less than adequate

00 NO

The assessment of treatment readiness

URICA SOCRATES VRS/VRS-SO PCI-OA CVTRQ/VTRS

Norms

Internal Inter-rater consistency reliability

Transitions to Better Lives

for theoretical advancem ent in the area. W hile this is im portant from an evidence-based approach to assessm ent, it is equally im portant in terms of gaining a better understanding of the construct and how it can (or should) be measured. Readiness is a com plex and dynam ic phenom enon: an individual is ready for som ething. In the case of offender treatm ent, he or she is ready to engage in a process (treatm ent) that will bring about behavioural change (non­ offending). This highlights two im portant considerations in terms of assessm ent. First, readiness can change over time. Som eone who is ready for treatm ent prior to the com m encem ent of a rehabilitation program m e may, for a vast num ber of reasons, shift along the readiness continuum throughout the treatm ent process. It may be, for exam ple, that readiness issues (such as anxiety about self­ disclosure in a group settings) are resolved in the early stages of a treatment program m e. For others, their experiences of the treatment in the early sessions may entrench and reinforce low levels of readiness (beliefs that correctional treatment isn 't confidential, or isn't effective, for exam ple). Second, its com plexity necessitates m ulti-m ethod assessm ent. The m easures described herein all provide inform ation about the individual but none provides the practitioner w ith a com prehensive assessm ent of all facets of readiness. There is a lim it to the extent to w hich self-report measures of the type described in this chapter can provide a com prehensive assessm ent of an individual's preparedness and ability to engage in treatment. Psychom etric assessm ents are one part o f a broader assessm ent of readiness, w hich should take into consideration an individual's level of risk and presenting needs. There is, however, an urgent need to undertake and publish validation studies using the m easures already available, as these m easures offer the practitioner a way of m aking decisions that are both informed and reliable. In a forensic context it is critical that practitioners are able to publicly defend their clinical decision-m aking. The research literature as it currently stands is an im portant first step, but if evidence-based assessm ent does underpin evidence-based treatm ents, then scientific evidence should guide the selection of readiness instrum ents in the same w ay as it should guide any other form of clinical assessm ent.

90

P art Two

Readiness and Offenders

Chapter 6

Interpersonal violence: the need fo r individualised services

It is w idely accepted that one of the most im portant initial tasks in violent offender treatm ent program m es is finding a w ay to engage participants in a therapeutic process. This is seen as a necessary, and perhaps even sufficient, condition for change to take place. In Chapter 12, Christina Kozar review s evidence to suggest that at least a quarter of the therapeutic change observed can be directly attributed to the nature of the relationship form ed betw een the client and the treatment provider. Kozar notes that clients w ho do not experience a collaborative purposeful therapeutic relationship tend to either leave treatm ent or do not m ake as m any gains as other clients, and this would appear to be particularly true for m any violent offenders w ho are referred to rehabilitation program m es (Ross et al. 2008). M any violent offenders do not readily view their aggression and violence as problem atic, and participants in violent offender treatment com m only (if not typically) experience serious, com plex problem s, personality disorders, and possibly outright m ental illness, which potentially leaves them unreceptive to intervention and likely to drop out of or not com plete program m es. This is a serious issue, given evidence to suggest that the risk of reoffending is increased follow ing the non-com pletion of violent offender treatm ent program m es. In their w ork D ow den, Blanchette and Serin (1999) have show n that program m e drop-outs have the highest rate of violent reoffending (40 per cent), com pared with both untreated (17 per cent) and treated (5 per cent) groups. A m ore recent review of offender treatm ent non­ com pletion across a wide range o f offender treatm ent program m es produced sim ilar findings (M cM urran and Theodosi 2007). High 93

Transitions to Better Lives

attrition rates are com m only reported in evaluations of program m es for violent offenders, particularly when program m es are offered in the community. A m ong court mandated dom estic violence offender program m es, for exam ple, non-com pletion rates of up to 75 per cent have been reported (Buttell and Carney 2008), although the com pletion rate of approxim ately 55 per cent for 16-week group-based program m es reported by G ondolf (2008) is probably more typical. Such findings underscore the im portance of considering the nature of treatm ent readiness in violent offenders, such that appropriate placem ent into rehabilitation program m es takes place. A num ber of the factors are likely to influence the w ays in which violent offenders engage with rehabilitation program m es. These are described elsew here in this book. For exam ple, attitudes and beliefs about offending and violence that potentially underm ine readiness are described in Chapter 3. M any violent offenders present with difficulties in regulating their em otions, notably anger, and in the treatm ent context this may also im pact on readiness (Chapter 3). In addition, som e violent offenders will also have long-term difficulties, som etim es associated w ith personality disorders (Chapter 9), that m ake the task of program m e delivery particularly challenging. The aim of this chapter, however, is not to repeat this m aterial, but to consider the need to match the needs of the individual offender to the particular type of program m e being offered. Too often correctional agencies m ake referrals to program m es on the basis of offences, and fail to adequately consider the heterogeneity that exists in offender groups. This can result in the delivery of standardised program m es that can be seen as irrelevant or unnecessary by participants. As such, even if m otivation exists to change aggressive and violent behaviour, the lack of an appropriate program m e m ight inhibit readiness. Day and H ow ells (2008) describe the problem as follows: It is tem pting (and som etim es adm inistratively convenient), in setting up treatm ent program m es for violent (or other) offenders, to define the problem topographically and to use topographic descriptions to allocate offenders to program m es. Thus 'sex offenders' would be directed to a sex offender program m e, 'violent offenders' to a violence program m e, and so on. The dangers of such an approach, although these can be offset by practical advantages, lie in the im plicit assum ption that individuals in the offender group share features and needs that go beyond the fact that they have all engaged in violent acts. (2008: 17) 94

Interpersonal violence: the need for individualised services

In the past decade evidence has increasingly dem onstrated that violent offenders are indeed heterogeneous in term s of their needs and the causal influences on their offending behaviour. Serin (2004), for exam ple, has suggested that violent offenders are sufficiently heterogeneous that endeavours to distinguish am ong types of offenders should be strongly encouraged and that current strategies to assign all violent offenders to anger m anagem ent program m es is both inefficient and ill-advised. Similarly, in his com prehensive review of treatm ent outcom es for violence program m es, M cG uire (2008) concludes that different sub-groups m ay be differentially responsive to existing treatm ents. The argum ent put forward in this chapter is that offering program m es that are m ore closely aligned with the needs of participants (or 'functional' approaches to treatm ent allocation, see Sturm ey 2007), is likely to lead to higher levels of engagem ent, participation, com pletion and ultim ately reduction in risk. Indeed, as a group, violent offenders can be considered particularly heterogenous in relation not only to their offences, but also to a range of other factors that include the triggers of their aggression, the role that em otional arousal plays, and gender differences. These are discussed below. Offence type

Nowhere is the heterogeneity of violent offenders more obvious than in relation to the nature of the offence. Sexual violence, for exam ple, is w idely acknow ledged to require different understandings from general violence.1 Sexual offences are som etim es classified as violent offences, and for som e offenders (for exam ple, those who com mit serious violent acts that have a sexual com ponent) the decision w hether to allocate a sexual offender or a violent offender rehabilitation program m e is not alw ays easy to make. W hichever choice is made, the offender m ay regard the referral as inappropriate to his or her ow n circum stances, and be reluctant to engage or participate. Violence that occurs w ithin fam ilies or intim ate relationships also has a num ber of features that potentially distinguish it from other form s o f violent offending. It typically occurs, for exam ple, behind closed doors, where the only w itnesses are victim s, and as a result the facts o f the offence are often difficult to establish. Even after conviction, offenders com m only dispute the evidence that is presented to the police a n d /o r the courts in relation to the frequency and intensity

‘Readiness in sexual offenders is discussed separately in the next chapter. 95

Transitions to Better Lives

of their violence, the reliability o f victim statem ents, and the extent to which they consider them selves personally responsible. M andated referral to a dom estic violence treatm ent program m e is thus often seen as further evidence that they have been treated either unfairly or unjustly (O 'Leary et al. 2009). In interview s w ith m en w ho had ju st begun a com m unity-based dom estic violence program m e, O 'Leary et al. (2009) reported that m any men expressed a sense of injustice about being there, because they believed, at the very least, that their partners were equally responsible for the circum stances that led to their order to attend the program m e. One participant spoke of attending the program m e 'under duress'. H e described it as follows: 'to be perfectly frank, I'm unfortunately in the situation where I d on't think I'm suitable for the course. I'm not innocent in relation to the few areas where I've crossed the line but w hat I am com pletely shocked at is the system where the other party is not brought to account' (2009: 170). W hen men were asked about the court or correctional services ordering them to attend the program m e, it w as not uncom m on for m en to be unclear about this process. M any men com m ented that they felt the length of the program m e w as excessive, w ith one m an com m enting that he had been put on bonds that had not lasted as long as the program m e. There was some apprehension from men about content and the fact that they would be in a group w ith other men. The follow ing quote highlights one m an's confusion about the referral process and program m e: 'Just a bit confused I suppose, no one told m e w hat its all about, I've waited about 6 m onths and then I find out it is 24 w eeks I have to com e or som ething, but I am curious as w ell, yeah, more than anything else, yeah. Yeah, I haven't got that I shouldn't be here attitude ... Just d on't know w hat to expect' (2009: 175). M ost m en in this study, however, did not com pletely dism iss the potential for benefits to arise from their attendance at the group, som etim es even in the face of claim s of innocence and unfairness from the m andatory order. There has been a great deal of discussion concerning the extent to w hich m entally disordered offenders have specific risk factors that differentiate them from general offender groups (see H ow ells et al. 2004a). The co-occurrence o f m ental disorder and violent behaviour is w idely acknow ledged, but the nature of the causal links (if any exist) between the two classes o f phenom ena rem ains controversial and uncertain, despite research efforts over several decades. Bonta, H anson and Law (1998) conducted a m eta-analysis of studies predicting recidivism in m entally disordered offenders and found that broadly 96

Interpersonal violence: the need for individualised services

sim ilar factors predicted reoffending in the m entally disordered as in non-disordered offenders. However, there are also a num ber of risk factors for violence that may be specific to disordered offenders. First, there is som e evidence to support the inclusion of the role of psychotic sym ptom s and substance abuse as crim inogenic, at least for some patients. Particular psychotic sym ptom s such as paranoia (Grossm an et al. 1995), and delusions involving personal targets (N estor et a l 1995), have been associated w ith violent behaviour. There have also been suggestions that experiences involving a loss of self-control (for exam ple, thought insertion) may lead to a loss of constraint on behaviour. Link and Stueve (1994) reported that patients who felt threatened by others and were unable to have control over their ow n thoughts were twice as likely to have been violent than those who reported other psychotic sym ptom s. Sw anson, Borum and Swartz (1996) reported that a com bination of substance abuse and these particular sym ptom s very strongly predicted future violence. Q ualitative research (for exam ple, Cham bers et al. 2008) has further suggested that a num ber of different causal pathw ays to violent offending can be identified, and that violent offences may be best conceptualised in term s of individual goals w ith respect to the offending (approach versus avoidance) and the m anner in which the individual attem pts to achieve these goals (passive versus active; see Ward et al. 2004a; Yates and K ingston 2006). Cham bers (2006) identified three separate pathw ays to violent offending which she referred to as 'honour offences' (follow ing a perceived threat to status or safety, characterised by situational violent reactions to perceived threats), 'punishm ent offences' (concerning either revenge or retribution towards a victim perceived to have com m itted a norm violation), and 'denial offences' (involving extrem e em otion occurring under conditions of extrem e stress). It follow s that offenders in each of these three pathw ays will benefit from quite different therapeutic responses, and should begin treatm ent w ith som e understanding about the how the program m e m ight be relevant to their offending, if they are to be considered 'read y' for treatment.

Hostile and instrumental aggression O ne of the most enduring distinctions m ade between different form s of violent behaviour has been betw een hostile and instrum ental form s of aggression (M cEllistrem 2004). W hereas hostile aggression com prises a triggering event (typically in the form of a frustration), an internal 97

Transitions to Better Lives

state of negative em otional arousal, and an im pulse to hurt or harm the provoking agent, instrum ental aggression is intended to secure an environm ental reward and, as such, negative em otional arousal is less likely to be present as an antecedent. Som e robberies clearly illustrate violence that is instrum ental in nature, w hile some though not all hom icides appear to be predom inantly anger-m ediated form s of violence (H ow ells 2008). A w ide variety of term s have been used in the literature to describe w hat is essentially the sam e distinction; these include angry, affective, hostile, reactive, hot-blooded versus non-angry, predatory, proactive, planned, cold-blooded and so on (see M cEllistrem 2004). O ffenders w ho engage in hostile aggression are typically seen as being in need of interventions that develop m ore effective self-regulation of anger and other negative affective states (Cavell and M alcolm 2007), w hile instrum ental aggressors have different and poorly understood treatm ent needs. There are a num ber of problem atic features of the angry versus instrum ental distinction (Bushm an and A nderson 2001; Duggan and Howard 2009). The first is that the distinction is m ost appropriately applied to acts than to actors. A violent individual m ay engage in both form s of violence, though a particular type of offence m ay still predom inate in his or her history. Second, the distinction betw een the tw o form s is often difficult to m ake in practice (Barratt and Slaughter 1998; Bushm an and A nderson 2001). Third, hostile aggression appears to be wrongly confounded w ith impulsivity, and instrum ental aggression with planned violence. M any crim es of violence indicate that angry reactions to a provocation can be carefully rehearsed and nurtured over time until a planned retaliation, delayed revenge, is enacted. W hile som e clinicians m ight readily equate instrum ental violence with 'co ld ' psychopathic characteristics, Patrick's (2006) analysis suggests a more com plex picture of the role of angry em otion in psychopathy (see H ow ells 2009). Current neurobiological and developm ental studies of violence appear to concur and to support the im portance of the an g ry instrum ental distinction. In a recent review of the neurobiology of violence published by the Royal Society (H odgins et al. 2008), Rutter (2008) concludes that violence is heterogeneous and that angry, instrum ental and sadistic form s o f violence should be distinguished. B lair's w ork on brain system s (2004, 2008) also suggests that there are different neuro-anatom ical pathw ays for angry and instrum ental violence. Blair observed different developm ental pathw ays to violence that are broadly consistent w ith these categorisations. Once again, this research suggests that m atching the needs of the offender with 98

Interpersonal violence: the need for individualised services

the type of intervention offered is likely to be im portant - and that if a m ism atch occurs, then levels o f readiness are likely to be low.

Gender differences The angry-instrum ental distinction does not exhaust w ork suggesting the heterogeneity of violent offenders. Work on gender differences suggests that m ales and fem ales differ in the pattern of and antecedents for aggression and violence (A rcher 2000, 2004; Cam pbell 2006; Graves 2006). W omen m ay have a stronger association betw een aggression and internalising conditions such as depression; a greater inhibition of aggression as a result of socialisation; and a stronger association of aggression with physical and sexual victim isation and subsequent post-traum atic stress reactions. Such gender difference may be best explained by improved acquisition of self-regulatory behaviours in fem ales rather than by instigatory differences (see Cam pbell 2006), and suggest that a violent offender program m e for w om en would require a different focus from that adopted in program m es developed for m en. M urdoch, Vess and Ward (in press), in a qualitative investigation of the processes underlying the violence perpetrated by w om en offenders, have suggested that the psychological context in w hich violence occurs can be characterised by substance abuse, poor em otional m anagem ent, poly-victim isation, lack of life goals, poor educational attainm ent, poor social support netw orks, and dysfunctional coping styles. They conclude that w om en offenders who are convicted of violent offences tend to have num erous and long-standing needs that require specific individualised approaches to rehabilitation, w hich should adopt a different focus from interventions that have been designed for male offenders.

Treatm ent readiness in violent offenders The heterogeneity of violent offender groups has m ajor im plications for how treatm ent is delivered. If violent offenders are functionally heterogeneous then it m akes little sense to allocate all violent offenders to the same program m e, unless the program m e were so structured and were sufficiently flexible that session content and therapeutic objectives could be individualised in some way. In practice, however, violent offender treatm ent program m es tend to be highly structured and m anualised, with relatively little opportunity to adapt 99

Transitions to Better Lives

program m e content to individual needs (Polaschek and Collie 2004). It is perhaps unsurprising then that program m e facilitators com m only report encountering hostility, resistance and non-com pliance from their clients - especially in the early stages of intervention. Renw ick and colleagues (1997) have pointed to the therapeutic pessim ism felt both by clients and therapists in institutional settings and to enduring problem s of low m otivation, treatment resistance and avoidance in violent offenders. They note the resentful, distrustful and even com bative style of som e participants in therapeutic groups. In addition, program m e facilitators can experience strong reactions to participants who are frightening or intim idating. In their focus groups with program m e facilitators, Kozar and Day (2009) encountered many instances of situations in w hich facilitators thought that they were going to be attacked, or had direct and indirect threats m ade towards them (for exam ple, one client asked a facilitator if she w as worried that anyone m ight hit her in a group; another facilitator described how she sat with her heart pounding and feeling very disturbed because a client was being hostile). This can lead facilitators to be m anipulated, both overtly and covertly, such that they deviate from treatment plans. K o zar's interview s w ith program m e facilitators further suggest that this m ay lead som e to avoid dealing directly w ith violence, or becom e overly reactive or punitive. M urphy and Baxter (1997) have also pointed to the need to consider the influence of situational and contextual factors on readiness to change in dom estically violent men, including the interpersonal dynam ics of the relationship in w hich the violence occurs, and the im pact o f changes in living arrangem ents for the individuals involved (such as separation, reunification). Existing research on readiness with violent offenders has tended to focus on identifying and responding to individual difference factors. H ow ells and Day (2003), in their analysis of treatm ent readiness in offenders referred to anger m anagem ent program m es, identified seven potential im pedim ents to therapeutic engagem ent. These included: the com plexity of the cases presenting w ith anger problems; the setting in which anger m anagem ent is conducted; existing client inferences about their anger 'problem '; the influence of coerced or m andatory treatm ent; inadequate analysis of the context of personal goals w ithin which the anger problem occurs; and gender and cultural issues. Sim ilar them es have been picked up in the dom estic violence literature. For exam ple, Gilchrist (2009) has identified ten potential im plicit theories held by dom estically violent m en, each of w hich potentially underm ine the extent to w hich perpetrators see their behaviour as problem atic (including: 'w om en are dangerous'; 100

Interpersonal violence: the need for individualised services

'the need for control'; 'w om en are objects'; 'en titlem en t/resp ect'; 'sex drives are uncontrollable'; 'real m an'). O thers have applied the transtheoretical model (TTM ) of change to violent offenders, a model that places em phasis on problem aw areness, self-efficacy and m otivation to change (see Casey et al. 2005). This w ork suggests that m any offenders who are referred to violent offender program m es can be classified as being in the early stages of change, or as low in treatment readiness. Indeed, the task o f engaging such participants often becom es the prim ary goal o f program m e providers.

Responding to low levels of readiness in violent offenders N ovaco, Ram m and Black (2001) have provided accounts of how program m es m ight be modified to m eet the needs of those w ith low levels of readiness, including, for exam ple, by introducing a more extended pre-intervention assessm ent, a preparatory phase to develop necessary skills, and an extended, m ore intensive program m e. Others have suggested that reduction of 'resistance' m ay be undertaken during individual pre-program m e therapy sessions (for exam ple Preston 2000), or at the beginning of a group program m e. For instance, the Canadian Intensive Program m e for Violent O ffenders (Preston and M urphy 1997) begins w ith a 'tw o-w eek m otivational module that addresses client interaction, com m itm ent and trust using techniques such as a cost-benefit analysis of program m e com pletion. This m otivational m odule enables suitable foundations for group w ork to com m ence, with the im plem entation of rules of conduct for the offenders in the group' (Cham bers et al. 2008: 281). A nother alternative is to offer individual treatm ent, utilising m otivational interview ing techniques (these are described in Chapter 10). Day and Carson (2009) have suggested that one of the m ajor therapeutic tasks in violence perpetrator program m es is to reach a shared understanding with the client about the nature, extent and seriousness of their violence. W hile general counselling techniques can be used to develop an affective bond (through the developm ent of rapport and trust), and treatm ent goals and tasks can be set in the process of developing change plans, it is less clear w hether the defining feature of the therapeutic alliance - collaboration - should be a goal in interventions for violent m en, given the tendency of som e to claim that their behaviour w as not serious or that their victim s provoked or deserved the violence. There is a choice that each practitioner m akes betw een confronting, persuading, cajoling and 101

Transitions to Better Lives

m otivating clients into some form of agreem ent w ith the basic goal of non-violence, and beginning work from the offend er's ow n fram e of reference. The form er is associated w ith a didactic, educational approach to group w ork, the latter w ith a more therapeutic approach. Thigpen and colleagues (2007) suggest that the provider should act as a m odel and dem onstrate anti-crim inal expressions of behaviour, and that the effective practitioner m ust be 'consistent' and 'unerring' in com m unicating pro-social and high moral values. At the sam e time, highly discrepant therapist-clien t values are likely to adversely affect treatm ent outcom es by im pairing the developm ent of a strong relational bond. Less direct consideration has been given to the im pact of external or system ic factors on low levels of treatment readiness. Day et al. (2009b) have observed that a m ajor contributing factor to low levels of readiness in dom estically violent m en is the inconsistency and delay in the crim inal justice and referral pathw ays for mandated dom estic violence offenders. For exam ple, court referral and adm inistration by correctional services to intervention program m es can take place after a significant time delay (som etim es years) since the offence. There m ay also be a need, at least in som e violent offender program m es, for the crim inal justice system to reinforce the link betw een the charges and attendance at a behaviour change program m e, and to reinforce and clarify the legally m andated nature of program m e attendance. In addition, Babcock, Green and Robie (2004) in their m eta-analytic review of dom estic violence program m es found that the m ost effective program m es incorporated 'retention techniques' for use w ith those w ho m ay be reluctant to attend com m unity base program m es. These included rem inder phone calls and follow -ups, as well as the use of what they referred to as 'em otion-focused' interventions to improve em otional aw areness and expression, em pathy and com m unication skills.

Conclusion Levels of treatm ent readiness, particularly in relation to problem recognition and m otivation to seek help, are likely to be particularly low in violent offenders w ho are referred to offender rehabilitation program m es. Part of this m ay be attributable to personal beliefs about offending w hich rationalise and justify violence. H owever, low levels of readiness may also be a function of referrals to program m es that are not perceived as being likely to address the particular needs 102

Interpersonal violence: the need for individualised services

of individual offenders. In this chapter we have focused on the m atch betw een the needs of the individual and the content of the program m e as an im portant determ inant of engagem ent in violent offender treatm ent program m es. The balance betw een program m e content that allow s facilitators to respect individual differences while m aintaining acceptable levels of program m e integrity is a delicate one to m aintain, but nonetheless likely to be critical in engaging this group of offenders in any process of behaviour change.

103

Chapter 7

Sex offenders: understanding low readiness

There have been a variety of law s passed since the early 1990s to protect the public from the perceived threat posed by sexual offenders. These law s allow for special sanctions such as indefinite civil com m itm ent, preventive detention, continued detention, extended com m unity supervision, registration and com m unity notification for those convicted o f sexual crim es. There has also been an expansion and refinem ent in the treatm ent approaches taken w ith sex offenders in an effort to reduce the risk of reoffending. Mixed em pirical findings regarding the effectiveness of treatm ent suggest that some but not all sex offenders achieve the desired benefits of treatm ent (H anson et al. 2002; M arques et al. 2005). Despite advances that have been m ade, sex offenders are often difficult to engage in treatm ent, perhaps especially so in response to the involuntary and potentially coercive treatm ent or supervision program m es that have been legally m andated. This chapter exam ines specific factors that may contribute to this difficulty, and offers suggestions to improve treatm ent readiness with this challenging population. The fram ew ork adopted here is the m ultifactor offender readiness m odel (M O RM ) (Ward et al. 2004b). A s presented in earlier chapters, this model asserts that treatm ent readiness is a function of both internal and external factors. The factors internal to the person are cognitive, affective, volitional, behavioural, and identity based. The external or contextual factors depend on circum stances, such as an adversarial legal process and coercive treatm ent; opportunities for treatm ent, such as the availability of specific program m es; resources, such as the availability of adequately trained and qualified staff, 104

Sex offenders: understanding low readiness

cultural sensitivity and a positive therapeutic m ilieu; interpersonal supports for the offender; and program m e characteristics, such as the type of treatment and its timing. This model suggests that an offender will be more ready to change offending behaviour when he or she has these personal characteristics to a greater degree, and w hen he or she occupies an environm ent that can effect and support such change. Rather than being a static quality of the individual or a progression through various stages, readiness for change also fluctuates over time, depending on variations in these internal and external factors.

Internal readiness conditions Cognitive factors

Sex offenders can exhibit a variety o f attitudes, beliefs and thinking styles that interfere with treatm ent readiness. A com m on feature am ong sex offenders is denial of their offending. Studies cited by Yates (2009) have reported com plete or partial denial in betw een 50 per cent and 87 per cent of sex offenders (Barbaree 1991; M aletsky 1991; Sefarbi 1990), and that denial is unrelated to actuariallymeasured risk to reoffend (Sim ourd and M alcolm 1998). Although denial has not been found in the m eta-analyses of H anson and his colleagues to be em pirically associated with higher rates of sexual recidivism (H anson and Bussiere 1998; H anson and M orton-Bourgon 2004, 2005), more recent evidence suggests that denial m ay predict recidivism for some types of offenders. N unes et al. (2007) found that denial was associated with increased sexual recidivism am ong low-risk and incest offenders and decreased recidivism am ong highrisk offenders. Regardless of the relationship w ith reoffence risk, the targeting of denial and m inim isation rem ains an em phasis within treatment program m es for sex offenders (see for exam ple Langton et al. 2008). M any sex offender treatm ent program m es will not admit offenders w ho deny their offending, and denial continues to be considered a serious obstacle to treatm ent participation and progress (ATSA 2005; Levensen and M acgow an 2004; Lund 2000; Schneider and W right 2001; W right and Schneider 2004). In order to understand denial in sex offenders and how we m ight constructively deal with its im pact on treatment readiness, it is necessary to be clear about w hat denial m eans, and perhaps more importantly, w hat function it serves for the individual. Different types

105

Transitions to Better Lives

of cognitive processes have been labelled as denial in the professional literature, ranging from com plete denial of com m itting an offence to not accepting responsibility, m inim ising the seriousness of the offending or the harm done to the victim , and various thinking errors (M arshall et al. 1999; Lund 2000). O ne study by Winn (1996) described seven different overlapping positions that m ay be considered form s of denial: denial of facts, denial of aw areness, denial of impact, denial of responsibility, denial of groom ing oneself and the environm ent, denial of deviant sexual arousal and inappropriate sexualisation of problem s, and denial of denial. Levenson and M acgow an (2004) subsequently noted that denial m ay som etim es refer to a com plete refutation of the facts related to an offence, but m ay also represent a continuum that includes m inim ising the im pact of the offending on victim s, and refusing to acknow ledge the severity or repetitive nature of their offending behaviour. Beyond w hat denial is, it is useful to consider w hat denial does. Denial serves a protective function, described as a form of selfpreservation (Winn 1996) or a self-protective strategy (Kear-Colw ell and Pollock 1997). Yates (2009) suggests that denial represents a norm al cognitive process that all people use to m aintain self-esteem , and that sex offenders m ay be particularly invested in som e form s of denial in light of the personal, social and econom ic costs to individuals w ho adm it to sexual offending. O ne way to consider denial in sex offenders is by w ay of the concept of a defence m echanism (Levenson and M acgow an 2004). Denial is defined in the D SM -IV-TR (A m erican Psychiatric A ssociation 1994) as a defence m echanism in w hich 'the individual deals w ith em otional conflict or internal or external stressors by refusing to acknow ledge som e painful aspect of external reality or subjective experience that would be apparent to others' (1994: 811). Defence m echanism s are considered autom atic psychological processes that protect the individual against anxiety and from the aw areness of internal or external dangers or stressors, often outside of the individual's conscious awareness. This introduces the question of w hether denial is a conscious m isrepresentation of som ething the individual knows to be otherw ise or a distortion of reality of which the individual is unaware, or incom pletely aware. Yates (2009) notes that the view of cognitive distortions as constituting denial arises from the assum ption that these are nearly alw ays deliberate and conscious distortions, as suggested earlier by M arshall et al. (1999). However, Yates provides various exam ples from the research literature of the m isperceptions and m isinterpretations that sex offenders are prone to m ake, such that 106

Sex offenders: understanding low readiness

they may not alw ays be intentionally distorting or denying the truth. The im plication is that denial and m inim isation by sex offenders is not alw ays deliberate. Regardless of w hether denial and related cognitive distortions are deliberate or largely unconscious, som e studies have attempted to em pirically address the assum ption that there is a significant relationship betw een denial, treatm ent engagem ent and treatm ent progress for sex offenders. Levenson and M acgow an (2004) found that treatm ent progress was correlated w ith higher levels of treatm ent engagem ent and lower levels of denial. They use the ATSA definition of denial as the failure of sexual abusers to accept responsibility for their offences, and conclude that their findings support the standards of practice that m aintain that adm itting to a sex crim e is a necessary condition for successful treatment. However, these authors note that it is unclear w hether denial causes an inability or unw illingness to engage in treatm ent, or w hether treatm ent creates a w illingness to let go of denial. This appears to be a crucial distinction. If engagem ent decreases denial, and decreased denial is im portant for treatment progress, then readiness to engage in treatm ent becom es an im portant goal, prior to or even instead of overcom ing denial. This becom es particularly salient in situations where denial of sexual offending is a criteria used to exclude sex offenders from treatment. Yates (2009) points out that although denial and engagem ent in com bination accounted for a significant proportion of the variance in treatm ent progess in the Levenson and M acgow an (2004) study, engagem ent was a stronger predictor than denial. She also notes that treatm ent progress was measured through therapist ratings that included adm ission of offending, presenting a potential m ethodological confound. Is overcom ing denial necessary for treatment progress and reduction of risk for sexual reoffending? Earlier positions on this issue considered the confrontation and elim ination of denial as an explicit requirem ent of effective treatm ent (for exam ple, Salter 1988). M ore recent researchers continue to support this prem ise (Levenson and M acgow an 2004). N oting that offender accountability is considered a central goal by alm ost all cognitive behavioural treat­ ment program m es, W right and Schneider (2004) have em pirically investigated treatm ent progress as a function of denial and offender explanations of accountability using the Facets of Sexual O ffender Denial (FoSOD), a self-report m easure designed to assess various aspects of denial throughout the treatm ent process w ith child victim sex offenders. These authors describe three processes of 107

Transitions to Better Lives

denial: refutation, m inim isation and depersonalisation. Refutation is a form of absolute, full, or categorical denial of sexual offending. M inim isation represents partial acknow ledgem ent that som ething about his or her behaviour is problem atic or potentially harmful. Depersonalisation, considered the m ost entrenched form of denial, is the rejection of the possibility that the offender is the kind of person w ho could com m it a sexual offence. W hile an offender engaged in depersonalisation may acknow ledge his (or her) involvem ent in a sexual offence and accepts responsibility for these behaviours, there rem ains an inability to recognise predispositions likely to influence future behaviour, and therefore denial of risk for future offending. W right and Schneider found that their measure showed significant reductions in denial only for those w ho progressed in treatment, but also that reductions in denial continued through more advanced levels of treatment. Conversely, denial scores changed little for those w ho did not advance in treatment. W right and Schneider consider their findings consistent with the view that denial consists of cognitive processes that function to m aintain distorted thinking and to excuse the offenders' responsibility for their behaviour throughout treatm ent. Their results are presented as suggesting that denial is a dynam ic factor inversely related to the treatm ent goal of accepting responsibility, and the reduction o f denial is therefore central to treatm ent progress. O thers have taken the position that focusing on denial is not so essential, at least not initially, and certainly not through aggressive confrontation. Recognising that denial is a form of self-preservation allow s the clinician to address the offender in w ays that do not rely on direct confrontation and the activation of defences that often result (Cooper 2005; Kear-Colw ell and Pollock 1997; W inn 1996). M arshall and colleagues (M arshall 1994; M arshall et al. 2006; M ann and M arshall 2009) have observed that effective therapeutic approaches reduce denial, and that denial may be dealt w ith clinically in different ways. They suggest that it is im portant to distinguish betw een offenders' cognitions that are related to risk (and therefore im portant to focus on in treatm ent), and those that are not. Post-hoc rationalisations and m inim isations that occur after the offence are not necessarily causally related to the offending process. As denial and m inim isation are not generally associated w ith risk o f sexual recidivism , their reduction or elim ination should not be a prim ary goal of treatm ent. Vigorously confronting denial and m inim isation, especially early in treatm ent w ithout an adequate therapeutic alliance, is likely to decrease rather than increase readiness. But according to M ann and M arshall 108

Sex offenders: understanding low readiness

(2009) there are cognitive phenom ena that are im portant for risk, and therefore are relevant when it com es to d iscussing the issue of treatment effectiveness. These occur prior to and during the offence, and are likely to reflect underlying schem as. Schemas and readiness

The various form s of denial that have been described in the professional literature appear sim ilar to the broader category of cognitive distortions that have been identified with sex offenders (Abel et al. 1984). Cognitive distortions have m ore recently been conceptualised as im plicit theories (Ward 2000; Ward and Keenan 1999). All of these cognitive phenom ena m ight be productively considered w ithin the concept of schem a (Baker and Beech 2004; M ann and Beech 2003; M ann and Shingler 2006; Thakker et al. 2007; Yates 2009). Schem as are defined as cognitive structures that serve to influence and direct the processing of incom ing stim uli by sim plifying and classifying the inform ation based on previous experiences (Beck 1964, 1967). Thus they originate from early life experiences, and are organised around various them es for understanding the self, others and the world. The content of schem as m ay or m ay not be available to conscious aw areness, and can include rules, attitudes, self-verbalisations, beliefs and assum ptions. From this perspective, denial and other cognitive distortions are seen as products of underlying schem as. Schem as are theorised to be stable and enduring, and they are activated by the presence of situational cues, particularly am biguous or threatening cues, as perceived through the filter of the schema. The treatm ent process can be seen as presenting a variety of am biguous or threatening cues that will activate the existing schem as of the offender, especially treatm ent that is vigorously confrontational or insensitive to an offend er's defences. Consideration of schem as may therefore offer a productive w ay to approach denial and other cognitive factors that im pede treatm ent readiness. M ann and Shingler (2006) recom m end the use o f 'life m aps' or 'life histories' to identify recurring thinking patterns and the previous experiences that have shaped these patterns. Recognising the w ork of Young (1990) on early m aladaptive schem as, the goal of intervention is not to change schem as, as this is unlikely to be successful in the tim e-lim ited and structured group program m es that constitute most sex offender treatment. This collaborative and supportive approach attem pts to allow the offender to understand and recognise the underlying schem as that influence his or her interpretation of and response to

109

Transitions to Better Lives

events, by explaining the concept o f schem as and helping offenders recognise their ow n schem as in action. This approach is likely to m ore effectively support the developm ent of a positive therapeutic alliance, thereby im proving treatm ent readiness. Finally, a cognitive factor that may strongly influence sex offenders' treatm ent readiness has been identified in the general treatment outcom e research as client expectations (A rnkoff et al. 2002; Garfield 1994; Kirsch 1990). Ward et al. (2004b) note that expectancies can com e from previous experiences of treatm ent, the experience of the assessm ent process, the reputation o f the treatm ent program m e and perceptions about treatm ent staff. To the degree that these expectancies are based on positive prior experiences, treatment readiness m ight be im proved, or at least not further com prom ised. H owever, it will be im portant to consider the individual offend er's perceptions of previous assessm ent and treatment experiences. A dversarial legal process and coercive treatment conditions may contribute to the form ation of negative and defensive expectancies by offenders. Concerns about confidentiality and the uses o f inform ation disclosed in treatm ent can becom e significant barriers to treatm ent readiness. These contextual factors, considered in more detail later in this chapter, m ay lead to negative expectations and related fears that w hat the offender says m ay be subsequently used against him, such as evaluations for extended supervision or involuntary civil com m itm ent follow ing his prison term. Affective factors

Affect is typically considered a broad term that includes em otions, m oods and feelings (Berkow itz 1999; Pow er and D algeish 1999 from H ow ells and D ay 2006). In w ays sim ilar to H ow ells and D ay in their approach to affective determ inants of treatm ent engagem ent with violent offenders, the affective reactions of sex offenders are likely to have a m ajor influence on their treatm ent readiness. Strong em otions have been found to play a role in the offence pathw ays for at least som e sex offenders (H ow ells et al. 2004b; Polaschek et al. 2001; Proulx et al. 1996; Ward et al. 2006), so that access to and experience of these em otions may be essential to the eventual m odification of these pathw ays. A relationship betw een schem as, as discussed earlier, and affect has been noted by Serran et al. (2003), w ho observed that offence-related cognitive schem as are typically only activated by particular em otional states. Yet som e offenders have trouble accessing and indentifying various affective states. An extrem e exam ple of this

Sex offenders: understanding low readiness

may be the profound affective deficits exhibited by psychopaths, which m ay also relate to difficulties engaging in or benefiting from treatment. Readiness for treatm ent in sex offenders m ay be related to affect in several ways. One hypothesis is that sex offenders (and treatment clients in general) are often m otivated to engage in treatment to reduce distress, m eaning that som e level of negative affect is necessary initially to prom pt the desire for treatment. Beutler, Clarkin and Bongar (2000), for exam ple, have argued that treatm ents for depression achieve their best outcom es am ong those w ith m oderate to high initial levels of distress. Yet if there is no appreciable a priori distress, as m ay be the case w hen a sex offend er's denial or other defence m echanism s are w orking effectively, there is little to stim ulate treatment readiness. In fact, the prospect of engaging in treatm ent, w ith the attendant requirem ent to adm it and disclose distressing offence-related inform ation, m ay be far more distressing than the offend er's usual affective state. Som e sex offenders may lack the ability, w hich could be viewed as a skill, to access and identify different em otional states. Such individuals m ay never have had m uch practice at this process, and have grow n up and alw ays lived in fam ily or cultural environm ents where expressing feelings (other than anger, perhaps) was not modelled or rewarded. These individuals m ay actually require help to acquire the skills of affect recognition and expression. In our view, these types of skills are likely to be m ost easily acquired w ithin the context of a safe and supportive therapeutic environm ent, as addressed below. For other sex offenders, offending may have provided a m echanism for em otional regulation (see Day 2009; H ow ells et al. 2004b; Ward and H udson 1998). Proulx, M cKibben and Lusignan (1996) found that negative m oods and conflicts were associated w ith 'overw helm ing' deviant fantasies and increased m asturbatory activity for both rapists and child molesters. N egative mood and interpersonal conflict may be a com m on experience for offenders in m any sex offender treatm ent program m es. If this occurs early in the treatment process, it might result in an increase in the very cognitions and behaviours that the program m e is intended to reduce. Howells, D ay and W right (2004b) note that reciprocal determ inism betw een behaviour and the environm ent helps us to understand how some individuals, including sex offenders, are extraordinarily skilled at creating negative and punishing environm ents for them selves, which

Transitions to Better Lives

in turn produces negative affective experiences. A nyone w ho has worked in a correctional treatm ent program m e for sex offenders will be fam iliar with those offenders w ho, under the rules and procedures of that structured and scrutinised environm ent, consistently create a conflictual dynam ic and the associated negative affect both for them selves and others. H ow ells et al. suggest that life skills training to assist offenders to develop the skills that m ake life and relationships m ore rew arding may have an im portant role to play, along with a specific focus on improved recognition and regulation of em otions. A ffective skills training and acceptance-based approaches such as m indfulness training to allow accurate experiencing and labelling em otional states should result in improved engagem ent in therapy (Day 2009; H ow ells and Day 2006). This could be particularly useful in the early stages of treatment to facilitate a secure therapeutic alliance, prior to more direct and forceful confrontation of behaviour and defences that m ay rupture the alliance and decrease treatment readiness. A particularly salient affective factor for treatm ent readiness with sex offenders is likely to be the im pact of guilt and shame in response to acknow ledgem ent of their offending. Proeve and H ow ells (2002, 2006) note that evaluation of the self distinguishes guilt from shame. Individuals experiencing guilt tend to focus more on specific behaviours, but do not tend to m ake the global negative evaluations of self associated with shame. Sham e involves the judgem ent of self as inferior or bad and can also involve the perception of being negatively judged by others as defective in some respect. Sexual offending m ay be considered a 'quintessential sham e-eliciting form of behaviou r' (Proeve and H ow ells 2006: 128). Sham e and guilt may have different im pacts on treatment readiness. The action tendencies associated with guilt are thought to include confession, apologising, and attem pting to repair the dam age inflicted by guilt-inducing behaviours (Barrett 1995; Frijda 1986). These actions may be conducive to increased levels of self­ disclosure and to treatm ent engagem ent. Sham e, on the other hand, is associated w ith action tendencies to hide the self from others, and therefore is associated w ith a w eaker therapeutic alliance and lower levels of treatment readiness. Effective treatm ent with sex offenders m ay require a m ovem ent from shame towards guilt, aw ay from a focus of 'w ho I am ' tow ards a focus on 'w hat I have done'. Confrontation in therapy, particularly w ithout adequate preparation regarding the role of confrontation in the therapeutic process, is likely to be experienced as an attack on the self, and actually increase levels

Sex offenders: understanding low readiness

of shame. Educating offenders about the differences betw een sham e and guilt, and explicitly focusing on the behaviours of offending rather than the self of the offender, may help to m inim ise defensive resistance and facilitate treatm ent engagem ent. The relationship betw een sham e, cognitive distortions, em pathy deficits and self-esteem has been recently addressed by M arshall and his colleagues (M arshall et al. 2009). C iting evidence that many sex offenders suffer from low self-esteem , these authors note that such individuals appear reluctant to attem pt novel behaviours, an inherent elem ent of treatm ent. Denial and cognitive distortions in such cases function to protect an offend er's already low self-im age. This is to be expected, and is an understandable response to the offend er's situation and experiences in the judicial and correctional system s, rather than necessarily indicating a defect in character. Em pathy is positively associated with guilt and negatively associated w ith shame. Integrating these concepts, M arshall et al. propose that sham e, cognitive distortions, and lack of victim em pathy serve to avoid further erosion of self-esteem . Such offenders will typically be difficult to engage in treatment (that is, will have low treatment readiness) unless efforts are m ade to overcom e these interrelated problem s. In an earlier w ork, M arshall and colleagues (1997) recom m end procedures for enhancing self-esteem that have dem onstrated effectiveness with sex offenders and caution against vigorously challenging or aggressively confronting offenders too early in treatm ent. Early stages of treatm ent should focus on engagem ent and m otivation, with goals of enhancing the offend er's self-esteem , reducing his sham e, and teaching him basic coping behaviours. These authors further note that treating the offender respectfully and dem onstrating em pathy and w arm th have also been shown to facilitate treatm ent-induced changes with sex offenders, features of the therapist that are addressed in more detail below under external readiness factors (for discussion of this issue see Chapter 3). They also caution that not all offenders will have low self-esteem or struggle w ith guilt or sham e, and therefore it is im portant to take into account offenders' unique personality factors. Personal identity factors

A nother set of internal factors that w ill have a strong ongoing effect on treatment readiness relate to the offend er's personal identity. Sex offenders w ith features of personality disorders and psychopathy m ay present som e of the m ost overt challenges to treatm ent readiness and the developm ent of an effective therapeutic alliance. A lthough findings

113

Transitions to Better Lives

have been m ixed, some studies have reported that alm ost 90 per cent of their samples exhibit features of at least one personality disorder (Berger et al. 1999). In a recent study, Craissati, Webb and Keen (2008) reported that in their sam ple o f 103 sex offenders w ho completed the M CM I-III, 73 per cent presented with personality dysfunction and 37 per cent reported personality dysfunction sufficient to w arrant possible personality disorder diagnoses. Psychopathy has also been reported in significant proportions of sex offenders, although in higher percentages of rapists than child m olesters and incest offenders (Forth and Kroner 1995, cited in Craissati et al. 2008). Furtherm ore, personality disorders have been associated with treatm ent attrition for sex offenders (Abel et al. 1988; Craissati and Beech 2001), and psychopathy in particular poses a variety of challenges to effective treatment. Problem s with treatment readiness in high-risk offenders with personality disorders have been specifically identified (H owells and Day 2007). In the m ultifactor offender readiness model (M ORM ), Ward et al. (2004b) have spoken of personal identity in term s of the pursuit of prim ary hum an goods in w ays that will shape an individual's sense of mastery, self-esteem , perception of autonom y and control (see Chapters 3 and 10). The perspective taken here is concerned w ith personal identity in term s of the offend er's sense of self, which relates back to the concept of schem as. Much has been w ritten in the general clinical and psychotherapy literature about self-schem as and their role in the treatm ent process, especially in the field of personality disorders. W hile som e sex offenders may present characteristics that are consistent w ith various aspects of personality disorder, the consideration of self-schem as m ay be more broadly useful in understanding im portant features of treatm ent readiness and resistance to change. Livesley (2003) has delineated a sequence of phases for the treatment process in m anaging personality disorders that com prises safety, containm ent, control and regulation, exploration, and finally integration and synthesis. H is description of the process of change em phasises the role of the therapeutic relationship in leading to increased self-know ledge, problem recognition, and the acquisition of alternative behaviours. A key to facilitating this process is an appreciation of the individual's self-states. D raw ing from H orowitz (1998), self-states are defined as states of mind consisting o f conscious and unconscious experiences with associated patterns of behaviour, which can last for short or long periods of time. Self-states are im portant for understanding the structure of self-experience and the flow of interpersonal behaviour.

114

Sex offenders: understanding low readiness

They are structured and initiated by underlying core schem as, which are activated by different (usually interpersonal) events or mood changes. Livesley (2003) states that m ost people experience several self-states, w hich som etim es conflict. O nce activated, the schem a can arouse strong em otions and other schem as that help to m aintain the self-state. Resulting behaviours can often evoke reactions from others that are perceived as confirm ing and reinforcing the underlying schem a, which m akes for a repetitive cycle that is difficult to modify. Livesley also argues that the m ost general schem a is the person's overall conception of the self, som etim es referred to as the theory of the self, and that m ore abstract schem ata are difficult to change. A s noted earlier, the relationship betw een self-concept and underlying schem as has been taken up by Young (1990, 1999), w ho identifies a set of d istinct early m aladaptive schem as that result in recurrent dysfunctional patterns o f behaviour that are selfperpetuating and resistant to change. These early m aladaptive schem as are organised into five broad dom ains, w hich are hypothesised to correspond to developm ental needs of childhood that were not met. These are disconnection and rejection, im paired autonom y and perform ance, im paired lim its, other-directedness, and overvigilance and inhibition. Space constraints do not allow for a detailed analysis and application of these m aladaptive schem as in this chapter, but it appears that several may be relevant to the sense of self often experienced by sex offenders. These include disconnection and rejection expectations regarding abandonm ent, m istrust, em otional deprivation, defectiveness, sham e, and alienation, w hich may be particularly relevant am ong child m olesters. Deficiencies related to impaired lim its, including entitlem ent, grandiosity and insufficient self-control or self-discipline, may be particularly relevant for rapists. A consistent observation by those w ho address the nature of selfconcept or identity from a schem a perspective is that self-schem as are difficult to change (Beck et al. 2004; Livesley 2003; Young 1999). Leahy (2001) considers resistance in cognitive therapy to be based in part on the need for self-consistency or self-verification. According to this principle, people are driven to achieve balance and consistency in their beliefs and roles, to m aintain control. Part of this process involves self-verification, w hereby individuals seek out and attend to inform ation that verifies their self-concept, regardless of w hether that self-concept is positive or negative (Sw ann et al. 1992).

115

Transitions to Better Lives

A nother source of resistance that will reduce treatm ent readiness is schem a avoidance. Young (1999) notes that w hen early m aladaptive schem as are triggered, the associated affective experiences can be intense and unpleasant. Therefore, individuals develop several types of both volitional and autom atic processes for avoiding the activation of these schem as, including cognitive, affective and behavioural avoidance techniques. Livesley (2003) states that m odifying schem as and associated m aladaptive patterns o f behaviour is an im portant part of the change process in treatm ent, and requires identification and recognition of these patterns. However, schem a avoidance may hinder recognition and acceptance, and often occurs w hen a schem a is identified before the individual is ready, or w hen the therapeutic alliance is in a com prom ised state. More recently, Livesley (2007) has advocated for the relevance of an integrated approach to the treatm ent of personality disordered offenders. The defining features o f personality disorder that he or she presents appear sim ilar to som e of the features that describe many sex offenders: failure to achieve stable and integrated representations of self and others; failure in the capacity for intim acy; and failure in adaptive functioning in the social group as indicated by failure to develop the capacity for pro-social behaviour and cooperative relationships. Livesley argues that interventions need to be directed towards personality pathology per se, because it directly affects com pliance and the ability to respond to interventions. This position suggests that it will not be sufficient to apply interventions found to be effective w ith non-personality disordered offenders while neglecting the personality disorder com ponent. Livesley recom m ends an integrated and m ultifaceted approach tailored to the needs of individual offenders, with an em phasis on facilitating the developm ent of more integrated and coherent personality functioning. C iting the work of Piper and Joyce (2001), he notes that the best results occur w hen treatm ent is individually tailored and the patient and therapist agree on a specific treatm ent contract. The com m on elem ents of treatm ent for personality disordered offenders, and by extension all offenders, are based on the therapeutic relationship, w ith attention given to m aintaining a stable therapeutic process. Behavioural factors

In the m ultifactor offender readiness m odel of Ward et al. (2004b), three types of behavioural factors that influence treatm ent readiness are distinguished, and are proposed as occurring in a tem poral

Sex offenders: understanding low readiness

sequence. First, current behaviour m ust be identified as a problem. Then, help m ust be sought for the problem atic behaviour. Finally, offenders m ust have com petencies to participate in the treatment process. We have already noted that denial of offending behaviour, or denial that it is a problem , m ay be com m on am ong sex offenders. We would argue that this does not preclude the engagem ent of sex offenders in a therapeutic process that eventually leads to behaviour change in the area of sexual offending. From the perspective of the Good Lives M odel, the goals of treatm ent are the securing of prim ary goods in more pro-social and less harmful w ays than the offender has behaved in the past (see Chapter 4). Ward et al. (2004b) have argued that to be ready for treatm ent, the individual m ust recognise a need to change: that is, that their offending creates problem s for them selves and others. If this is defined more broadly, at least initially, than an exclusive focus on sexual offending, it is more likely that a need for change in the offend er's life will be recognised and accepted. This im plies that an offender m ay be ready to work on a particular problem , but not the one that the therapist view s as relevant and central to offending. This approach calls for there to be a collaboration betw een the offender and the service provider to identify treatm ent aim s that w ill help the offender m eet broader personal goals. To do this, both the offender and the therapist have to be ready for a treatm ent process that identifies and w orks towards the m eaningful life goals of the offender rather than exclusively avoidance goals to reduce risk. This issue is taken up further in the external readiness factors section later in this chapter. Volitional factors

Volition refers to an intention to pursue certain goals, and is closely associated w ith w hat is often termed m otivation in the offender treatment literature. Yet despite its long history in the field of offender rehabilitation, the construct of m otivation has been criticised as am biguous (Drieschner et al. 2004) or even unnecessary (Draycott 2007b). Ward et al. (2004b) conceptualise volitional factors, including m otivation, to involve the form ation of an intent to pursue certain goals and the developm ent and im plem entation of a plan to achieve these goals. The goals m ust therefore be relevant and valued by the offender if he or she is to be m otivated to pursue them, including the goals of treatment. O ther aspects o f volition include the belief that one is capable of exercising choice and can directly control im portant personal outcom es. In the context of offender rehabilitation, this

Transitions to Better Lives

inclu des the ability to con sent to treatm ent, to m ake d ecisions based on the costs and benefits to oneself, and to m ake d ecisions w ithout coercion. M otivation for treatm ent is not b est conceptu alised as a static trait of the ind ividu al, but rather as a fluctuating state dep endent on d ynam ic factors (D rieschn er et al. 2004) and the relevance o f the goals to the individual at a given point in tim e. B arrett, W ilson and L ong (2003) have specifically addressed m otivation to change in sex offend ers, and found that m otiv ational m easures varied over the cou rse o f treatm ent. They also noted that m otivation is dynam ic and m ay be im pacted by internal, extern al and alliance issues. Their study highlighted the influence o f environm en tal d eterm inants on m otivation. M otiv ation and oth er asp ects of treatm ent read iness m ay therefore be as m u ch a function o f external factors as internal factors.

External readiness conditions Circumstance factors

O ne of the d efining features o f sex offend er rehabilitation in the crim inal ju stice con text is the degree of coercion that com pels offend ers into treatm ent (Vess 2009). Treatm ent is often m andated , and relatively few offend ers seek help w ith their sexual offending and its related problem s voluntarily, particularly in correctional settings. H ow ever, it has been argued that the experience o f coercion is not entirely a function o f w hether treatm ent is m andated; furtherm ore, Day, Tucker and H ow ells (2004) have distinguished coerced treatm ent from pressured treatm ent. R ehabilitation activities are pressured in the sense that the d ecision o f w hether to undertake a program m e is influenced by negative con sequ ences for n on ­ participation. This is a com m on elem ent of treatm ent program m es that w ill be seen as favourable, o r even necessary, by parole boards w hen m aking d ecisions about release from prison, for exam ple, although p articipation in the program m es w ill n o t be m andated in a form al sense. In a recent m eta-analysis o f offend er coercion in treatm ent, Parhar and colleagues exam ined 129 stud ies of various offend er rehabilitation program m es (Parhar et al. 2008). To avoid the false d ichotom y o f m andated v ersu s v olu ntary treatm ent, these researchers created five levels, ran gin g from m andated involu ntary to non-m andated v o lu ntary program m es across both custod y and

Sex offenders: understanding low readiness

com m unity settings. The strongest effect sizes were found for voluntary treatm ent w ithout coercion, and the w eakest effect sizes were found for mandated treatment. Furtherm ore, treatm ent setting had a significant effect, w hereby mandated treatment had no effect on recidivism when the program m e w as delivered in a custodial setting. The authors suggest that custodial settings m ay increase the perception of forced treatm ent, w ith negative consequences for treatment responsiveness. D ay et al. (2004) point out that coercion and pressure are not sim ple objective facts, but rather have more to do with the offend er's subjective experience regarding freedom of choice. H owever, a significant problem is that a num ber of sex offenders appear unlikely to choose treatment voluntarily. Grubin and Thornton (1994), for exam ple, found that 41 per cent of sex offenders in the UK said that they would only participate in treatm ent in order to gain parole. Consistent with these findings, Burdon and Gallagher (2002) concluded that coercion rem ains the only effective m eans of protecting society from sex offenders. These authors also noted that while coercion increases treatm ent retention, it does not necessarily im prove treatm ent outcom e, although they cite trends in the research findings that suggest treatm ent can transform the perception of coercion and lead to better outcom es. If m ost sex offenders com e to treatm ent because it is mandated or coerced, w hich can have negative effects on treatm ent com pliance and ultim ate outcom es, an im portant question arises: w hat can be done to m odify the subjective experience of coercion and im prove treatm ent readiness? Coercion is experienced w hen treatment is not congruent w ith the individual's personal goals and there is a subsequent perceived m isalignm ent betw een program m e goals and desired outcom es. One challenge is therefore to help the offender to identify w ith and invest in (that is internalise) the goals of the treatm ent program m e. M eeting this challenge will depend largely on several interrelated external conditions, including resource factors, location factors, opportunity factors and program m e factors. Resource factors

Engaging coerced offenders in treatm ent is a task that requires high levels of therapeutic skill (Day et al. 2004). Thus one of the crucial resources for enhancing treatm ent readiness w ith sex offenders is the availability of adequately trained and effective staff. Sex offenders have reported that they assessed the quality of the treatm ent

Transitions to Better Lives

program m e based on their perception of the therapists' com petence, and that they relied on their perception of the therapists to determ ine w hether or not to get involved in therapy (D rapeau 2005). It is only relatively recently that greater attention has been paid to the skills and qualities necessary to be an effective sex offender therapist (Beech and H am ilton-G iachritsis 2005; H arkins and Beech 2007). M arshall and colleagues have written extensively about the therapeutic process w ith sex offenders (M arshall 2005; M arshall et al. 2005; Serran et al. 2003). They recom m end that a non-confrontational approach, where confrontation is defined as a harsh challenging of clients that seem s likely be perceived by the client as denigrating, is adopted by clinicians. W hile distorted and dysfunctional perceptions, beliefs, attitudes and schem as of sex offenders need to be challenged in treatm ent, the critical issue is how to do this effectively w ithout dim inishing treatm ent readiness. Research has revealed that therapist qualities of w arm th, empathy, respect, rew arding, and directive are all related positively to indices of beneficial change in treatm ent (M arshall 2005). W hile unqualified support of sex offenders in treatment is not called for, directive and supportive therapists can help clients recognise problem s and consider the benefits of behaviour change. Referring back to the internal readiness factors discussed earlier, rather than seeing offenders as resistant to treatm ent, therapists should conceptualise m inim isation, denial and avoidance as a desire to avoid sham e and to protect self-esteem . Pushing against these defences is unlikely to enhance treatment readiness and more likely to im pede treatm ent progress. Rather, resistance should be seen as a signal to sw itch therapeutic strategies rather than as a reason to attem pt to coerce the client into accepting the therapist's position (Serran et al. 2003). M uch of the currently available therapy for sex offenders occurs w ithin a group treatment form at, and therefore sex offender therapists need to be properly trained and qualified in group process skills. The im portance of group leadership in producing cohesiveness, appropriate group norm s, and the instillation of hope have long been recognised for running effective groups and producing therapeutic change (Yalom 1995; Yalom and Lieberm an 1971). H ighly cohesive groups are associated w ith positive outcom es, and qualities such as leader support have a clear effect on cohesion, expressiveness and other positive group process indicators in sex offender treatment groups, as reported by Beech and H am ilton-G iachritsis (2005). These researchers suggest that group therapy skills are particularly im portant for engaging sex offenders in treatm ent, given the coercive 120

Sex offenders: understanding low readiness

circum stances of m any program m es. U nder these conditions, it is particularly im portant for program m es to select and train their treatment staff well. Programme factors

O ther therapist qualities associated w ith treatm ent engagem ent and a positive treatm ent outcom e include hum our and flexibility. H owever, M arshall and colleagues (M arshall et al. 2005) have noted that flexibility is often constrained by the overly detailed treatm ent m anuals of m any sex offender treatm ent program m es. Flexibility refers to the need for therapists to adapt their strategies to each client's particular characteristics, throughout the treatment process (M arshall 2005). M arshall et al. (2005) consider strict adherence to detailed m anuals a m istake, as they discourage the flexibility and sensitivity required by the therapist to prom ote effective engagem ent in therapy. Flexibility extends to the identification of treatment goals. It has already been noted in the section on internal factors above that treatm ent readiness w ill be influenced by the degree to w hich treatm ent appears to encom pass goals that are personally relevant and valued by the individual. If the treatm ent program m e, and by extension the therapists, insist that the offender pursue goals that are defined by the program m e as im portant rather than those identified by the offender, poor treatm ent readiness is likely to result. This is especially salient in mandated treatm ent settings where subjective perception of coercion is likely to be inherently high. In one au th or's experience (JV) in California's Sexually Violent Predator treatm ent program m e, m any of the offenders com m itted for treatment considered them selves m ore as political prisoners than legitim ate treatment clients, and efforts to engage them in treatm ent to pursue risk reduction goals defined by the program m e were fraught w ith difficulties. Som etim es this low state of treatm ent readiness could be gradually overcom e through the skilful therapeutic w ork of highly trained and experienced staff, but there was often an enduring lack of perceived alignm ent betw een the goals of the program m e and the goals of the offender. M arshall et al. (2005) sum m arise five general concerns about current treatment approaches with sex offenders as follows: (a) there is an excessive em phasis on negative issues in both targets of treatm ent and language used by treatm ent providers,

121

Transitions to Better Lives

(b) there is a failure to explicitly encourage optim ism in clients and encourage their belief in their capacity to change, (c) there is a general absence of an explicit attem pt to w ork collaboratively with clients, (d) the role and influence of the therapist has been all but neglected, and (e) there have been few attem pts to provide clients w ith goals that w ill result in them leading a more fulfilling and prosocial life. (2005: 107) It is argued that as more of these concerns are addressed in treatment program m e design and function, treatm ent readiness and treatment effectiveness will improve. M ovem ent aw ay from earlier positions that advocated that the therapist m ust im pose the goals of treatment (see Salter 1988) towards more collaboration - a practice in which the therapist w orks w ith the client to define together the nature of the client's problem s and to agree on a process for w orking together - has been recom m ended as an approach to apply from the initial contact with the client (M arshall et al. 2005). A ttention to these concerns may also help to reduce the attrition rate from m any program m es, reported as high as 30 -5 0 per cent, as predictors of treatm ent attrition are seen as indicators for program m e im provem ent rather than shortcom ings in the offenders (Beyko and W ong 2005). Opportunity, location and support factors

There are three brief points to m ake regarding these external readiness factors. One is the need for the availability of treatm ent in a variety of settings in order to fully address the needs of sex offenders. Although the risk principle of the risk -need -resp onsivity m odel suggests that treatm ent resources should be applied to higher-risk offenders, this can result in an im balance and potential m isdirecting of scarce resources. A s an exam ple, the State of California releases over 700 sex offenders a m onth from its prison system , but less than 1 per cent are com m itted as 'sexually violent predators' (Vess et al. 2004; Vess 2005). The cost of treating this small m inority of offenders in a high-security facility exceeds $100,000 per offender each year, over the course of an indefinite length of treatm ent from w hich few offenders have been recom m ended for release to the community. In light of findings cited earlier in this chapter that m andated treatm ent in custodial settings has dem onstrated negligible effects on recidivism , we m ight question w hether this balance in the distribution of treatm ent opportunity resources is optim al.

122

Sex offenders: understanding low readiness

A related issue under opportunity factors is the lim ited opportunity for offenders to practise and dem onstrate im provem ent in the dynam ic risk factors know n to be associated w ith risk of sexual reoffending w hile they are incarcerated in a custodial setting. It is often difficult or im possible to create the conditions that offenders will face in their lives outside the institution, and it is difficult to extrapolate current risk assessm ent m easures from the custodial setting to the anticipated functioning of the individual once released to the community. Treatment and assessm ent and reduction of risk can therefore struggle to reach any sem blance of ecological validity. This m ay be a lim iting factor in the effectiveness of custodial treatm ent opportunities, and m itigate against the trend towards longer prison sentences and post-sentence detention that has developed in recent years (Vess 2008, 2009). The third issue under this topic is related to support, which is often influenced by opportunity and location. The readiness of an offender to enter and stay in treatm ent is likely to be a function of the degree of support available to him or her. Yet m any offenders are im prisoned in facilities far from fam ily and friendship netw orks, particularly in large correctional system s that routinely m ove offenders across different facilities. O thers w ill have been rejected by fam ily and friends in response to their sexual offending or continued criminality. In these circum stances, the presence o f supportive clinicians, prison officers and custodial or com m unity corrections staff w ho support and facilitate engagem ent in treatm ent will be particularly important. Yet not all custodial and supervision staff, or even all clinical staff, will have positive and supportive attitudes towards sex offenders. This lack of consistent and overt support is likely to dim inish treatm ent readiness.

Summary and conclusion This chapter has reviewed the internal and external treatment readiness factors as they apply to sex offenders. It is proposed that treatment readiness can be im proved by understanding both the general principles of treatm ent readiness presented throughout this book, as w ell as the particular features of sex offender treatment covered here. To facilitate treatm ent readiness, efforts can focus on m odifying the client, m odifying the treatm ent, or m odifying the setting. M odifying the client will involve recognising and adapting

123

Transitions to Better Lives

to the needs, defences and schem as presented by individual sex offenders. Readiness may be enhanced by m eeting the client with a clear understanding of these individual differences at the start of treatm ent, as w ell as throughout treatment, as readiness is a dynam ic feature that will fluctuate throughout the treatm ent process in response to internal and external factors. M odifying the treatment m ay involve greater flexibility in defining the goals of treatm ent, and allow ing therapists to adapt their approach to meet the needs and characteristics of individual clients (see Chapter 10). H aving w elltrained and experienced staff w ho facilitate effective treatm ent groups and a positive therapeutic environm ent is essential to developing and m aintaining treatm ent readiness. Creating an appropriate social clim ate in w hich to conduct effective therapy m ay present one of the biggest challenges to treatm ent readiness, particularly in the current environm ent of m any custodial settings. A lthough, it is heartening to note that the developm ent of prisons that are more treatm ent focused and supportive has begun in som e jurisdictions (Laws and Ward, in press). Facilitating the opportunity for offenders to participate in supervised leave program m es and com m unity-based treatm ent program m es m ay further prom ote treatm ent readiness and treatm ent effectiveness in the transition from custody to community. Greater attention to the factors that prom ote treatm ent readiness is considered necessary if w e are to continue to m ake progress in reducing recidivism and assisting offenders to achieve better lives.

124

Chapter 8

Substance use and readiness

Substance use is regarded as a pervasive problem in m any countries. In addition to the direct and indirect econom ic costs, there are negative social (e.g. fam ily dysfunction, unem ploym ent) and health (e.g. HIVAIDS) consequences associated w ith use. O f particular concern is the w ell-docum ented association betw een substance use and crim e (Graycar 2001; H iller et al. 1999; M akkai 2000), the frequency and severity of w hich has been directly linked to substance use levels (Anglin et al. 1999; Chaiken and Chaiken 1990; M akkai and Payne 2003). W hile alcohol is frequently im plicated in offending, particularly crim es involving violence, it is the w idespread use of illicit substances that is cause of widespread concern. For exam ple, betw een 60 and 80 per cent of arrestees test positive to at least one illicit substance (Kouri et al. 1997; M ilner et al. 2004; Prendergast and M augh 1994), w hile for incarcerated offenders, approxim ately two-thirds of first-tim e male and three-quarters of first-tim e fem ale prisoners report a history of problem atic drug use in the six m onths prior to incarceration (Makkai and Payne 2003). These figures rise to 80 per cent in m ales and 90 per cent of fem ales for second and subsequent incarcerations (Departm ent of Justice 2002). A pproxim ately 55 per cent of fem ales (Johnson 2004) and 52 per cent of m ales (M akkai and Payne 2003) meet the criteria for substance dependence while in prison, around 55 per cent of offenders relapse into drug use w ithin one m onth of release from incarceration, w hile some 75 per cent of offenders have reported using substances during their incarceration (Nurco et al. 1991). The typical crim inal justice response to the d ru g s/crim e nexus has been the im plem entation o f a broad range of initiatives, the intensity 125

Transitions to Better Lives

of w hich differs according to where interventions are offered. At the lower end, a range of police-based interventions target first or second-tim e offenders detected as being in possession of m inor am ounts of illicit drugs (such as cannabis or prescription drugs) a n d /o r drug im plem ents. These interventions are low intensity and typically involve an educational com ponent, ranging from on-thespot w ritten m aterial, through telephone-based education sessions, to m eetings with a specialist drug counsellor (W undersitz 2007). Court-based drug diversion initiatives sit at the interm ediate level and target offenders charged w ith crim es at the lesser end of the severity spectrum (for exam ple property, driving and good order offences) that are directly linked to the use of illicit drugs (although eligibility criteria can differ across jurisdictions). Treatment in diversion program m es is often relatively short term (three to four m onths) and m ost often undertaken at the pre-sentence stage. M ore serious offences are dealt w ith in form al drug courts, the first of which was established in the US around two decades ago, and which now exist in m ost w estern jurisdictions (Payne 2006). D rug courts deal with repeat offenders w hose crim inal behaviour (excluding violent or sexual offences) is directly connected to long-term drug dependency (usually heroin and am phetam ines), and for w hom the likely sanction is a term of im prisonm ent (potential participants must be eligible for bail or release). Treatment program m es usually last for 12 months, during which tim e the offender undergoes regular drug treatm ent as well as being provided w ith support in areas such as accom m odation, financial advice and health care (W undersitz 2007). Frequent and regular random urinalysis is m andatory and offenders w ho persistently fail this com ponent or repeatedly fail to com ply with other conditions set dow n by the court m ay be term inated from the program m e and sanctioned w ith a short term of im prisonm ent. Prison-based treatm ent program m es are at the upperm ost end of the crim inal justice continuum , w ith incarcerated substance users representing that sub-group of the crim inal population w hose offences are serious enough or w hose offending is sufficiently repetitive to justify im prisonm ent. In terms of treatm ent options, m ost correctional agencies offer opiate replacem ent therapy (although this is not as extensive in the USA as in other countries) in addition to a range rehabilitation program m es that are psycho-educational or psychologically based. An influential treatm ent paradigm (again, outside the USA) is that of offender rehabilitation, nam ely the risk need s-responsivity m odel, w hich is based on w ork conducted by Canadian researchers (see A ndrew s and Bonta 2006). This approach 126

Substance use and readiness

advocates that interventions should be targeted towards higher-risk offenders (risk principle), aim ed at changing those areas of need that are functionally related to offending (needs principle), and delivered in w ays responsive to the individual (responsivity principle). Risk is typically assessed using actuarial instrum ents, and treatment approaches are often cognitive behavioural in orientation. The issue of being ready for treatm ent is an im portant consideration at any of the above junctures. It m ay not be practical or appropriate to assess readiness at the earliest stages of intervention (for exam ple, police diversion), but once an individual's behaviour has reached the critical point of being arrested a n d /o r charged w ith a crim inal offence, it is relatively easy for the individual to becom e entrenched in a d ru g s/crim e cycle that can ultim ately lead to a term of im prisonm ent. The crim inal behaviour of those offenders who are sentenced to prison is likely to be m ore serious a n d /o r more entrenched than in those w ho are diverted, and if behaviour change is to be successful a com prehensive assessm ent of the offend er's readiness to change (and identification of factors that m ight com prom ise that readiness) is likely to be im portant. A s noted in Chapter 4, this assessm ent needs to be sufficiently broad, using m ultiple m ethods and sources. It should also take into account a num ber of factors w ith the potential to im pact on an individual's level o f readiness for treatm ent. Som e of these factors are addressed in this chapter.

Physiological considerations A starting point in any discussion of behaviour change in substanceusing offenders is the role o f addiction and dependence. W hile there is no laboratory test for dependence, it has been operationally defined in the D SM -IV-R as a pathological condition manifested by three or more of seven criteria. Two of these - tolerance and w ithdraw al - indicate neurologic adaptation (that is, physiological dependence). Physiologic adaptation on its own, however, is neither necessary nor sufficient for a diagnosis of substance dependence. Diagnosis requires that the individual show a com pulsive desire for and use of a particular substance or substances, even in the face of adverse consequences (A m erican Psychiatric A ssociation 1994). The propensity for addiction is also thought to have a genetic com ponent, w ith published heritability estim ates of 0.34 for m ales dependent on heroin, 0.55 for m ales dependent on alcohol (Tsuang et al. 1996), and 0.52 for fem ales dependent on m arijuana (Kendler and Prescott 1998). 127

Transitions to Better Lives

These figures are com parable to the heritability for such disorders as Type 1 diabetes (0.30-0.55; Kyvik et al. 1995) and adult onset asthm a (0.36-0.70; M cLellan et al. 2000). Drugs of dependence have an effect on the brain circuitry involved in controlling m otivated and learned behaviours (Koob and Bloom 1998). Drugs such as alcohol, cocaine and opiates all have significant effects on the dopam ine system , albeit through different mechanisms. Cocaine increases synaptic dopam ine by blocking re-uptake into presynaptic neurons, am phetam ines produce an increase in the presynaptic release of dopam ine, w hile opiates and alcohol disinhibit dopam ine neurons leading to an increase in firing rate. Given the dopam ine system is a reward system associated with feelings of euphoria, m any drugs of dependence (including cocaine and opiates) act to produce supranorm al stim ulation of the reward circuitry. The com bined actions of the neuroanatom y and neuropharm acology also produce an im m ediate and profound desire for re-adm inistration which, at a particular dose, frequency and chronicity, produces enduring (possibly perm anent) pathophysiologic changes to the reward circuitry, neurochem ical levels, and the stress response system (Kreek and Koob 1998). The exact am ount required to m ake such changes is unknown. D ependence can be understood as a chronic, relapsing condition due to the integration of the dopam ine reward circuitry and the m otivational em otional, and m em ory centres w ithin the limbic system (Childress et al. 1999). This interconnectedness is responsible for the substance user not only experiencing the pleasures of reward but also learning specific signals for reward and to respond in an anticipatory w ay (M cLellan et al. 2000). The repeated pairing of substance use with an object (needle), person (drug-using friend), or even an em otional state (anger, depression) can result in rapid and entrenched learning. This process helps to explain how a previously substance-dependent individual w ho has been abstinent for a long period will experience significant, conditioned physiological reactions (w ithdraw al, craving) when encountering the conditioned object, person or em otional experience previously associated w ith substance use (Childress et al. 1999). As noted above, chronic substance use can result in neuroanatom ical and neurochem ical dam age leading to im pairm ent across a num ber of cognitive dom ains (A haronovich et al. 2006). W hile there are generalised cognitive deficits across all classes of addictive substances (m ost notably decision-m aking deficits), subtle differences have been noted for specific substances due to distinct m odes of action (Rogers 128

Substance use and readiness

and Robbins 2001). Chronic users of cocaine, for exam ple, show deficits in attention, decision-m aking and problem -solving (Cunha et al. 2004; Tucker et al. 2004), w hile those with alcohol abuse disorders have been show n to have problem s acquiring new coping behaviours, learning and retaining new m aterial, and engaging with therapeutic interventions (Fals-Stew art 1993; M cC rady and Sm ith 1986; A lterm an and Hall 1989). Chronic m arijuana m isuse is associated with impaired mem ory as well as deficits in attention, abstraction and executive functioning (A haronovich et al. 2008; Lam ers et al. 2006), while M DM A (Ecstasy) use has dosed-related long-term deficits in verbal memory, processing speed, and planning and problem -solving (H anson et al. 2008). Cognitive im pairm ents may contribute to substance abuse and dependence in at least two ways. First, the various cognitive deficits can increase the likelihood of drug-seeking behaviour. For exam ple, decision-m aking deficits may lead substance-dependent individuals to persist with risky behaviour even in the face of know n negative consequences (labelled 'm yopia for the future'; Clarke and Robbins 2002). Research by Bechara and colleagues (Bechara and Dam asio 2002; Bechara et al. 2002) has shown that prolonged substance use can result in users (a) becom ing insensitive to future consequences regardless of the em otional valence, (b) show ing learned feedback about reward but not punishm ent guided by long-term decision-m aking, or (c) becom ing hypersensitive to reward but hyposensitive to punishm ent. Second, persistent use can interfere with the individual's capacity to participate in rehabilitation program m es w ith an educative and cognitive em phasis which, in turn, can lead to poor retention and high relapse rates. A dequate cognitive function is a requirem ent for m any of the activities in substance-use programm es. It is therefore reasonable to assum e that cognitively impaired substance users participating in cognitively oriented treatment may encounter difficulties since many treatment m odalities require the verbal, m em ory and learning skills that are often impaired as a result of chronic use.

Offending pathways The process by w hich an individual becom es involved in substance use m ay also influence decisions about the need to change either their offending behaviour or their substance use. A lthough clear evidence exists for a relationship betw een substance use and crim e, a direct causal link has yet to be established. From the considerable debate 129

Transitions to Better Lives

that has taken place, three possible explanations have em erged: (a) drug use leads to crim e; (b) crim e leads to drug use; and (c) any relationship betw een drugs and crim e can be explained by a set of com m on causes (Gorm an and W hite 1995). With respect to the proposition that drugs lead to crim e, G oldstein (1985) has identified three w ays this m ight occur. First, the pharm acological m odel posits that intoxication (disinhibition, poor judgem ent, cognitive-perceptual distortions) and its by-products (w ithdraw al, enhancem ent of psychopathological disorders, sleep deprivation) can lead to crim e (either w hile under the influence or as a consequence of substance use). Second, the econom ic m otivation m odel assum es that drug users com m it crim e (theft, burglary, drug dealing) to generate incom e to support their addiction. Third, the system ic m odel proposes an intrinsic link betw een the system o f drug distribution and violent crim e (such as 'turf w ars', physical assaults related to the collection of drug debts, robbery of dealers or buyers). A ccording to the second proposition, crim e leads to substance use because individuals involved in deviant behaviour are more likely than non-deviant individuals to find them selves in situations where substance use is condoned a n d /o r encouraged. It is the involvem ent in this type of subculture that provides a context for substance use (Collins and M esserschm idt 1993; W hite 1990). It has also been suggested that deviant individuals m ay use drugs in order to selfm edicate or to provide an excuse for their deviant behaviour (Collins 1993; Khantzian 1985). Finally, the com m on cause explanation argues no direct causal link betw een drugs and crim e but proposes instead that the two are related through com m on causes (for exam ple, childhood abuse, early school failure, fam ily problem s, tem peram ental traits, anti-social personality disorder, neighbourhood disorganisation; see W hite et al. 1993; W hite and Gorm an 2000). Subcultural norm s have also been im plicated in this explanation as substance use and crim e can help individuals achieve m em bership and status w ithin the subculture (Gorm an and W hite 1995).

Heterogeneity in the substance-using population A n im portant consideration w hen m aking decisions about treat­ ment, and by extension treatment readiness, is the com plex and heterogeneous nature of substance use and dependence and the im pact that this has on the individual offender presenting for treatment. D ependence occurs as a function of the interaction betw een physical 130

Substance use and readiness

(physical adaptation to the substance), psychological (reasons for substance use), and social factors (availability of drugs and drugusing lifestyle), all of which are strongly influenced by personal and environm ental factors (Gow ing et al 2001). Consequently, substance users can differ m arkedly in term s of their age; the length and severity of their substance use; the presence of co-m orbid conditions; the extent, type and severity o f offending; their levels of psychosocial functioning; the m otivations for and functions of substance use; their social status; and their treatm ent history. Importantly, they will also differ in term s of their level o f m otivation to change (or modify) their substance use, the extent to w hich they can engage in the treatm ent process, and the length of tim e any changes gained in treatment can be sustained (Sim pson et al. 1997). An exam ple of this heterogeneity is evident in Germ an and Sterk's (2002) study of crack cocaine users. Based on in-depth interview s w ith active users, the authors identified four sub-groups: stable users; tempted users; grappling users; and immersed users. The m ost salient dim ensions differentiating each typology were the level of protective strategies used by individuals to avoid negative consequences, and the frequency of use (which w as directly related to daily life structure and access to crack cocaine). Recently, W ittchen and colleagues (W ittchen et al. 2009) reported on a longitudinal study investigating the relationship betw een sustained cannabis use and m ental health problem s. W hile a large proportion of participants (59 per cent) were identified as 'unproblem atic users', 14.4 per cent had 'prim ary alcohol use d isord er', 17.9 per cent had 'delinquent can n ab is/alco h o l DSM -IV abuse', while the rem aining 8.5 per cent had 'cannabis use disorder with m ultiple problem s'. The probability of unm et treatm ent needs was highest for those falling into the category of prim ary alcohol disorder or m ultiple problems. Finally, a substantial body of research also points to the heterogeneity in alcohol abuse and dependence as a function of genetic predisposition (see M aher et al. 2002; van der Zw aluw and Engels 2009), culture (Bau et al. 2001; Delva et al. 2005), sex (Oxford et al. 2003), deviant peer influence (Li et al. 2002; W iesner et al. 2008), and co-m orbid m ental health problem s (M aher et al. 2002; Shum w ay and Cuffel 1996).

Compulsory and coerced treatm ent W hile use of the crim inal ju stice system to direct offenders to treatm ent is com m onplace in the US, its use has been the subject of fierce 131

Transitions to Better Lives

international debate (Klag et al. 2005; Wild 2006; Wild et al. 2002). Critics opposing coerced or com pulsory treatm ent have argued that individuals w ho are forced into treatm ent are less motivated (Brecht et al. 1993) and show greater treatm ent resistance (Shearer and O gan 2002), which subsequently com prom ises positive therapeutic and legal outcom es. O ther research has show n reductions in both drug use and crim e in mandated or legally coerced clients that is sim ilar to (and in som e instances better than) voluntary clients (Farabee et al. 1998; H iller et al. 1998; M arlow e 2001). Ethical concerns inevitably arise in any discussion of mandated or coerced treatm ent. There is a need to ensure that treatm ent does not supplant basic civil or hum an rights in m eeting the larger societal goal of reducing the risk substance users pose to others (K leinig 2004). W hile the issue of coercion and its relationship to treatment readiness has been articulated in Chapter 2, there are som e issues specific to substance-using offenders that require further com ment. Policies supporting the use of coerced treatm ent rest on three main assum ptions: (1) substance use and crim e are so highly correlated that a reduction in the form er will result in a concom itant reduction in the latter; (2) because the treatm ent substance use has been found to reduce crim inal behaviour, at least part of that behaviour is related to substance use; and (3) even substance-using offenders w ho do not perceive they have a substance-use problem have been shown to benefit from treatm ent (Prendergast et al. 2009). There is a strong body of evidence that supports all three assum ptions (for exam ple, A llen et al. 2001; A nglin and Perrochet 1998; Dow den and Blanchette 2002; Wild et al. 2002). The extent to w hich coerced clients and those who do not recognise or acknow ledge problem atic substance use can becom e engaged in the therapeutic process and have positive gains that are com parable to voluntary clients depends to a large extent on w hether the individual perceives the treatment as coerced. This is an im portant consideration, particularly as m uch of the coercion literature fails to distinguish betw een objective, external or legal sources of coercion (courts, prisons, parole boards) and the ind ividual's subjective experience (Prendergast et al. 2008). There is also an assum ption that crim inal justice clients undergoing coerced or mandated treatm ent do so against their will. A study by Farabee, Prendergast and A nglin (1998) revealed that m ost of the clients interviewed reported that they would have entered treatment in the absence of pressure from the crim inal justice system . Specific to the issue of readiness to change and coerced treatm ent, Gregoire and Burke (2004) used the Readiness to Change Q uestionnaire (Rollnick 132

Substance use and readiness

et al. 1992) to assess mandated and non-m andated substance users presenting as outpatient clients. A significantly higher proportion of the mandated group reported being in either the action stage or the contem plation stage. Better outcom es for mandated as compared to non-m andated treatm ent participants remained even w hen prior treatm ent and addiction severity w ere controlled. G ender differences were also noted, w ith fem ales being more likely to be in the action stage, independent of either severity o f substance-use or legal coercion.

Sex differences A lthough it is w idely acknow ledged that the aetiology and subsequent rehabilitation needs of m ale and fem ale offenders are different, much less is know n about the specific needs of fem ales presenting w ith substance-use problem s. W hile this m ay be a product of the considerable im balance betw een the male and fem ale offending populations, there has been a not unsubstantial increase in the num ber of fem ale offenders incarcerated for substance-related offences over the past decade. For m any of these w om en, offending is directly related to their involvem ent in drug-related crim es. Substance use is also m ore likely to precede the offending behaviour of fem ales than it does for m ales (Makkai and Payne 2003). For exam ple, in one A ustralian study of police detainees (M ilner et al. 2004), tw o-thirds of respondents reported having used illegal drugs prior to or w ithin the sam e year as their first offence. This suggests that for a substantial proportion of fem ale offenders, substance use plays a critical role in the aetiology of their offending behaviour. The origins of fem ale substance-use would appear to be embedded in the psychosocial problem s and traum atic life events that women frequently experience. Fem ale offenders w ith substance-use histories are more likely to have em ploym ent, mental health (depression, anxiety, bipolar affective disorder, psychosexual disorders, eating disorders, post-traum atic stress disorder) and fam ily problem s (fam ily dysfunction, lack of adequate role m odels) than their male counterparts. They are also m ore likely than m ales to have required help for em otional problem s at a younger age and to have attempted suicide (Brady and A shley 2005). Fem ales are also significantly more likely to exhibit recent physical, em otional or sexual abuse (Gentilello et al. 2000) and report more problem s related to physical and sexual abuse and dom estic violence victim isation than m ales (Green et al. 133

Transitions to Better Lives

2000). The physiological effects o f substance use also display gender differences, with fem ales tending to report m ore physical problem s and be m ore vulnerable to the physiological effects of substance use (W asilow -M ueller and Erickson 2001). G ender differences have also been noted in the reasons w hy fem ales initiate substance use. For exam ple, they are m ore likely to use substances to alleviate either physical or em otional pain (Corcoran and Corcoran 2001; Langan and Pelissier 2001) and more likely to have been initiated into both substance use and anti-social behaviour bv m ale partners (H enderson et al. 1994). It is notew orthy that the gender differences described above frequently m irror those of fem ale substance users w ho are non­ offenders. Perhaps more importantly, given the m uch harsher societal response to fem ales with substance-use problem s as com pared to m ales (Van der W alde et al. 2002), these problem s also parallel gender differences found w ithin the non-offending female populations (Brady and A shley 2005). O n a more positive note, despite the extent and com plexity of the problem s experienced by fem ales w ith substanceabuse histories, there would seem to be a 'gender paradox' at work: first, w om en are no more likely than m ales to relapse follow ing treatm ent, and second, they are m uch m ore likely than m en to engage in treatm ent (Fiorentine et al. 1997).

Co-morbidity Substance use disorders frequently co-occur w ith a range of other m ental health and behavioural disorders. Large epidem iological studies in the US, for exam ple, report lifetim e prevalence rates for co-m orbid substance disorders and mood disorders (26 per cent), anxiety disorders (28 per cent), psychotic thought disorders (7 per cent), and anti-social personality disorder (18 per cent) (Kessler et al. 1997; Regier et al. 1990). Reiger and colleagues also noted that more than half of those identified as m eeting the criteria for a drug disorder also had a m ental disorder o f som e type. The figures are sim ilar in clinical populations for patients seeking treatment for substance use and co-m orbid m ood disorder (41 per cent), anxiety disorder (26 per cent), post-traum atic stress disorder (25 per cent), severe (psychotic) mental disorder (17 per cent) and borderline personality disorder (17 per cent) (M cG overn et al. 2006). Recent research indicates gender differences in the pattern of comorbid conditions and temporal onset, in both general and clinical 134

Substance use and readiness

populations. For exam ple, fem ales w ith substance-use disorders report higher rates of anxiety, depression, eating disorders and borderline personality disorder, w hile m ales report higher rates of anti­ social personality disorder (Zlotnick et al. 2008). G ender differences in tem poral onset show that as com pared to m ales, fem ales tend to have prim ary psychiatric disorders that antedate the onset of substance-use disorders. This would indicate a difference in the aetiological relationship of substance use and co-m orbid disorders from that found for m ales (Kessler 2004). The relationship betw een addiction, traum a, and post-traum atic stress disorder is also stronger in fem ales. This is particularly so for fem ale victim s of childhood abuse as well as those exposed to violence in adulthood (Kendler et al. 2000). Rates of co-occurring substance use and psychiatric disorders in male (Sw artz and Lurigio 1999) and fem ale offenders (A bram et al. 2003) are sim ilar to those found in the general and clinical population. For exam ple, it has been reported that fem ale offenders undergoing substance-use treatm ent are twice to three times more likely to meet the diagnostic for m ajor depression, post-traum atic stress disorder, borderline personality disorder, or any affective, anxiety, or psychotic disorder (Zlotnick et al. 2008). A nti-social personality disorder is m ore prevalent in male offenders. As com pared to their male counterparts, fem ale offenders report m ore severe substance-use histories and more frequent co-m orbidity prior to incarceration (M essina et al. 2006) and lifetim e prevalence for depression, anxiety, and psychosis (M essina et al. 2003). The com bination of a substance-use disorder and co-occurring psychiatric condition has a detrim ental im pact on treatm ent engage­ m ent and therapeutic outcom es due to the com plex interaction of the two problem s (DiClem ente et al. 2008). For exam ple, individuals with a dual diagnosis that includes substance-use disorder have been found to not adhere with m edication regim es, to experience sym ptom exacerbation, and have frequent episodes of rehospitalisation, poor social adjustm ent, and poor prognosis (Drake et al. 1996). Treatment is also characterised by a lack of therapeutic engagem ent, low levels of m otivation to change, and frequent relapse. The more severe the substance-use problem and co-m orbid psychiatric condition, the greater the likelihood that dysfunctional thought processes, impaired decision-m aking capabilities, and poor insight will im pair the ability to either recognise or seek out and participate in treatment.

135

Transitions to Better Lives

T reatm en t considerations The above d iscu ssion highlights the heterogeneou s nature of the substance-u sing population. G iven this heterogeneity, it is im portant that treatm ent providers end eavou r to m atch services to ind ividual need, taking into accoun t levels of m otivation and read iness to change w hen m akin g d ecisions abo u t treatm ent type and intensity. O ne o f the first con sid erations in d eterm ining an in d iv id u al's treatm ent needs is their level of m otivation to change. M otivation is at the core o f successfu l behav iou ral change, although m any substance-u sing offend ers can present as bein g am biv alen t about their substance use and the need to change (although even am bivalen t clients can benefit for treatm ent if they rem ain in treatm ent). M otivation reflects the degree o f interest o r con cern an ind ividu al has about the need for change, their goals o r intentions, the need for them to take responsibility, the capacity to sustain change that is m ade, and the need to have ad equ ate incen tiv es to change (D iC lem ente et al. 2004). Perhaps the m ost influential m odel of m otiv ational tasks and intentional behav iou r change is Prochaska and D iC lem en te's (1984) transtheoretical m odel, described in C h apter 4. A staged m odel o f change, it includes pre-action (pre-contem plation, contem plation, preparation), action, and m aintenan ce stages. P re-action tasks are preparatory: for exam ple, acknow led g in g that substance use is p roblem atic, creatin g goals, form u latin g intention, p lanning o f and com m itm ent to change. The actio n stage is w here beh av iou r change is initiated (ceasing or red ucing substance use) w hile the m aintenance stage involves su staining any change o v er time. A ccord ing to D iC lem ente (2005), forward m ovem ent through these stages can only occu r if stage-specific tasks are sufficiently w ell accom p lished. T he m odel also posits that behav iou ral change is likely to occu r only after the ind ividual has 'recy cled ' through the stages several tim es. The activ ities and experiences used to com plete the various tasks and facilitate progress through the stages are referred to as P rocesses o f C hange. These processes, w hich inclu de both co g n itiv e /e x p e rie n tial (con sciou sn ess raising, self re-evaluation, environm en tal re-evalu ation , em otional arou sal) and behav iou ral activ ities and strategies (self-liberation, stim ulus control, cou nter con ditionin g, con tin gency m an agem en t, and helping relationships), are relevan t to the tasks required at each of the different stages in ord er that the task goals o f each stage be m et, that forward m ovem ent through the stage p rocess occurs, and that behaviou ral change ou tcom es are achieved (P rochaska et al. 1992). 136

Substance use and readiness

For problem atic substance use, m otivation and intention are critical com ponents of treatm ent readiness. Increased m otivation has been positively associated w ith treatment engagem ent, attem pts to stop substance use, reduced alcohol consum ption, treatment retention, sustained abstinence, and m ore positive life outcom es (Joe et al. 1998; Project M ATCH Research Group 1997). This highlights the im portance of assessing m otivation to change in conjunction w ith an assessm ent of treatm ent readiness. The m ost frequently used m easures have been described in Chapter 4. O f these, the U niversity of Rhode Island Change A ssessm ent (U RICA) (M cConnaughy et al. 1983, 1989) has versions designed specifically for alcohol and drug use (see A ppendix). In addition to the original 32-item scales, both have shorter versions that can also be used to assess change over time (that is, pre- and post-treatm ent). Rehabilitation efforts can then be targeted at the individual's location in this cycle. For exam ple, education program m es aimed at im proving m otivation m ay be m ost appropriately aim ed at those offenders who are reluctant to enter more form al treatment program m es. An alternative strategy is the use of m otivational enhancem ent techniques (such as m otivational interview ing; M I) w hich m ight be used to engage the individual in the sorts of cognitive (for exam ple, psycho-education) and experiential (for exam ple, dram a therapy) activities that m ight prom ote preparation for change. Described m ore fully in Chapter 10, M I is a collaborative process w hereby the therapist helps identify the potential for change w ithin the client and facilitates the change process. Readiness for change can be increased w hen M I serves to enhance engagem ent and reduce reactance. This is especially pertinent w ith offenders, w ith whom engagem ent in change is often difficult to achieve. M atching services to the pre-treatm ent characteristics of the client is perhaps the sim plest form of adapting treatm ent to the needs of the individual. For exam ple, tailoring services to the specific needs of fem ales, youths or m inorities is a basic approach to adapting treatment for the client. Similarly, a history of negative outcom es in previous treatm ent episodes m ight lead a program m e to prescribe a higher dosage or greater variety of services for a new client than would ordinarily be provided to new adm issions. This, too, is a sim ple form of adapting treatm ent to the needs of the individual. A t a m ore sophisticated level of adaptive program m ing, the nature and intensity of services is continuously readjusted as a consequence of the client's ongoing perform ance in treatm ent. For exam ple, if a client fell below an effective threshold for engagem ent in treatment (such as m issing a predetermined num ber of scheduled counselling 137

Transitions to Better Lives

appointm ents), he or she m ight be reassigned to a m otivational enhancem ent intervention or m ight be encouraged to try a different and more desired form of treatment. The notion of m atching client needs to treatment intensity has informed two of the largest substance-use treatm ent evaluations undertaken: Project M ATCH (Project M ATCH, 1997) and Project REFORM (Wexler et al. 1991). Project MATCH was an eight-year m ulti-site study conducted in the US that was designed to test the assum ption that treatm ent outcom es could be improved by carefully m atching individuals to therapeutic approaches for alcohol a b u se / dependence, based on their personal characteristics. W hile positive gains were only noted for those identified as having low er levels psychological distress, this m ay have been a function of the intensity of program m es offered, all of which would fall into the category of low intensity under current best practice guidelines (such as 12-step program m es, 12-session CBT-based coping skills, and brief m otivation enhancem ent). Individuals with higher levels of distress are more likely to benefit from more intense treatments. This is noted by Wexler, Blackm ore and Lipton (1991) when w riting about Project REFORM . In their view, prisoners assessed as having 'm oderate' level substanceuse problem s should be assigned to non-residential program m es (reside in m ainstream but undertake program m e) w hereas those offenders w ith severe or chronic substance-use problem s should be directed to isolated, residential treatm ent program m es. A lthough there are few published accounts of good practice in offender substance-use program m ing, an analysis of Project REFORM (Wexler et al. 1991) suggests the follow ing principles and guidelines for program m es in correctional system s that seek to reduce recidivism. • Assist the offender to identify personal im pedim ents to recover. • Provide the offender with incentives, positive or otherw ise, to participate in recovery program m es. • Separate participants from the general prison population as soon as possible. • Reinforce pro-social behaviours rather than attem pt to directly reduce the frequency of negative behaviour. • Establish clear, unam biguous rules and consequences for breaking such rules.

138

Substance use and readiness

• Establish clear behavioural contingencies. • Em ploy ex-offend ers/ex-ad d ict staff to serve as role m odels. • M aintain treatm ent integrity, autonom y, flexibility, openness, and fiscal and political support. • Establish continuity of intervention, from the outset of custody to its term ination. • Establish program m e effectiveness.

evaluation

system s

and

analyse

cost-

The US N ational Institute on Drug A buse also specifies the follow ing principles of effective program m ing: • No single treatm ent is appropriate for all individuals. • Treatment needs to be readily available. • Effective treatm ent attends to m ultiple needs of the individual, not ju st his or her drug use. • An individual's treatm ent and service plans m ust be assessed continually and m odified as necessary to ensure that the plan meets the person's changing needs. • Rem aining in treatm ent for an adequate period of tim e is critical for treatm ent effectiveness (m inim um of three months). • Counselling (individual and group) and other behavioural therapies are critical com ponents of effective treatm ent for addiction. • M edications are an im portant elem ent of treatm ent for m any patients, especially when com bined with other behavioural therapies. • Addicted or drug-abusing individuals w ith co-existing mental disorders should have both disorders treated in an integrated w ay • M edical detoxification is only the first stage of addiction treatment and by itself does not change long-term drug use. • Treatment does not need to be voluntary to be effective. • Possible drug continuously.

use

during

treatm ent

must

be

monitored

139

Transitions to Better Lives

• Treatment program m es should provide assessm ent for H IV /A ID S , hepatitis B and C, tuberculosis and other infectious diseases, and counselling to help patients to m odify or change behaviours that place them at risk of infection. • Recovery from drug addiction can be a long-term process and frequently requires m ultiple episodes o f treatment.

Conclusion Given the chronic, relapsing nature of substance-use disorders, individuals w ho present for treatm ent (w hether voluntarily or in response to som e level of coercion) m ight not be m otivated to change and, therefore, are unlikely to be treatm ent ready. W hile this may be the case, there is evidence to support the proposition that once in treatment, the experience o f the therapeutic process can result in positive gains irrespective o f m otivation levels. There are, however, a num ber of factors internal and external to the individual that have the potential to negatively im pact on treatm ent outcom es, even in those w ho wish to participate. Chronic substance use can, for exam ple, result in physiological change that increases the likelihood of treatm ent failure (for exam ple, through relapse or an inability to participate fully in the treatment process due to cognitive deficits). The heterogeneity of substance-use problem s is also a potential im pedim ent to treatm ent success. Finally, the presence of a co ­ occurring psychiatric disorders may result in the individual being differentially m otivated with respect to each condition to the extent that neither is adequately addressed. This all suggests that treating substance-use problem s in a crim inal justice context is likely to be particularly challenging, and highlights the need to carefully assess those w ho are referred to program m es to ensure that they are both able and w illing to undertake a process of behaviour change. The fram ew ork for understanding treatm ent readiness in this book offers a structure by w hich offenders can be assessed and a rationale for offering com prehensive assessm ents. There is a need to ensure that the scope of any assessm ent is sufficiently broad to enable treatm ent to be m atched to the needs and characteristics of the individual presenting for treatment.

140

Chapter 9

Readiness and treatm ent engagement in personality disordered offenders: towards a clinical strategy 1

In the decade since the advent o f the m illennium , therapeutic and rehabilitative interventions for offenders have proliferated in m any jurisdictions and the confidence of forensic and clinical practitioners appears to have grown. Proliferation and confidence are likely to be a product, in part, of the substantial evidence base that has becom e available, indicating that treatm ent and rehabilitation program m es can indeed have an im pact in changing im portant outcom es, such as recidivism , and suggesting that positive outcom es are associated w ith particular features of program m es, such as attention to the 'risk, needs and responsivity principles' (Andrews and Bonta, 2006; Hollin and Palmer, 2006, Ward and M aruna 2007). The grow th in confidence has had a num ber of positive effects. One of the latter has been the em ergence of healthy conceptual and em pirical critiques of the W hat Works m ovem ent (Ward and M aruna 2007), another has been the developm ent of w hat m ight be termed a positive psychology perspective on the goals of treatm ent and rehabilitation, with an em phasis on 'good lives' (Ward and Brown 2004). A third effect has been an expansion of the field to accom m odate other populations. W hile W hat Works program m es developed m ainly in crim inal justice system s, m any offenders, particularly high-risk offenders, are to be found w ithin forensic m ental health services. ‘We would like to thank Matthew Tonkin, Allison Tennant, Lawrence Jones, Mary McMurran and Kerry Sheldon for their contributions to my thinking and the NIHR (CLAHRC) programme for supporting work in the Institute of Mental Health, Nottingham University. 141

Transitions to Better Lives

Typically offending behaviours such as violence or sexual offending occur in the context of the presence of a m ental disorder, and adm ission to the health system under mental health legislation is seen as preferable to adm ission to the crim inal justice system (M cM urran et al. 2008a). There is some evidence that the crim inogenic needs (and hence treatm ent targets) for m entally disordered and non-m entally disordered offenders overlap substantially and that the causal influence of mental disorder per se for serious offending is som ewhat uncertain (Duggan and Howard 2009; H ow ells, in press). It would follow from the latter that w hat have becom e established types of offender rehabilitation program m es are also likely to be necessary in forensic m ental health services, though progress in im plem enting and evaluating such program m es in forensic m ental health system s has been slow (H ow ells et al. 2005; M cM urran et al. 2008a). It is likely that distinct problem s arise, or will arise, as interventions develop for the m entally disordered offender. The features of disorders them selves have the capacity to im pair the individual's ability to benefit from program m es and there is a consequent need to adapt such program m es to accom m odate the characteristics of the m entally disordered person, as would be suggested by the responsivity principle (A ndrew s and Bonta 2006). O ne increasingly recognised area in w hich all treatm ent program m es m ay be deficient is that of insufficient attention to responsivity. It has been proposed in this book that an im portant dim ension of responsivity is the readiness of the person for the program m e and the readiness of the program m e itself for the particular population to w hich it is to be delivered. The assum ption here is that low readiness produces low engagem ent, which in turn influences non-com pletion of treatm ent and ultim ately poor treatment outcom es. Certainly, there is evidence for non-com pletion of program m es in offender populations, particularly in com m unity program m es, where betw een a third and a half of starters fail to com plete. There is less evidence available as to com pletion rates in m entally disordered populations, though engagem ent problem s appear to be com m on for people with schizophrenia, particularly w hen substance m isuse is also present (Tait et al. 2002). Hodge and Renw ick (2002) have described extensive problem s of m otivation and engagem ent in m entally disordered offenders and these have been explored b y Sainsbury, K rishnan and Evans (2004). Such problem s m ay be accentuated for the personality disordered offender (PDO). The PDO has, in the past, been seen as falling into the gaps betw een m ental health and crim inal justice services. Personality 142

Readiness and treatment engagement in personality disordered offenders

disorder itself has been neglected w ithin m ental health services and it is only in the past decade that there has been a considerable increase in clinical and academ ic interest in personality disorder, particularly in the links betw een personality disorder and risk of violence (Duggan and Howard 2009; M cM urran and Howard 2009). In the United Kingdom , increased interest has stim ulated the rapid developm ent of new clinical assessm ent and therapeutic services, particularly for those w hose personality disorder is seen as severe and w hose ensuing risk is deem ed to be very high. The Dangerous and Severe Personality Disorder (DSPD ) program m e, for exam ple, an initiative of the UK governm ent, has given rise to a range of high security and com m unity services for this population (see H ow ells et al. 2007, for an overview). It was acknow ledged from the outset of the D SPD program m e that these patients would pose particular challenges to services and this has proven to be the case, though the units created have now been in operation for more than five years. Patients with severe personality disorders and high risk fall at the extrem es of dim ensions of mental disorder and of offending and thus m ight be anticipated to be particularly difficult to engage. This population includes a significant group w ho m eet the criteria for psychopathy (Sheldon and K rishnan 2009). H em phill and Hart (2002) have reviewed evidence that individuals w ith psychopathic traits resist the role of being a m ental health patient and that their treatm ent engagem ent is low. Difficulty in form ing a therapeutic relationship w ith therapists, need for control and dom inance, and im pulsivity in the treatm ent environm ent m ay all w ork against treatm ent engagem ent in such individuals (H emphill and Hart 2002; H ow ells and Day 2007). It is im portant that readiness factors are seen as both intrapersonal and situational (see the M O RM m odel, Chapter 1). Thus there is a need to investigate the situational context of adm ission to a hospital as well as the social and environm ental aspects of the treatment environm ent. Typically, for exam ple, D SPD patients have been trans­ ferred to a high secure hospital setting from a prison where they have either failed to com plete m ainstream offender program m es or been perceived as unlikely to be able to cope with them, often because of personality disorder traits. For DSPD patients, they may have been transferred from prison under the M ental H ealth A ct and against their will, late in their sentence, when they may have been anticipating release into the community. It m ight be expected that such circum stances would, at least initially, enhance resentm ent, resistance to treatment and perceptions of unfair coercion. 143

Transitions to Better Lives

Treatm ent completion in the personality disordered M cM urran, H uband and O verton (subm itted) have conducted a system atic review of com pletion o f therapy in people with personality disorder and also o f correlates of non-com pletion (drop­ out). M cM urran et al. found three categories of readiness factor to be present in published studies: patient characteristics, need factors and environm ental factors. The authors reported a m edian non-com pletion rate of 37 per cent in studies of PDOs, pointing out that this figure was higher than that (20.5 per cent) reported in a recent review of RCTs of psychological treatm ents for people w ith PD (Duggan et al. 2007) but slightly low er than that found (47 per cent) in a m eta-analysis of drop-outs from general psychotherapy. As M cM urran et al. suggest, non-com pletion rates for PDO s may be com parable to those for other psychological disorders but the rate is nevertheless substantial and likely to have a num ber of adverse consequences. W hat is unclear is w hether crim inological characteristics and psychological disorder interact w hen they are co-m orbidly present w ithin the individual. Do PDOs have particularly poor com pletion rates? A factor likely to confound answ ers to the latter question is the likelihood that such patien ts/offen d ers will often be found in secure forensic mental health or prison institutions where there are often strong pressures, even perceived obligations, to attend therapy sessions and to com plete treatm ent, w hereas m any personality disordered people w ho are not serious offenders are likely to be treated in a com m unity setting where the patient can more easily drop out. Treatment program m es for the personality disordered offender, particularly where the PDO is deem ed to present a high risk (H owells et al. 2007) are likely to involve a high degree of coercion (Day et al. 2004). In the D SPD population, for exam ple, offenders in prison may be involuntarily transferred to a high-security forensic mental health service under the M ental H ealth Act in England and Wales, in order to receive treatm ent. W hile studies of recidivism in the past have reported mixed results in term s of the im pact of coercion on outcom es such as reduction in recidivism , Parhar et al.'s (2008) com prehensive m eta-analysis provides evidence that m andated treatm ent is more likely to be ineffective, particularly w hen treatm ent is located in a custodial as opposed to a com m unity setting. Individuals diagnosed w ith anti-social personality disorder (ASPD), in particular, have typically been seen as problem atic to treat, with a resulting rejection of them by mental health services (Duggan 2009). Duggan identifies several 'attitudinal barriers' to treatm ent of ASPD, 144

Readiness and treatment engagement in personality disordered offenders

including perceptions by clinicians that those w ith A SPD may be difficult to w ork with because of traits of rule-breaking, egocentricity and defensiveness, and also the 'lukew arm ' com m itm ent of such people to interventions and therapies. It is reasonable to conclude that the negative perceptions, expectations and pessim ism of clinicians and services, although often based realistically on previous experience of w orking with such populations, nevertheless may som etim es constitute a situational factor im pairing readiness to engage in patients and offenders. One im plication is that m easurem ent of, and developm ent of interventions to modify, such expectations are im portant tasks for the future (discussed below). O ne notew orthy feature of previous studies of low engagem ent has been the pejorative w ay in w hich low engagem ent has been described. The patient may be described as 'resistant' or 'unm otivated'. Such descriptions tend to pathologise the person and to ignore the possibility that non-engagem ent m ay be a product of legitim ate and rational concerns and of problem s being in the treatm ent environm ent rather than w ithin the person (see Chapter 1). It is apparent from the above discussion that a host of readiness and engagem ent issues arise in relation to PDOs and that these m ay be even m ore m arked for those adm itted to high-secure services of the DSPD type. The failure to engage in such services also has large potential econom ic costs, in that treatm ents have been developed that are often labour and resource intensive. Such an investment of resources is vitiated should patients fail to becom e engaged in treatm ent or drop out from it.

Conceptions of readiness U tilising H ow ells and D ay's (2003) analysis of im pedim ents to therapeutic engagem ent in patients w ith anger problem s, Ward et al. (2004b) define treatm ent readiness as 'the presence of characteristics (states or dispositions) within either the client or the therapeutic situation, w hich are likely to prom ote engagem ent in therapy and which, thereby, are likely to enhance therapeutic change'. Ward et al. suggest that these factors can reside w ithin the person, the context, or w ithin the therapy or therapeutic environm ent, and outline a num ber of internal and external factors that are likely to influence the extent to w hich an individual approaches and ultim ately engages w ith the treatm ent being offered. In the readiness model (the M ORM ) outlined by these researchers, internal readiness factors fall w ithin 145

Transitions to Better Lives

particular dom ains: cognitive, affective, identity-related, volitional (m otivational) and behavioural skills. The cognitive dom ain, for exam ple, includes personal beliefs about the nature of the problem , therapy expectations and self-efficacy beliefs, w hile the affective dom ain includes excessively high or excessively low subjective distress. Volitional factors refer to the personal goals of the patient, which may be either consistent or inconsistent with the goals of the treatm ent offered. External readiness factors include lack of social support (staff, fam ily and friends) for engagem ent in treatment, responsivity failures in the treatm ent program m e itself, and poor therapeutic clim ate w ithin the treatm ent setting. There m ay also be w ider influences that affect the degree to w hich patients engage with treatm ents, associated with, for exam ple, the organisation of care, the roles ascribed to different professional groups, the im pact of national and local-level policies, and other organisational factors. H ow ells and Tennant (2007) have suggested that the M O RM model, when applied in a personality disorder service, directs the clinician to ask a series of straightforward questions: • H ow does the patient perceive the therapeutic program m e? • Does the patient believe he or she is capable of change? • H ow does the patient perceive the therapeutic staff and the system ? • W hat expectancies has the patient acquired, for exam ple from previous exposure to therapy? • W hat is the patient's affective state? • W hat em otional reactions does h e /s h e have to their offending and other related problem s? • W hat are the personal goals o f the patient? • Are personal goals congruent w ith the goals of the program m e? • W hat is the self-identity of the patient and is it com patible with treatm ent? • Does the patient have the behavioural and cognitive skills required b y the program m e? • Does the alliance?

146

patient

have

the

capacity

to

form

a

therapeutic

Readiness and treatment engagement in personality disordered offenders

• W hat support is given by staff and other patients for therapeutic engagem ent? • W hat support is given by relatives and friends for therapeutic engagem ent? • H ow does the patient perceive coercion into treatment? • H as the program m e been adapted for the (DSPD) population? • Is the clim ate of the service supportive of engagem ent, therapy and of change?

External readiness conditions External readiness factors are im portant to consider in treatment services for the PDO. A high proportion of personality disordered offenders will be found w ithin the prison system or w ithin secure forensic m ental services (M cM urran et al. 2008). W hile PDOs are likely to have m any needs in com m on w ith offenders (particularly violent and sexual offenders) w ho have not been diagnosed as having a personality disorder, they w ill also have distinctive needs that m ay require adaptations of the content of program m es and of their style of delivery (Tennant and Howells 2010). To implem ent a sex offending program me w ith PD sex offenders in exactly the same way as would be done with non-PD sex offenders would infringe the Responsivity Principle and be an instance of low program m e readiness (Jones 2010). This latter problem may be dim inished in part by the high probability that m any offenders in m ainstream offender program m es in prison settings would prove to be diagnosable as having a PD, were they to be clinically assessed. To this end, future research is needed, evaluating the level of PD in those attending m ainstream non-PD program m es in the crim inal justice system , to determ ine whether and how they m ight differ from those already diagnosed as personality disordered. It is possible that features of institutional life, particularly in prisons, w ork against engagem ent in therapeutic program m es in some cases and that the services are thus 'unready'. The provision of therapy is typically not a prim ary goal for prison systems. Even where therapeutic goals are acknow ledged as im portant, for exam ple in specialist therapeutic prisons, they are secondary to the custodial and deterrence functions of im prisonm ent. PDO services located within prisons face the challenge of m aintaining a therapeutic rather than a 147

Transitions to Better Lives

custodial regim e and the effort required is necessarily greater than it is w ithin high-security m ental health services, where understanding the individual, the need for treatm ent enabling patients to improve their w ell-being, and social adjustm ent have historically been core, high-priority goals. It must be acknow ledged, however, that in the past m ental health services have also struggled to prevent institutional environm ents from becom ing 'toxic' (Davies 2004). Hodge and Renw ick (2002) and Ginsburg et al. (2001) have identified a range of organisational factors that can im pede readiness, including perceived lack of safety, poor facilities, security versus therapy disputes, adversarial staff-service user relationships, pejorative labelling of those in treatm ent and reinforcem ent o f non-engagem ent w ithin the inform al institutional culture. The identification and m easurem ent of features of settings and organisations that m ight im pede engagem ent in treatm ent is an im portant task for the future. W ithin both m ental health and correctional settings the construct of therapeutic clim ate has becom e an im portant one. The m ost relevant body of scientific w ork for understanding, m easuring and m odifying negative environm ents and m ilieux is that on social clim ate, pioneered for over 30 years by Rudolph M oos (see Timko and M oos 2004) in a range of service settings, including health, mental health and correctional institutions. O ne of M oos's m ost im portant contributions has been the Ward A tm osphere Scale (available in different versions for different settings; M oos 1997). In recent years the M oos scales have been subject to critiques (for an overview' see Schalast et al. 2008) that suggest problem s with item content being outdated, and the time and effort required for com pletion being too great for repeated clinical use; and criticise the low internal consistency of som e scales, and discrepancies betw een the scales and the factor structure of the m easure (Rossberg and Friis 2003). In response to such criticism s, N orbert Schalast at the U niversity of Essen in Germ any has developed a new clim ate measure, specifically designed for use in forensic psychiatric wards. This is a 15-item instrum ent (named EssenCES: Essen Clim ate Evaluation Schem a). The m easure has three factor-analytically supported scales: Therapeutic Hold, Patients' Cohesion and M utual Support, and Experienced Safety. The Therapeutic Hold scale assesses perceptions of the extent to w hich the clim ate is supportive of patients' therapeutic needs; the Patient Cohesion scale assesses w hether mutual support of a kind typically seen as characteristic o f therapeutic com m unities is present; and the Safety scale assesses tension and perceived threat of aggression and violence. 148

Readiness and treatment engagement in personality disordered offenders

In a recent validation of EssenCES, Schalast and colleagues collected data in 17 forensic mental hospitals in Germany. H igh internal consistencies were found for the scales and good support for the expected factor structure. Convergent validity was dem onstrated in terms of correlations with related m easures, including jo b satisfaction in staff and perceived therapeutic m ilieu (Schalast et al. 2008). Since this initial validation study, EssenCES has been translated into a num ber of different languages, including Japanese, Dutch, Finnish and English. The English translation of EssenCES has been used in English high- security settings, particularly in high secure services for PDOs (H ow ells et al. 2009). In this latter study, internal consistencies, factor structures and correlations w ith m easures of ward experience, clim ate and m ilieu were broadly sim ilar to those reported by Schalast et al. (2008) in their Germ an sample. EssenCES has considerable face validity as a method for operationalising environm ental readiness factors in that perceived supportiveness o f therapy by staff and fellow patients in the setting, as m easured by EssenCES, are likely predictors of engagem ent (Ward et al. 2004b). N evertheless, there is a clear need to test directly the hypothesis that levels of engagem ent will be higher in wards and units w ith m ore positive therapeutic climates. A com parison of clim ate in prison as opposed to health service units would also provide an opportunity to test the hypothesis (above) that environm ental support for engagem ent is likely to be low er in prison settings. Therapeutic clim ate is a dynam ic rather than a static factor and likely to change over time, thus a measure such as EssenCES could be used as a dependent m easure in p re-p o st com parisons for service changes or initiatives to im prove engagem ent in treatment in PDO or other groups. Initiatives may take the form of increased staff training, changes in unit philosophies and m ethods of w orking or even im provem ents in the physical design and layout of buildings.

Assessing MORM characteristics in a personality disordered offender population The M ultifactor O ffender Readiness M odel (M O RM ) provides a useful fram ew ork for conceptualising and classifying potential im pedim ents to engagem ent and has influenced recent studies w ith offender populations, where psychom etric m easures have been developed derived in part from the M ORM (see Chapter 5). An alternative approach, particularly suited to assessing long-stay personality 149

Transitions to Better Lives

disordered offenders in clinical health settings, is to use routinely collected clinical interview m aterial (often extensively available in such intensive services) to classify reasons proffered by patients them selves as to reasons for not engaging in or failing to com plete treatm ent. In a recent prelim inary study of this sort Sheldon, H ow ells and Patel (subm itted) studied clinical interview m aterial in this w'ay and devised a brief manual based on the M ORM to be used by raters. In this case the patients were severely personality disordered and high-risk in-patients in a high security hospital. The rating system is shown below, adapted from Sheldon, H ow ells and Patel. M ultifactor Offender Readiness Rating Manual: internal and external factors

Internal readiness factors influencing treatm ent com pletion and engagem ent Cognitive factors: These include appraisals and beliefs about the treatm ent on offer, for exam ple its relevance and likely effectiveness, beliefs about the therapists, the crim inal justice or health system and about self-efficacy (perceived ability to m eet the requirem ents of the program m e). Examples: 'I d on't trust the staff'; 'You can't do treatm ent in places like this'; 'These program m es d on't w ork'; 'It w ould n't be any use to m e'; 'I will never get out so there is no point'; 'I struggle in groups'; 'It's too com plicated'; 'I w on't be able to do the hom ew ork' 'I am not attending because I d on't like the other group m em bers'. A s the cognitive factor is likely to be a com m on one, we also indicated w'hich of the follow ing cognitive dom ains applied: (a) self-efficacy belief (b) negative staff evaluation (c) negative patient evaluation (d) negative program m e evaluation (e) negative outcom e expectation (f) negative system evaluation (of unit or forensic service). Affective factors: This includes the level of general distress of the patient but also specific em otional reactions to previous offending that m ight lead to low engagem ent (for exam ple, shame). Therefore, any statem ents indicating that em otional arousal, or its absence, has contributed to low engagem ent would be categorised as affective. Examples: 'I w as in too m uch of a state'; 'H e w as overw helm ed w ith fear about telling others about his past'; 'I was too disturbed to be able to participate'; 'I felt too self-conscious to participate'; 'I am frightened of other participants and their reactions'. It is of note that

150

Readiness and treatment engagement in personality disordered offenders

som e of the descriptions that would count as 'affective' m ay also qualify as 'cognitive' or 'behavioural'. For exam ple, 'I couldn't cop e' could reflect self-efficacy beliefs, em otion or a lack of group skills. In general, an em otional state m ust be indicated as an antecedent to drop-out: for exam ple, sadness, fear, anger, distress, guilt, sham e or their opposites. Volitional factors: This com prises the personal goals being pursued by the patient, the location of offending behaviour w ithin the goal system of the patient, that is, the functions it serves for them (see Daffern et al. 2007), the effectiveness o f strategies for achieving personal goals and the congruence of personal goals with the explicit or implicit goals of the therapeutic program m e being offered. Generally this would include any indication that a person is not engaging or dropping out because they have other priorities (other goals they are pursuing). Included in this category would be non-engagem ent or drop-out due to m edico-legal reasons. Examples: 'C ontrolling my temper is not im portant in my life'; 'I d on 't w ant to im prove m yself'; 'All I w ant is to do time and not get into trouble'; 'I will be out in six m onths so there is no point in therapy', 'I have done therapy before and dealt w ith these problem s'; 'H is solicitor appears to have advised him not to cooperate in treatm ent, if he w ants to get out of the unit quickly'. Behavioural factors: This includes evaluations of their problem behaviours (for exam ple, previous offending) as not actually being a problem , help-seeking behaviours and, importantly, the diverse behavioural and cognitive skills required to becom e engaged in the treatment program m e on offer. Examples: 'I cannot talk in groups'; 'The hom ew ork is too tough'; 'I can't take criticism '. Identity factors: This includes core values and beliefs that constitute the person's identity. Identity is closely related to how personal goals are prioritised and indicates the kind of life sought and, relatedly, the kind of person he or she would like to be. The im portant issue for readiness is that an individual's personal identity m ust allow for the possibility of change and of an offence-free lifestyle and is not based too strongly on being an offender. Examples: 'I am a villain, not a patient'; 'I am not som eone with mental problem s'; 'M en don't talk about feelings'; 'I am not the sort of person w ho does therapy'.

151

Transitions to Better Lives

External readiness factors influencing treatment completion and engagement Circumstances: N on-engagem ent or drop-out is attributed to som e aspect of external circum stances. Examples: T h e facilitator (therapist) left the unit, w ithout a replacem ent'; 'therapy clashed w ith urgent m edical appointm ents'. Location: N on-engagem ent is attributed to therapy not being available in the setting. Example: 'he moved to another ward, w hich m ade attending the program m e im possible'. Opportunity: W ishes to engage in treatm ent but the opportunity is not available. Example: 'accepted for violence group, but its not available till next year'. Resources: W ishes to engage but attendance is im possible for resource reasons. Example: 'therapy was cancelled because a room was no longer available'. Multi-disciplinary team (M DT) exclusion from therapy: This applies where a patient has been excluded from therapy as a result of an external decision. Example: 'Patient behaved inappropriately in therapy group', 'Failed to follow group rules'; 'Perceived by clinicians as having too m any active treatm ents'. Transfer to another facility: Patient is transferred to a p riso n /o th er forensic setting, m aking treatm ent continuation im possible. Support: Perceived lack of encouragem ent and support from family, friends, fellow patients or staff. Examples: 'M y father kept telling me not to participate'; 'O ther people on the ward think this therapy is daft'. Programme: Low engagem ent is perceived as caused by som e deficit in the program m e itself (m ore than w ithin the patient). Examples: 'I am in the substance m isuse program m e, but I have never drunk alcohol or taken drugs'; 'This patient has brain dam age and is unable to cope w ith D BT' (overlaps w ith Behavioural factors above). The reasons for drop-out proved to be straightforw ard and reliable to rate, though some statem ents m et the criteria for more than one 152

Readiness and treatment engagement in personality disordered offenders

category: for exam ple, cognitive, affective, and volitional. Cognitive factors form ed the m ost endorsed category, com prising negative beliefs as to self-efficacy (T am terrified of failing again'); negative evaluations, particularly low trust, o f therapy staff ('I am misled by the professionals w ho run the groups'); as well as negative evaluations of fellow therapy participants ('H e will not go to D ialectical Behaviour Therapy any more as [another patient] is in that therapy and he does not w ant to spend time w ith him '). In this study external, situational factors were recorded, though less frequently than intrapersonal factors. Exclusion of the patient from the therapy by the responsible clinical team was a com m on external reason, often because of inappropriate or therapy-interfering behaviour in the therapy group by the patient.

An integrated multi-level strategy for improving readiness and engagement in a personality disordered offender service Little is know n, as yet, about w hat a m ulti-level and com prehensive approach m ight look like and how effective it m ight be, though this is the focus of a long-term project, funded by the N ational Institute of H ealth Research (CLA H RC) at the Institute of M ental H ealth at N ottingham U niversity in the United Kingdom . It is proposed here that a prospective strategy m ight have the follow ing com ponents, reflecting the M O RM m odel as applied at the individual, program m e and organisational levels. (a) Identify the extent and nature o f the problem.

The first step in a strategy in a clinical service is clearly to establish w hat is the extent of engagem ent and non-engagem ent in the therapeutic program m es that are delivered. For this to occur there needs to be in place an auditing system to m onitor engagem ent and non-engagem ent. It is the experience of the author that form al system s for recording and auditing engagem ent, including program m e com pletion, are not universally present in services. (b) Determine in what contexts non-engagement is evident

Treatment program m es for personality disordered offenders may be m ultifaceted, including treatm ents focused on personality traits and disorder and also on associated patterns of offending such as sexual offending, violent offending or substance misuse-related 153

Transitions to Better Lives

offending (H ogue et al. 2007; Tennant and H ow ells 2010). In addition to these psychological therapies, services often include occupational, educational and other program m es intended to facilitate a 'good life' (Ward and Brown 2004). The question thus arises w hether an individual's engagem ent pattern is consistent across these various therapeutic program m es. A sim ple correlation coefficient based on the association betw een rated levels of engagem ent for a group of patients in different therapy contexts is likely to be useful. Table 9.1 provides illustrative data of this sort for patients in a high-security unit for personality disordered offenders. A t a service level it is also desirable to establish w hich program m e w ithin the system is m ost problem atic in terms of low engagem ent. A patient m ay have full engagem ent in occupational and educational program m es but show m inim al participation in form al psychological therapies. Such inform ation illum inates w hether non-engagem ent has trait aspects or is situationally dependent. Evidence that engagem ent is good in one area but not in others may suggest hypotheses as to w hat program m e features need to be present for engagem ent to occur and also prom otes creative thinking as to how generalisation m ay be encouraged. (c) Identify what form non-engagement takes

N on-engagem ent can take m any form s, including refusal to attend sessions, failure to com plete the required program m e sessions ('dropTable 9.1

Engagement in therapy contexts

Ratings were highly positively correlated (rho, p < 0.001), suggesting that patients who engage well in one environm ent/activity lend lo engage well in all contexts. Named nurse sessions Named nurse sessions Day care/O T Therapy groups Social learning groups Individual sessions

Day care

Therapy groups

.71 .65 .80

.65 .87

Social Individual learning counselling groups sessions

_ .68 .73 .63 .81

.67

-

Note: We would like to thank Dr Kerry Sheldon and Ms Gita Patel for their help in collecting this data. 154

Readiness and treatment engagement in personality disordered offenders

out'), treatm ent-interfering behaviour w ithin sessions, the breaking of therapy rules, failure to com plete required therapeutic tasks, such as betw een-session hom ew ork, or failure to form a therapeutic alliance (Drieschner et al. 2004; Drieschner and Boom sm a 2008). (d) Work at the individual level

The assessm ent of the individual patient or offender in terms of intrapersonal factors relevant to readiness and engagem ent is a necessary part of any strategy. A range of possible assessm ent m ethods exist (see Chapter 5). The purpose of such an assessm ent is to inform a psychological form ulation o f the readiness and engagem ent needs of the individual. A theoretical fram ew ork or m odel, derived from previous studies, is necessary for devising a form ulation, as is the case in form ulating individuals' clinical problem s them selves (Sturm ey 2009). The M ORM model (Ward et al. 2004b; Sheldon et al., subm itted) is an exam ple of such a m odel, as discussed in earlier sections of this chapter. The non-engagem ent of the person in therapy can be mapped and form ulated in terms of such factors, to produce a dynamic readiness profile. For exam ple, an individual's drop-out from treatm ent m ight be explained in terms of their negative self-efficacy beliefs ('I can't cope w ith the group treatm ent setting, I can 't get the w ords out'), his perceived lack of support for engagem ent by significant others and his difficulties in trusting therapists ('they are out to hum iliate m e, not help m e'). (e) Implement individual level interventions

Such a form ulation and profile would directly suggest individualised intervention strategies. Clinical nursing staff, for exam ple, might develop a program m e to im prove self-efficacy beliefs, either through cognitive therapy to shift w hat may be an inaccurate self-appraisal or by providing behavioural training to im prove group therapy-related skills. Strategies to im prove trust in staff are likely to be long term and to require the build-up of trust and a positive relationship through the positive interactions in the course o f everyday life and experience of the general m ilieu, rather than a form al, tim e-lim ited 'therapy'. The shifting of unhelpful negative beliefs, in the setting of an institution or residential unit m ay require an interdisciplinary and consistent strategy, so that such beliefs are underm ined through educational, occupational, nursing, social w ork, psychiatric and psychological staff, w orking in unison and with a shared understanding of what is 155

Transitions to Better Lives

to be targeted and achieved. A lthough an individualised, form ulationdriven approach m ay be desirable, there is clearly also room for group interventions such as m otivational interview ing (M cM urran 2002) when it is congruent w ith the form ulation made. An individual-level and m ultidisciplinary form ulation of engagem ent problem s allow’s for creative, tailored approaches to enhancing engagem ent. A person w ho has never achieved any w'orking alliance w ith a mental health professional, about w'hom they m ay be suspicious and hostile, m ay be enticed into an alliance through w orking productively on educational or other problem s they them selves acknow ledge (Jackson et al. 2010). Equally, an enthusiasm for m usic, art, cooking, sport or other occupational activities can form the first step on a path leading to positive relationships with staff and to the beginnings of engagem ent in form al psychological therapies. D eveloping readiness and engagem ent is a m ilieu task, rather than one addressed only in circum scribed therapy sessions (Burrow es and N eeds 2009). The latter authors also draw attention to the im portance of the 'catalyst for change'. The individual, the environm ent and the catalyst are likely to interact rather than be independent influences. A broad strategy would need to incorporate the situational factors identified by M O RM and the barriers to change discussed by Burrow es and N eeds (2009). (f) Work at the programme level

Given the argum ent (above) that non-engagem ent m ay reflect the inappropriateness ('unreadiness') of the therapeutic program m e for the individual, w hen the Responsivity principle has been violated, any strategy needs to involve scrutinising the therapeutic program m es being delivered. The critical question is w hether the program m e has been adequately adapted to meet the particular needs and characteristics of the personality disordered offender. There is a danger that treatment program m es for this population m ay fall betw een two stools, in that PDOs may differ from personality disordered people w ho do not offend but also differ from offenders w ho do not suffer from a personality disorder. The therapy program m es on offer have typically been derived either from generic treatm ents for personality disorder or from offence-focused (crim inogenic) program m es (H ogue et al. 2007). How offence-focused program m es m ight need to be adapted has only recently begun to be addressed (Jones 2010; Sainsbury 2010). A strategy to im prove readiness and engagem ent in a service would need to give priority to such program m e factors. At

156

Readiness and treatment engagement in personality disordered offenders

a research level, there is a need to form ally com pare adapted and non-adapted interventions in term s o f their relative effectiveness. (g) Work at the institutional or organisational level

The prevailing culture, clim ate and modus operandi of a service would need to be addressed in a com prehensive strategy, or at least assessed to ensure that there are not factors present that m ight w ork against readiness and engagem ent. A s suggested above, negative beliefs about treatm ent on the part o f individual patients or offenders could reflect correspondingly negative beliefs on the part of clinicians and service m anagers, w ho may subscribe to view s that such individuals are not treatable. Staff education is thus a potential strategy for positive change. Explicit and enthusiastic support for, and prioritising of, engagem ent in therapy at a policy level by senior clinicians and m anagers of the service is im portant, as is the clear identification and reinforcem ent of individuals (staff and patients) w ho are potential m odels for good practice. Therapeutic clim ate (above) needs to be routinely assessed and addressed when found to be deficient. Finally, the structural context of a service m ay be vitally im portant (how and w hy individuals are referred; relationship to funders and external stakeholders). Control over such factors m ay be lim ited, but a proper analysis and intervention strategy requires that these broader factors receive consideration. In conclusion, dealing w ith problem s related to low treatm ent readiness and engagem ent are fundam ental and as yet largely unaddressed challenges for services for the personality disordered offender. The viability of and public and governm ental support for innovative treatment program m es for PDOs w ithin forensic mental health and crim inal justice services will depend on our capacity for rigorous analysis and creative solutions for the problem s identified in this chapter.

157

Part T h re e

Clinical a nd Therapeutic Approaches to W orking with Low Levels of Readiness

Chapter 10

The modification o f low readiness

In this book we have understood treatm ent readiness as the presence of characteristics (states or dispositions) w ithin either the client or the therapeutic situation that are likely to prom ote engagem ent in therapy, and thereby are likely to enhance therapeutic change. According to this definition, behaviour change in relation to persistent offending behaviour requires the presence of certain internal and external readiness conditions. It follow s from this that w hen w orking w ith those offenders w ho are assessed as 'not ready' to change (and w ho may have previously been regarded as 'resistant', 'u n b eatab le' or 'challenging'), a useful starting point will be to identify those internal and external conditions that are required for engagem ent in a rehabilitation program m e to occur, and then m odify these as required. Given that our definition o f readiness incorporates client, program m e and setting factors, increasing readiness can occur by m odifying any or all of these factors. In this chapter we discuss each of these in turn, and then review the evidence for one of the m ost w idely used approaches to im proving levels of offender motivation: m otivational interviewing.

Modifying the setting M aking changes to the setting in which rehabilitation program m es are offered often requires m edium - to long-term planning. Setting factors can include the physical and social environm ent in which

161

Transitions to Better Lives

program m es are offered, and these are often hard to influence. Such term s are used broad ly to refer to the extent to w hich a particular unit or institution is likely to be con du cive to therapeutic change. Investigation s into the experiences o f those w ho are held in secure facilities reveal that they typically experience a nu m ber of barriers to change in their environm en ts, and express a range o f concerns. These inclu de a subjective sense o f failure, p ow erlessness, and the im pairm en t of social identity, con cern s about surv eillance and the over-regu lation o f their behaviou r, and w'orries for their personal safety (Q uirk and L elliot 2002; Toch and A dam s 2002; Zam ble and P orporino 1990). Such exp eriences are lik ely to be particu larly salient and intense for prisoners, given the ad ditional con straints on their behav iou r and actuality o f close d aily m onitoring that is a defining feature o f the correctional environm ent. In ad dition, the prevalence o f aggression and self-harm in prisons suggests that con cern s about personal safety m ay be w ell founded. The m ost relevan t body o f scientific w ork for understand ing, m easuring and m od ifying negative therapeu tic environm en ts and m ilieu x is that on social clim ate, pioneered and used for over 30 years by R u dolph M oos (M oos 1997) in a range of service settings, inclu d ing health, m ental health and correctional institutions. O ne o f M o os's m ost im portant con tribution s has been the Ward A tm osphere Scale (available in different versions for different settings, inclu ding p risons), a 100-item scale that p u rports to m easure ten aspects of the social clim ate of a unit or institution and is com pleted by staff and patients. In recent years, the M o os scales have been subject to a nu m ber o f critiqu es (for an overview see Sch alast et al. 2008; see also C hapter 9) id entifying problem s associated w ith outdated item content, the tim e and effort required for com p letio n in d isturbed and unm otivated popu lations, the length of the m easures for repeated clinical use, the low internal con sisten cy o f som e scales, and d iscrepancies betw een the scales and the factor structure o f the m easure. T his has prom pted new scales to be validated (Sch alast et al. 2008) that allow for repeated ad m inistration such that chang es in clim ate over tim e can b e assessed . It is only then that the effects o f any attem p t to change the social clim ate o f prison u n its can be ju dged . These m ight include, for exam ple, u sing the sentence p lanning process to m ove offend ers to less secure environm en ts to receive program m es, or to prom ote the active involvem ent of prison officers in program m es as a m eans o f d ev elop in g a m ore therapeu tic cu lture in w hich pro-social behav iou r is b o th m odelled and reinforced o utsid e of the therapeu tic session. D eveloping m ore treatm ent supportive settings through training 162

The modification of low readiness

staff in the value of program m es is another obvious w ay in which readiness m ight be improved. There are also a num ber of organisational and adm inistrative measures that m ight be used to m odify the setting. For exam ple, in relation to the tim ing of program m es, lower security prisoners or those nearing com pletion of their sentences m ight be given the option of attending com m unity-based treatm ent program m es while on day leave. System ic issues also appear to exert a great influence on treatm ent outcom es in com m unity-based program m es. Day and Carson (2009) have observed that court referral and adm inistration by correctional services to intervention program m es for dom estically violent men, for exam ple, often takes place after som e time (som etim es up to a year or m ore) after the offence occurred. Such delays can cause offenders to question the relevance of program m es, leading som e to claim that they no longer have the need to attend, or that they have a poor recollection of the circum stances of the offence. In addition, the pow ers to legally enforce an offend er's program m e attendance can at times be inconsistently applied by both the courts and probation and parole officers, and in m any cases non-attendance leads to few (or lenient) statutory consequences. For offenders to be ready to receive com m unity-based program m es they need to be aware that the crim inal ju stice system is seeking particular outcom es from their attendance and participation, and that the consequences for non-attendance will be uniform ly applied.

Modify the programme A nother option, and one that is identified m ost explicitly in the responsivity principle, is to m odify the treatm ent or program m e that is offered. This is discussed in C hapter 6 in relation to the treatment of violent offenders, and involves am ending treatm ent m ethods and program m e content such that it is better suited to the needs of the particular client group. This may be in term s of the literacy level required to participate m eaningfully in the program m e, or the cultural or gender appropriateness of the program m e m aterials. Given that m any offenders will arrive in rehabilitation program m es w ith low levels of readiness, it is particularly im portant that the content of early sessions is appropriate for these presentations. O f course, this is alm ost im possible to achieve in program m es that have rolling intakes (where new m em bers join an existing group as they are referred), as is the case in m any sexual offender and dom estic violence program m es. 163

Transitions to Better Lives

M odifying the program m e may also involve changing structural com ponents of the program m e itself, such as the num ber of sessions offered. Some offenders may not be w illing or indeed able to com m it to a program m e that lasts over a year (or a very intensive treatm ent such as that offered in therapeutic com m unities), but may be w illing to engage in program m es that require less com m itm ent. A lthough readiness for treatm ent is likely to be reasonably consistent across m ost treatm ent m odalities, there m ay be som e occasions where readiness factors differ for different types of treatm ent. For exam ple, it has been suggested that clients w ho have low levels of distress m ay be better suited to m ore prescriptive interventions (Beutler et al. 2000), while those clients w ho have high levels of distress or specific needs m ay be better suited to individual m ethods of delivery rather than group-based approaches. Different types of rehabilitation program m e tend to adopt different approaches to the m anagem ent of therapeutic engagem ent. The treatment of sexual offenders is relevant here, as it is relatively com m on for sexual offenders to enter treatm ent m aintaining either that their offending did not happen or that it was not problem atic. Low levels of problem recognition and m otivation would in som e program m es (such as in some substanceuse program m es) exclude offenders from participation. In sexual offender treatm ent, however, denial, unless extrem e, is not alw ays grounds for exclusion (see Chapter 7). In our view, the early stages of these program m es should not be considered as treatm ent, but m ore as a preparatory stage of treatm ent where the prim ary task is to increase readiness. The w ay in w hich the program m e is delivered and the extent to w hich program m e facilitators are able to respond on a m om ent-by-m om ent basis to the changing needs of offenders will also be critical in both the successful form ation and the m aintenance of a strong therapeutic alliance (see Chapter 12). This is a skilled task, even in program m es that are predom inantly psycho-educational in nature. Theories of behaviour change offer useful fram ew orks for guiding decisions about the appropriateness of particular approaches for individuals w ho are at different points in the process of changing their behaviour. A lthough a num ber of m odels of how therapeutic change occurs have been developed, the transtheoretical model of change (TTM ) is probably the m ostly w idely researched and clinically useful m odels of behaviour change that exists. The TTM suggests that problem resolution typically occurs follow ing a progression through a sequence of change stages, each one characterised by different attitudes, thoughts, beliefs and values. The individual 164

The modification of low readiness

typically experiences a grow ing aw areness of the problem , form ulates a decision to do som ething differently, develops change strategies w hile in a transitional phase, and finally im plem ents those strategies (see Casey et al. 2005). The key organising construct o f the TM M is the notion of stages (Velicer et al. 1998). Irrespective of w hether som eone is in or outside form al treatm ent (and for virtually any type of problem behaviour), behaviour change is thought to occur in a series of six identifiable stages: precontem plation (no w ish to ch a n g e/n o recognition of a problem ); contem plation (intention to change problem behaviour within the next six m onths); preparation (intention to take im m ediate action, usually m easured as w ithin the next m onth); action (characterised by specific, overt m odifications w ithin the past six m onths); m aintenance (relapse prevention); and term ination (change process is co m p lete/n o further need to prevent relapse). O ptim al progress through each stage is achieved via different processes of change (which m ay be either overt or covert).1 The term 'process' refers to w hat an individual does to bring about change in affect, behaviour, cognitions or relationships (Prochaska et al. 1988). According to Prochaska and DiClem ente (1986), therapeutic interventions should be guided by the processes deem ed m ost appropriate to a particular stage, to ensure progression from one stage to the next. The Levels of Change dim ension of the TTM integrates the various processes and stages o f change w ithin the context of five interrelated but distinct levels of psychological problem s that can be addressed in treatment: sym p tom /situational problem s; m aladaptive cognitions; current interpersonal conflicts; fam ily system conflicts; and intrapersonal conflicts (Prochaska and DiClem ente 1984, 1992). A gain this is useful in terms of how to m odify program m es w hen offenders present w ith low levels of readiness as it helps to place the level at which different interventions are offered, and draw attention to areas that m ay be overlooked (for exam ple, program m e content may focus too heavily on m aladaptive thinking styles, and underestim ate the influence of fam ily system s on behaviour change). A dopting a transtheoretical approach also requires that both therapist and client ‘Ten processes have received the greatest empirical support in the literature (Prochaska and DiClemente 1983; Prochaska et al. 1988). Of these processes, five are experiential (consciousness raising, dramatic relief, environmental reevaluation, social liberation, and self re-evaluation), and five are behavioural (stimulus control, helping relationships, counter-conditioning, reinforcement management, and self-liberation). 165

Transitions to Better Lives

agree (at least im plicitly) about w hich level an identified problem will be attributed to, and at which level both client and therapist will w ork to change the problem behaviour. According to Prochaska and DiClem ente (1986), the trend in psychological therapies has been to attribute psychological problem s to one or two levels, with intervention subsequently focused on these levels (for exam ple, behaviourists focus on the sym ptom and situational determ inants, cognitive therapists on m aladaptive cognitions, family therapists at the fam ily /sy stem s level, and psychoanalytic therapists focus on intrapersonal conflicts). The assim ilation m odel (A M ) (Stiles et al. 1991; Stiles 2000) is an alternative model of problem resolution, developed prim arily on the basis of observations made across a series of intensive psychotherapy case studies. According to this m odel, 'therapeutic progress consists of the assim ilation of problem atic experiences into the client's schem ata' (H onos-W ebb and Stiles 2002: 407). The model describes the likely needs of an individual client at each stage of assim ilation. For exam ple, at Stages 0 -2 the problem atic experience is largely outside of the client's consciousness. Stages 3 - i are characterised by an ability to acknow ledge the existence of the problem and to being able to com m unicate a clear statem ent of the problem , and stages 5 -7 represent stages in which the problem is solved and mastered. In addition, the assim ilation m odel suggests that as individuals pass through the stages of assim ilation, they experience a corresponding sequence of em otional reactions, w ith psychological distress increasing as a function of problem aw areness (stages 0 -3 ) and then decreasing w ith problem clarification, solving and solution (stages 4 -7 ) (Stiles et al. 1991). A particularly im portant idea in both the transtheoretical and assim ilation m odels is that interventions should be matched to the individual's stage of change. The term 'm atching' is com m only used to describe strategies that tailor treatm ent to clients' baseline characteristics, w hereas the term 'ad aptive' is com m only used for strategies that continually readjust the interventions during the course of treatm ent (see M arlow e et al. 2007). It follow s that the early m odules of the rehabilitation program m es should reflect content that aim s to increase both self-aw areness and problem aw areness in w ays that m ay be particularly appropriate for those participants in the earlier stages of change. It is im portant that facilitators also attend to the needs of those in the later stages, w ho m ay benefit from m ore skills-based approaches. Thus, an individual w ho is pre­ contem plating change m ight be assisted to develop greater insight 166

The modification of low readiness

into his or her offending and encouraged to think carefully about any decision to (or not to) change. The notion that client needs can change over time suggests that different therapeutic tasks for clients may be required. For exam ple, the therapeutic task for clients at the early stages of assim ilation is to increase problem awareness. Low problem aw areness is another w ay of describing w hat has been referred to as 'd enial', although this term can have different m eanings, ranging from denial of guilt, denial o f responsibility, denial of victim impact, and so on. It also suggests that both program m e content and delivery styles should reflect these needs. For exam ple, problem aw areness exercises, such as victim im pact work, and more experiential m ethods of delivery (such as role play and group discussion) m ay be m ost valuable in the earlier stages of a program m e. As the problem enters aw areness, the requirem ent then m oves to facilitating em otional expression. O nce the problem has been recognised (from stage 3 onw ards), the task is to develop a clearer understanding of the onset, developm ent and m aintenance of the problem , before w orking actively towards solution. Thus, it is suggested that an appropriate therapeutic response is one that m eets client requirem ents at a given stage of assim ilation, such that the offender progresses to the next stage of change.

Modifying the offender For practitioners, attem pting to m odify the individual client often represents the m ost realistic and achievable w ay of changing levels of treatm ent readiness. Those offender characteristics m ost closely associated w ith readiness fall into three dom ains: the cognitive, the affective, and the behavioural. There are a num ber of different w ays in which these individual low readiness factors can be addressed, and indeed m any of the approaches and interventions described below (and elsew here in this book) are already used inform ally in rehabilitation program m es. Less com monly, however, are they conceptualised and delivered in relation to the need to im prove pre­ program m e levels of treatm ent readiness. Responding to feelings o f coercion

H ow ells and D ay (2003) have suggested that for therapeutic goals to be realised, client personal goals need to be constituted by, supportive of, or consistent w ith, the therapeutic goal. This is a theme

167

Transitions to Better Lives

of Chapters 3 and 10 of this book. Therapeutic change, therefore, involves the therapist exam ining the goals that the client is pursuing, how these goals are organised, and how they are being regulated or mis-regulated (Karoly 1999). Karoly (1993) further stresses the need to exam ine treatm ent targets in the context of broader client goals and the m otivational salience of change. H e suggests, also, that 'therapeutic failures of various kinds (prem ature term ination, resistance, relapse etc) can result from the therapist-assessor's failure to appreciate the structural relation betw een tim e-lim ited treatm ent goals and life goals in general' (1993: 279). Given that m any clients m ay not, at least in the early stages of intervention, have personal goals that are consistent with program m e goals (that is, com m unity safety), or have goals that are incom patible with program m e goals (for exam ple, to fulfil the obligations of the order rather than change behaviour), it is perhaps unsurprising that program m e facilitators com m only find difficulties in engaging clients in a change process, and often report encountering hostility, resistance and difficulties in engaging men in therapeutic change. Barber (1991) has discussed case work with the 'involuntary client' from a social w ork perspective, m aking the follow ing observation: Work with involuntary clients m ust begin with the recognition that the interaction betw een the w orker and client is based on conflict rather than co-operation, that social w ork with involuntary clients is a political, not a therapeutic, process involving the socially sanctioned use o f power. The political nature of this activity becom es obvious w hen one calls to mind w hat it m eans to be an involuntary client in the first place. (Barber 1991: 45) H e goes on to note that the 'ultim ate beneficiary of the w ork done will be those w ho in som e w ay suffered from the client's aberrant behaviour in the first place' (1991: 45). Thus for Barber at least the role of the w orker is to negotiate a settlem ent betw een the client and society at large, and it is im portant that the client is aware of this from the very outset of any intervention. Barber proposed a sixstep model of w hat he termed 'negotiated casew ork' (see Table 10.1 below ). An advantage of this approach is that it clarifies both the goals of the program m e and the responsibilities of the practitioner to other agencies, particularly in regard to reporting or evaluating clients for legal purposes.

168

The modification of low readiness

Table 10.1

Negotiated casework

Steps

Description

1 Clear the air.

Begin by directing attention to the order that led to the meetings. Read out a copy of the court order, and elicit client perspectives.

2 Identify legitimate client interests.

Attend to any objections to engagement in treatment and the goals of the client.

3 Identify non-negotiable aspects of intervention.

Clarify w hich aspects cannot be compromised and the reporting requirements (e.g. attendance).

4 Identify negotiable aspects of intervention.

Define the problem and identify possibilities for change.

5 Negotiate the case plan.

The aim here is to make decisions about the way forward, identifying goals and responsibilities.

6 Agree on criteria for progress.

Clarify how judgements will be made about whether the case plan is working or not, and what happens if the client fails to comply with aspects of the case plan.

Source: Adapted from Barber (1991).

We have suggested elsew here in this book (Chapter 2) that coercing offenders into attending rehabilitation program m es (or placing legal pressure on them to attend) is unlikely in itself to lead to poorer outcom es. Rather, we suggest, it is the perception of coercion that will determ ine how an offender approaches treatm ent. Even when offenders perceive they are being coerced, it also seem s likely that pre-treatm ent anti-therapeutic attitudes can change over the course of a program m e, such that therapeutic gains (risk reduction) can occur. That is not to say, however, that the needs of those w ho perceive coercion are the same as those who do not. Engaging coerced clients in treatm ent is a task that requires great therapeutic skill. O ffenders w ho feel coerced m ay arrive in treatm ent w ith high levels of antipathy tow ards both program m es and program m e providers. The w ay in w hich therapists respond to this hostility is likely to be a critical factor in w hether perceptions of coercion dissipate over

169

Transitions to Better Lives

time. In these circum stances, treatm ent engagem ent can be enhanced through the provision of accurate inform ation about an individual's legal obligations and consequences o f non-participation, as well as by acknow ledging explicitly the extent to w hich the therapist is w orking for the com m unity as well as for the individual. Offenders in rehabilitation correctly perceive that pressure is exerted to m ake them engage in program m es but also seem to accept such pressure as 'fa ir' w hen it is m ade clear that it is their decision to accept or refuse treatm ent (Rigg 2002). Reducing the reality and the perception of excessive and 'u n fair' coercion should be an im portant objective in rehabilitation, for both ethical and practical reasons. Similarly, Cosyns (1999) argues that to m axim ise treatm ent efficacy in situations of coercion, an agreem ent of m utual trust and com m on goals should exist betw een the therapist and coercing party. Both parties should hold the client's best interests param ount. It seem s that em phasising the negative consequences of non-participation, threatening, and using other form s of coercion such as close m onitoring and severe penalties are likely to be less effective than providing accurate and honest inform ation about an offend er's circum stances2 (M axw ell 2000; Young 2002). There has been m uch less discussion of the possible role of positive reinforcement for attending treatment, although the introduction of treatment-oriented diversion program m es for m entally disordered a n d /o r substance-using offenders represents a significant developm ent in this area (M urphy 2000). Potentially, there are a number of ways that programm e providers in correctional settings can offer incentives for participation. For exam ple, these may include offering salaries to those w ho attend program m es that are equivalent or greater to those offered at prison workplaces, or additional visits from fam ily and friends. Those offenders for w hom program m e participation is highly aversive m ay need to be treated differently. It would seem untherapeutic at best, and unethical at worst, to enforce a treatment that was likely to traum atise a client, and it seem s unlikely that these clients would benefit from the treatment. There is little point in enforcing treatm ent that is not going to w ork, and endorsing such an approach would seem to use psychological treatm ent as a form of punishm ent. For the m ajority of offenders w ho feel coerced into treatm ent, however, we w ould argue that interventions to im prove treatm ent readiness are likely to be helpful. C o ercion is distinguished from pressure partly by its exclusive focus on negative consequences or punishment as a means of ensuring compliance. 170

The modification of low readiness

Problem recognition and m otivation to change are im portant conditions of treatm ent readiness (see Ward et al. 2004b) and interventions such as m otivational interview ing (see below ) are thought to be useful in im proving m otivation to attend treatment. Exposure to group treatm ent in the form of less intensive psychoeducational program m es m ay also help to allay anxiety in those w ho are particularly apprehensive about treatment. Similarly, particularly distressed offenders may benefit from learning som e life skills in a less coercive 'offence-related' program m e prior to participation in 'offence-specific' program m es. Cognitive factors

We have argued that in the cognitive dom ain, some degree of problem recognition is likely to be an im portant readiness factor, along w ith confidence in the types of program m es and services provided. In Chapter 5 we noted that violent offenders have often been regarded as 'resistant' and unsuitable for treatm ent, m ainly because of their tendency to deny or m inim ise their abusive behaviour. Instead, they tend to blam e others for their problem s or justify their use of violence as a reasonable response to provocation. Violent offenders can also be extrem ely m istrustful, paranoid and suspicious of entering into a therapeutic relationship where they fear they m ay be perceived as vulnerable. Such negative attitudes and beliefs are likely to im pede their readiness to engage in program m es and offenders need to be given feedback that their hostility, w hile understandable, is inappropriate (see Cham bers et al. 2008). N egative attitudes about program m es a n d /o r program m e providers m ay be able to be addressed through giving offenders access to inform ation about what program m es actually involve (through videos or even docum entary film), or contact with other offenders w ho have previously participated successfully in the programm e. Affective factors

In the affective dom ain, som e level of general distress (including anxiety and depression, guilt or rem orse) is com m only identified as a potentially im portant readiness factor. One im plication of the w ork described in Chapter 3 is that there is a need for rehabilitation program m e providers to pay attention to the assessm ent and m odification of the affective states o f their clients if they are to be successfully treated. W hile this may seem obvious to mental health practitioners where the focus is on the individual w ell-being of the 171

Transitions to Better Lives

client, it m ay be less so for those w orking in correctional settings w ho are aware of the em pirical evidence that psychological discom fort per se and anxiety are unlikely to be risk factors for reoffending (Bonta 1997), and thus not appropriate targets for intervention. Individualised assessm ents, such as those proposed by Daffern and H ow ells (2002) are, in our view, necessary to help determ ine not only the initial treatm ent needs of the offender (and the extent to which these differ from identified crim inogenic needs) but also guide decisions about the com position of treatm ent groups. For som e, specific interventions to im prove treatm ent readiness, such as m otivational interview ing (see below ) m ay be particularly helpful in reducing anxiety about program m e participation. Day, Tucker and H ow ells (2004) also suggest that exposure to group treatment in the form of less intensive psycho-educational program m es may also help to allay anxiety in those w ho are particularly apprehensive about treatment. Victim aw areness program m es (see Day et al., in press) m ay serve to increase levels of rem orse and guilt, such that offenders are m ore motivated to seek help to address the causes of their offending behaviour. O ther offenders m ay need to learn how to recognise and respond appropriately to their em otional states. For those w ho experience too m uch negative affect, m ental health interventions (including pharm acotherapy and counselling) may help an offender contain high levels of arousal to a level that he or she can still engage in treatment. Peer education and m entoring schem es, or access to other services (such as the chaplaincy or culturally-based groups) can also be im portant here as they have the potential to im prove levels of social support and thereby encourage a person to engage in and sustain treatm ent over time. M any rehabilitation program m es respond to anxiety about participation by starting with an extensive discussion about the lim its of confidentiality and the agreeing of group rules that are likely to facilitate disclosure. It is, however, the skills of the facilitators in developing a safe environm ent that are likely to be the most influential. W hile facilitator authenticity (including such things as directness, acceptance, empathy, respect, self-disclosure, nonjudgem ental attitude) is likely to be im portant in overcom ing client resistance (M ilgram and Rubin 1992), it is also apparent that som e clients will have greater problem s in trusting the facilitator and the group process than others and this will require a higher level of skill from the facilitator.

172

The modification of low readiness

Low motivation

Low m otivation m ay be addressed using problem recognition and decision-m aking techniques often used in m otivational interview ing (M iller and Rollnick 1991). M otivational interview ing (MI) has also been designed to offer clients a relatively safe and non-threatening introduction to therapy, such that any anxieties about entering treatment are allayed. H emphill and Hart (2002) have discussed w ays of w orking w ith psychopathic offenders to increase their m otivation. They identify four m otivational strengths of psychopaths, associated w ith a need to feel superior to others, a desire for and tolerance of novelty, good interpersonal skills, and a desire to be in control. They propose, am ong other things, that interventions that suggest that crim inal lifestyles have low status, that help offenders feel in control of their treatm ent and em phasise self-sufficiency, can all help to m otivate the psychopathic offender to engage in treatment. Brief interventions are often used to increase readiness to change and strengthen self-efficacy (M cM urran 2002). There are two characteristics of a brief intervention that exert the greatest influence on readiness to change: the therapeutic style of the program m e staff; and the extent to w hich the client receives personal feedback. M otivational interview ing is one exam ple of a brief treatm ent approach that is used to increase problem recognition and the probability of treatment entry. MI has been described as 'a client-centered, directive method for enhancing intrinsic m otivation to change by exploring and resolving am bivalence' (M iller and Rollnick 2002: 25). This form of intervention is a collaborative method where the therapist finds the potential for change w ithin the client, and facilitates the change process. It thus has the potential to enhance engagem ent, reduce reactance and increase readiness for treatment. M otivational interview ing generally involves som e discussion regarding the argum ents for and against the individual changing their substance-use behaviour (decisional balance), feedback about the risks and benefits of continued substance use, and a description of the available treatm ent options (see Table 10.2). Brief interventions based on MI typically involve exam ining the costs and benefits of change, identifying high-risk situations associated with substance use, discussing life goals and how substance use affects these, and learning about the stages of change concept. MI is thought to be particularly useful at the beginning of treatm ent, providing the foundations of a level of m otivation to engage in therapy and change.

173

Transitions to B e tte r Lives

Table 10.2

Motivational interviewing

Component

Description

Express em pathy

A crucial attitude for therapists to exh ibit w hen w orking w ith violent offenders is one of acceptance. This does not m ean agreem ent or approval of their anti-social attitud es or behaviou rs or labellin g the offender as, for exam ple, anti-social or deviant. Instead, skilful, respectful and reflective listening to the offender, w ith an expressed desire to understand h im /h e r, is crucial. It is im portant for the therapist to understand that an o ffen d er's am bivalence is norm al and the therapist should dem onstrate an un derstand ing o f the clien t's perspective. A cceptance and respect help to build a w orking therapeutic alliance, and support the clien t's self­ esteem and self-efficacy, w hich is an im portant condition for readiness and m otivation to change.

Develop discrepancy

This refers to building a discrepancy betw een the target b ehaviou r (e.g. violent offending behaviour) and the ind iv id u al's values, beliefs and goals. With M I the therapist has the opportunity to change the clien t's perceptions of discrepancy w ithout creating a feeling of bein g pressured or coerced. It is im portant for the therapist to clarify im portant goals for the offender and to explore the consequ ences of the clien t's past and potential violent behaviour. W hen this technique is successful the client rather than the therapist presents the reasons for change.

Avoid argum entation

Any argum ents from the therapist in response to the offen d er's denial or resistance are counter­ productive. T h e therapist should be m indful that an argum ent can breed defensiveness, and resistance from an offender should be seen as a signal to change strategies. Clearly, if dis­ agreem ents can be resolved, not only w ill appropriate b ehav iou r b e m odelled, bu t also efficacy can be gained through the therapeutic process.

Table 10.2 continues opposite

174

The m o dification o f lo w readiness

Table 10.2 continued C o m p o n en t

D escrip tio n

Roll with resistance

Therapists need to rem ain non-defensive and calm when faced with a hostile and resistant offender and avoid confrontation. In m anaging hostile clients it is important not to oppose the resistance but to understand it, learn its directions and move w ith its tensions. If resistance does emerge, the therapist should help the offender to shift h is/h e r perceptions by refram ing h is/h e r cognitive distortions to create a new momentum towards change. It is important for the therapist to em pow er the offender to find solutions to h is/ her own problems and elicit self-motivational statem ents. The therapist encourages the offender to believe that s h e /h e has the ability to change. The offender is motivated to take responsibility for choosing and undertaking personal change. The therapist should be affirm ing of the offender's strengths, w hich develops the sense of selfefficacy, an im portant foundation for sustaining change. The therapist should em phasise the value of choice - offenders are more likely to be com mitted to a freely chosen course of action, such as readiness to engage in therapy, rather than when they feel they have been coerced or pressured to do so.

Support self-efficacy

Source: Adapted from Miller and Rollnick (2002).

This is especially pertinent with offenders, for whom engagement in change is often difficult to achieve. The efficacy of motivational interviewing has been subjected to two recent meta-analytic reviews (see Burke et al. 2003; Hettema et al. 2005). Studies reviewed in both meta-analyses were all randomised control studies which, in addition to the random assignment of participants to treatment and control groups, included (a) at least one group or individual intervention with components of MI,3 (b) at

3The research literature indicates that the m ost widely used M I approach is one where the client is given feedback based on individual assessment

175

Transitions to Better Lives

least one control condition or com parison group, and (c) adequate m easurem ent pertaining to pertinent target areas. In the Burke et al. (2003) m eta-analysis, adaptations of m otivational interview ing (AMIs) across a range of problem areas (alcohol problem s, drug addiction, sm oking cessation, diet and exercise, HIV-risk behaviours) were exam ined. Twenty of the 30 controlled clinical trials were related to substance m isuse (alcohol and drugs). Similarly, of the 72 clinical trials reviewed by Hettem a et al., 31 related to alcohol abuse and 14 to drug abuse (62.5 per cent of the total studies). W hile the m ajor findings in the Burke et al. (2003) m eta-analysis w ere m ixed, there were nonetheless positive indicators in support of the use of M I as a brief intervention strategy. For exam ple, AM Is w ere found to be significantly more effective than either no-treatm ent or control conditions, with m edium effect sizes for drug addiction (rf = .56) and sm all to m edium for alcohol {d = .25 to .53), depending on the target m easure used. This contrasts w ith the near zero effect {d = .02) for A M Is w hen com pared to other active treatm ent m odalities, although it should be noted that A M Is achieved these sam e results in substantially few er treatment sessions (approxim ately three to four). In this respect, A M Is could be considered a m ore cost-effective approach than, for exam ple, cognitive behavioural therapy w ith this client group. In term s of sustained efficacy, those studies with a sufficiently long follow -up period (N = 9) showed that the effect size at 20 w eeks (rf = .13) was com parable to the average effect size of 67 w eeks (rf = .11). In other words, irrespective of the com parison group (alternative treatm ent modality, no treatm ent, or control), the effects of A M Is did not reduce significantly over time. Clinical im pact was also prom ising, with 51 per cent of clients w ho received AM I for drug and alcohol problem s (N = 7 studies) show ing either im provem ents or abstinence on m easures taken betw een four w eeks and four years post-treatm ent. In fact, 54 per cent of clients who receive a single (stand alone) AM I intervention showed noticeable im provem ent, w hile 43 per cent of A M Is used as a prelude to further

using standardised measures of drug or alcohol use. This feedback relates to the level of severity on the target symptom as compared to population norms, and is delivered in an MI style, and includes elicitation from the client of possibilities for change, all of which is done in a non-threatening manner. This feedback approach is considered an adaptation of MI (AMI), as the process includes more than just MI. The term AMI can thus be applied to interventions incorporating additional MI techniques, but which maintain the core MI principles (see Burke et al. 2003: 844). 176

The modification o f low readiness

treatm ent also im proved. By com p arison , significantly few er clients im proved or abstained from d ru g or alcohol w ithout treatm ent (38 per cent) or w ith treatm ent as usual (35 per cent). T his translates to an im provem ent in client success rates from one-third to onehalf follow ing A M I, or the d ou blin g o f abstinen ce rates from one in five to tw o in five. Finally, A M Is w ere found to have the sam e level o f effect on social im pact m easures (d = .47) as target sym ptom s, ind icatin g the positive con sequ ences of treatm ent for a broad range o f im p ortan t life problem s beyond substance-related sym ptom s. A particu larly im portan t im plication o f the Burke et al. (2003) m eta­ analysis is the ind ication that A M Is m ay in fact be m ost efficacious as treatm ent preludes. The researchers point out w hile 'it is rare that a treatm ent can be efficaciou s both as a stand-alone treatm ent and as a treatm ent used ad junctively to enhan ce the efficacy o f a variety of other treatm en ts' (2003: 858), their results indicated that this w as the case for the A M Is. And w hile they acknow led ge that m ore detailed research is necessary, it would also seem that the feedback com ponent m ay be m ore critical to the success o f A M Is than the m otivational interview ing per se. In the m ajority o f studies review ed b y H ettem a et al. (2005), M I w as seld om used alone bu t m ore often in com bination w ith either feedback or som e other form o f treatm ent (such as edu cation, selfhelp m anuals, relapse prevention, cognitive therapy, skills training, A lcoholics A nonym ous, stress m anagem ent). T he d uration of interventions ranged betw een 15 m inu tes and 12 hours (average d ose about tw o sessions; M = 2.24 h ou rs, SD = 2.15), and in the vast m ajority of studies (74 per cen t), intervention s had been standardised by either a m anual or specific training. C on sistent w ith other system atic review s, H ettem a et al. found con sid erable variability in effect sizes across stud ies, even for those w ithin specific problem areas (for exam ple, the observed effects in alcohol studies ranged from dc = 0 to m ore than 3.04). W hat this su ggests is that d espite using the sam e treatm ent m od ality w ith the sam e target p opu lation, different effects w ere obtained across both sites and across populations. The im plication here is that variatio n in M I d eliv ery can substantially im pact on outcom e. A nother observation m ade by the researchers, and one that is incon sisten t w ith the find ings of B urke et al. (2003), w as the tend ency for MI trends to dim inish o v er tim e. The com bined effects across all ■‘An effect size of d = 1 represents a betw een-group difference of one standard deviation. 177

Transitions to Better Lives

studies showed that w hereas rf = .77 at the 0 to one month point post-treatm ent, it was reduced to .39 at > 3 to 6 m onths, .30 at > 6 to 12 m onths, and .11 at follow -ups longer than 12 months. The exception to this finding w as in studies where the additive effect of MI was assessed (where MI is used at the com m encem ent of a standardised or specialised treatm ent). In these studies, the effect of MI in term s of improved outcom e was either m aintained or increased over time, generally rem aining in the vicinity of rf, = .60. Across all problem dom ains, the H ettem a et al. (2005) m eta-analysis found the strongest support for MI efficacy w as in the area of substance use. The m ean effect size across 32 trials w hich focused on alcohol abuse was .41 at post-treatm ent and .26 across all follow -up points (range -.0 8 to 3.07), with the largest effect sizes (all > .70) reported in studies which com pared MI with either (a) a no treatment condition, wait-list controls, or education, or (b) adding MI to standard treatment. W here MI was used to treat illicit drug use, the effect sizes ranged from 0 to 1.81, with effects sizes on average being larger at early rather than later follow -up (.51 versus .29). M cM urran (2009a) has recently prepared a review of M I with offenders, identifying a total o f 13 outcom e studies w ith offenders and six dissertations. M ost of these (N = 10) studies had been conducted with substance-using offenders. Although there were marked variations across studies, the conclusion o f the review was that M I can lead to im proved retention in treatm ent, m otivation to change and reduced offending, although M cM urran notes the im portance of m aintaining the integrity of treatm ent. Thus, although MI w ith offenders probably falls short o f fulfilling the criteria for an evidencebased intervention, it is an approach that is likely to be particularly appropriate for use with m any offender groups, especially when integrity is increased through the application of practice guidelines.

Conclusion In this chapter we have discussed som e of the w ays in w hich low levels of treatm ent readiness m ight be modified. This includes m odifying the setting, the program m e, and the individual offender. A num ber of different approaches are described as having potential, although few em pirical studies have exam ined the effectiveness of different approaches. M otivational interview ing is one intervention that has attracted the m ost attention from researchers, and there is som e evidence to support its efficacy w ith offenders. If attention can 178

The modification of low readiness

also be paid to the environm ents in w hich program m es are offered and the m atch betw een the needs of individual participants and program m e content and structure, then it is likely that low levels of readiness can be successfully modified such that offenders are both w illing and able to participate in program m es that seek to address the causes of their offending.

179

C h ap ter I I

Goal-focused interventions with offenders Mary McMurran

Treatment to reduce the likelihood of reoffending operates through requiring offenders to make a range of personal changes, including changing ways of thinking, communicating, behaving and socialising. Motivating offenders to make these relevant personal changes and enabling them to sustain these changes are core aspects of clinical practice. Goal perspectives have proven useful in conceptualising motivation for behaviour in general and change in therapy in particular. This approach also has application in conceptualising motivation for change in offenders. In this chapter, the focus is on goals as a motivational construct and the methods by which offenders may be encouraged to set and pursue pro-social goals.

Goal perspectives

In evolutionary terms, goals are specific representations of what is needed for survival. Maslow (1943) proposed a hierarchy of needs from basic physiological needs (oxygen, food, water), through higher order needs for safety, belonging, esteem and self-actualisation. More recently, researchers have focused on psychological rather than physiological needs, identifying primary needs for autonomy, competence, relatedness, self-transcendence and spirituality (Deci and Ryan 2000; Emmons 2005). Deci and Ryan (2000) define these as 'innate organismic necessities', as opposed to acquired motivations, and say that these needs must be fulfilled for optimal health, well­ being and interpersonal functioning. 180

Goal-focused interventions with offenders

Innate needs are satisfied through the pursuit of specific goals, which are the 'consciously accessible and personally m eaningful objectives that people pursue in their daily lives' (Em m ons 2005: 732). As Deci and Ryan (2000) point out, innate needs are w hat give specific goals their potency and they are im portant to understanding goal choice, pursuit and attainm ent. Karoly (1993) defined goals as 'states toward which people intentionally aspire and actively w ork to bring about (or to avoid)' (1993: 274). Goals are action-oriented and the person strives to attain an identifiable end point. These end points may be of many different types and levels, for instance losing w eight, passing an exam , being patient w ith one's children, and being a kind person. Goals are not stable or static and the effort put into achieving goals is variable, m aking goals a useful m otivational construct (Karoly 1993, 1999). Em m ons (2005) holds goals as central to hum an functioning. Goals are the concrete expression of a person's life purpose, they represent how people structure their lives, and striving towards w orthw hile goals is a key determ inant of hum an health and happiness. In short, a person's goals are synonym ous w ith m eaning in life, since they are the route by w hich the individual fulfils his or her needs for com petence, relatedness, self-transcendence and spirituality. Goal strivings or personal concerns have been identified as the m iddle of three units of personality, w here the first is basic tendencies (what one has), the m iddle is goal strivings (w hat one does), and the third is personal narratives (who one is) (M cA dam s 1995). Goal-based psychological therapies focus upon a person's life goals and goal strivings, attending to the number, range and content of goals a person m ay be pursuing, the attainability of the goals, the likelihood of satisfaction from goal attainm ent, and w hether goals conflict or cohere (M ichalak and Grosse H oltforth 2006). The first step in this therapeutic approach is to define and identify life goals.

Assessing human goals O ne theory of m otivation in w hich goal-striving plays a central role is the Theory of Current Concerns, developed originally to understand problem drinking (Klinger and Cox 2004a). In the Theory of Current Concerns, goal pursuit is referred to as a 'current concern', w hich is a tim e-lim ited process initiated when a person becom es com m itted to a goal and ending w hen a goal pursuit is terminated. 181

Transitions to Better Lives

Cox and Klinger (2004b) identified 11 life areas in w hich most people aim for satisfaction through goal-setting and striving: (1) hom e and household m atters; (2) em ploym ent and finances; (3) partner, fam ily and relatives; (4) friends and acquaintances; (5) love, intim acy and sexual m atters; (6) self changes; (7) education and training; (8) health and m edical m atters; (9) substance use; (10) spiritual m atters; and (11) hobbies, pastim es and recreation. These life areas are highly sim ilar to the 'prim ary hum an goods' (valued aspects of hum an functioning and living) identified in Ward and colleagues' Good Lives M odel of offender rehabilitation (see Chapter 3; Ward and Brown 2004; Ward and Stew art 2003). The Good Lives M odel has as its underlying principle that offenders, like all hum an beings, seek satisfaction in certain life areas that contain the essential ingredients of hum an well-being. W hile we all seek satisfaction in these areas, Ward and colleagues suggest that the offender does so in problem atic or distorted ways. The challenge of offender rehabilitation is to encourage offenders to pursue goals that reduce risk and goals that build up a positive, non-crim inal identity that helps sustain socially acceptable behaviour (Ward et al. 2007). Focusing on the life areas, Cox and Klinger (Cox and K linger 2002; K linger and Cox 2004b) developed the Personal Concerns Inventory, which is an assessm ent of current concerns along with a num ber of appraisal dim ensions. Respondents are asked to identify their goals in each of the life areas and then rate each goal on a num ber of scales tapping goal value and likely attainability. We used this assessm ent procedure w ith a sam ple of 129 adult m ale prisoners to identify their life goals (M cM urran et al. 2008c). These m en identified goals consistent with those treatm ent targets identified in the risk -n e e d sresponsivity m odel of offender rehabilitation as likely to reduce reoffending (Andrews and Bonta 2003). They said that they wanted to stop offending, and were aw are that to do this they needed to be self-controlled, find and keep jobs, have stable accom m odation, quit drink and drugs, change support netw orks, and find new leisure pursuits. In short, prisoners seem to want what professionals think they need in rehabilitation. Furtherm ore, consistent w ith the Good Lives M odel, prisoners expressed life-enhancing goals, such as w anting a better lifestyle, gaining w ork experience, having good fam ily relationships, gaining skills, and getting fit and healthy. These aspirations translate into approach goals that are likely to provide the rew ards that sustain a person in persisting w ith goals related to risk reduction. So, w ithout being too naive about the possibility that prisoners' responses were coloured by the dem ands of the situation, 182

Goal-focused interventions with offenders

it seem s that goal-based approaches may be useful in offender rehabilitation. The question that arises here is, if prisoners have these positive goals, w hy do they not manage to pursue them successfully? People may have innate tendencies to strive for intrinsically satisfying goals, but these tendencies flourish only where there are supportive conditions (Ryan and Deci 2000). Social exclusion may underm ine intrinsic m otivation in that som e individuals do not have the w herew ithal or the opportunities to pursue positive life goals (Bonner 2006). Disadvantage m ay have been life-long, deriving from fam ily poverty, disadvantaged neighbourhoods, or poor educational opportunities. Alternatively, disadvantage m ay be a consequence of crim e, w hich often leads to reduced employability, unstable living arrangem ents and disrupted fam ily and social netw orks. These problem s define poor hum an and social capital, m eaning a lack of resources that people can use to achieve their goals (Farrall 2004). Social changes need to be m ade to help offenders com m it to and realise pro-social goals; however, in this chapter the prim ary focus is on individual aspects of goal choice and pursuit.

Motivational structure The Personal Concerns Inventory (PCI) has been used to investigate m otivational structure (Cox et al. 2000, 2002). Respondents rate their identified goals on a num ber of dim ensions, including im portance, com m itm ent, achievability, likelihood of attainm ent, control over attainm ent, im m inence of attainm ent, and happiness resulting from attainm ent. A nalysis of these ratings revealed two factors: (1) an adaptive m otivation factor, characterised by high perceived likelihood of goal attainm ent, expected happiness when goals are attained, and com m itm ent to goal striving; and (2) a m aladaptive m otivation factor, characterised by rating goals low in im portance, expecting no great am ount of happiness at goal achievem ent, and having low com m itm ent to goals. In substance m isusing sam ples, the adaptive factor has been shown as inversely predictive of quantities of alcohol consum ption (Cox et al. 2002) and positively predictive of readiness to change (Cox et al. 2000). This suggests that there m ay be value in developing this assessm ent for use w ith offenders. Scores on the PCI from 129 m ale prisoners replicated the two original factors: adaptive m otivation and m aladaptive m otivation (Sellen et al. 2009). The replication of these two factors suggests 183

Transitions to Better Lives

that offenders' m otivational structure is sim ilar to that of other populations. An index of adaptive m otivation calculated from ratings correlated positively w ith self-reported internal reasons for entering program m es. An index of m aladaptive m otivation correlated positively with being in the pre-contem plation stage of change as assessed by the U niversity of Rhode Island Change A ssessm ent (U RICA) (M cConnaughy et al. 1983, 1989) and negatively with staff ratings of m otivation for and com pliance with treatm ent. Additionally, som e m odest changes in m otivational structure were observed after the com pletion of treatm ent program m es, with improved adaptive m otivation scores and reduced m aladaptive m otivation scores. In the process of conducting our research on the PCI, we noticed that the process of identifying and rating goals was w ell received by prisoners. They reported that the interview, which lasted for betw een one and two hours, helped them identify and clarify their life goals. Additionally, in an offender adaptation o f the PCI, we asked prisoners to identify how offending or being in prison helped or hindered goal attainm ent. This was intended to clarify the im pact of offending on their life plan and hopefully m otivate them to engage in treatment aimed at reducing their offending. A prelim inary investigation into this m otivational effect w as exam ined w ith a small sam ple of sex offenders w ho were reluctant to engage in treatm ent program m es, w ith som e positive indications of a m otivational shift, although this needs further investigation (Theodosi and M cM urran 2006).

Focusing on goals in treatm ent The aim in this section is to outline a goal-focused approach to w orking w ith offenders. The overall aim is to assist offenders to identify their personal goals and help them identify ways of w orking positively and pro-socially towards goal attainm ent. A goalfocused approach begins by helping the client review, rationalise and prioritise his or her goals. The underlying purpose is to shift the offend er's m otivational structure in a positive direction by helping him or her aim for valued and achievable goals that form a coherent whole. Then, an action plan for w orking towards attaining life goals is specified, w ith engagem ent in treatm ent program m es being one option. A second goal-focused approach is problem -solving skills training. The path to attaining life goals is fraught w ith m inor hassles and m ajor life events that can throw us off course, and therefore we need to possess troubleshooting skills. Problem -solving skills training 184

Goal-focused interventions with offenders

can augment major life goal strivings by helping people avoid maladaptive or anti-social solutions to day-to-day problems. Goal setting

Based upon the PCI and systematic motivational counselling, which is the intervention that follows from it (Cox and Klinger 2004c), a system for identifying and prioritising life goals, examining obstacles and strengths in relation to goal pursuit, and form ulating specific action plans is presented in Tables 11.1 and 11.2. This Goal Positioning System (GPS) is completed with the client over a number of sessions. The information is typed into the system at the end of each therapy session and both the client and the therapist retain a copy after each session to record the progress of therapy. Table 11.1 shows a completed example of a GPS. Therapy sessions are introduced by explaining the importance of intrinsically rewarding life goals to the client. Life goals may be to do with 'being' - for instance being happy, being loved, or being healthy - but these being goals need to be translated into 'doing' goals: that is, doing the things that will bring happiness, love and good health. Doing goals are often extrinsically motivated in that their attainment is not intrinsically satisfying, but it does assist with the eventual attainment of intrinsically satisfying goals. An example for offenders might be attending treatment programmes: this is not an intrinsically rewarding activity, but it m ay assist with the attainment of intrinsically rewarding life goals, such as good relationships, a satisfying home life and stable employment. Care must be taken to ensure that extrinsically motivated goals actually do relate to important life goals. The pursuit of money or material goods, for instance, may become disassociated from any basic human need and this pursuit may actually detract from basic need fulfilment, leading to poorer well-being (Ryan and Deci 2000). The client is then presented with the goal schedule with the explanation that the life areas represent the most important components of human happiness. The first step is to consider each of these areas and identify the principal personal goals. Second, the areas should be ranked in priority order so that the most important, urgent, or manageable areas can be worked through first. Then, focusing on each area in turn, identify the major obstacles to goal attainment. These will include personal (such as substance use), interpersonal (such as conflict in relationships), and social obstacles (unemployment, for example). The focus should then turn to a

185

Table 11.1

Goal positioning system

Life area

General goals

Ranking

House and home

Get my own flat.

1

No money,

Employment and finances

Be self-employed.

3

Need to start from scratch.

Education and training

Get a. painling and decorating qualification.

None.

Family and relatives

Get back on good, terms with mv j mum.

Mum doesn't want to know me.

Friends and acquaintances

Make new friends who don't gel into trouble.

8

Problems

Most of my friends go out drinking and get into fights.

Strengths

Impaci of offending

Specific goals

I want to work for my money.

Employers don't want someone with a record.

• Work with my brother.

My brother does painting and decorating - he's said T can do some work with him.

Tt might be hard to get a job with my record.

• Plan work with my brother.

Can gel ihe training in prison.

Being in prison has helped!

• Complete the course.

My slep-dad and brother still speak lo me.

Mum threw me out of Ihe house because of mv drinking and crime.

• Write Mum a letter saying I'm doing courses in here.

My brother and his males are more settled and don't get into trouble.

• Spend more lime If 1 hadn't been with those people wilh my girl­ friend, brother, 1 wouldn't have committed the familv 14 and mv «/ offence. friends who go out for a quiet drink and don't fight.

Love, Stick with my intimacy and girlfriend. sexual matters

4

She doesn't like me gelling pissed, fighting and ending up in trouble. She said she'd leave me if I get in trouble again.

Health and medical matters

Cut down my drinking.

5

I spend too much time and money in the pub and I get into fights.

Spiritual

T3e happy

9

I've got no flat, no job, my girl­ friend is pissed off with me, and my mum has chucked me out.

6

Tt's what T do!

matters

Leisure, hobbies, pastimes

Stop spending mv time in the pub.

She hasn'l dumped me yel.

She's fed up wilh me gelling senl lo prison.

♦ Gel work wilh my brolher. ♦ Cut oul ihe drinking and leave ihe boys to fight and drink if they want to - T want to make a fresh start.

I can drink sensibly. J

I've got con­ victions for violence - that's due to drink.

♦ Drink sensiblv with mv brother, my girlfriend and mates who don't fight.

I am young enough to change - my brother will help me.

Drinking, crime, ♦ Cut down and prison make drinking. me unemployed ♦ Get work. and make mv ✓ girlfriend and family pissed off with me.

I have started learning martial arts.

I'll get fit in prison • Get back into by not drinking Tai Kwon Do and going to the practice, gym.

Transitions to Better Lives

person's strengths: that is, the assets that may be called upon in the service of goal pursuit. These too will be personal (such as physical fitness), interpersonal (for exam ple, a supportive m other), and social (such as stable accom m odation). A ttention is draw n to the methods of goal striving by asking offenders about the role of their offending in goal attainm ent. Does offending help or hinder goal attainm ent? O ffending m ay help in attaining som e life goals but is likely to interfere with others. For instance, burglary may assist with the attainm ent of m aterial goals, but burglary itself or the crim inal justice consequences of it m ay interfere with harm onious relationships (relatedness), em ploym ent prospects (achievem ent), or feelings of self-satisfaction (self-transcendence). M ethods of goal striving that are likely to lead to successful attainm ent of intrinsically satisfying states should be identified and translated into general goal-directed actions. The next stage is to form ulate a single action plan (see Table 11.2). This stage consolidates and further operationalises the goals. Persistence in goal strivings depends in large part on the value the individual places on the goal outcom e. People engage more enthusiastically w ith goals that are set by the self, com pared with those that are externally directed (Deci and Ryan 2000). This has im plications for w orking w ith offenders, w hose goals are often externally directed in the sentence planning process. Even where sentence plan goals are agreed upon by the offender, there is suspicion that they are not genuinely adopted but rather that the offender is sim ply creating a good im pression. The skill is to assist the offender to identify his or her ow n valued goals and devise non-dam aging w ays of striving for goal attainm ent. Entry into offender treatm ents m ay be one method of w orking towards general life goals and if the offender identifies this as a specific goal then this level o f self-determ ination m ay enhance engagem ent in the treatment program m e. Recidivism outcom es are better for offenders who enter treatm ent voluntarily com pared to those w ho are mandated or coerced (Parhar et al. 2008). Persistence in goal strivings also depends upon the individual's expectation that an action will result in a specified outcom e (Vroom 1964). That is, if a person thinks his or her goal-directed actions are unlikely to lead to positive results, that person is unlikely to put much effort into doing w hat needs to be done to attain the goal. Expectations of success vary w ith goal topography (M cM urran and Ward 2004). A goal is more likely to be attainable if it is SMART; that is, specific, m easurable, achievable, rew arding, and tim e-lim ited. Goals that are specifically stated w ith m easurable outcom es allow the individual to set personal perform ance standards w hich m ay be monitored in 188

Table 11.2

Action plan.

Specific goals

Steps

Start a painting 1 Speak to my brother about and decorating selling lip a business together. business. 2 Complete the painting and decorating course in prison. 3 Get information about selling up a business. Cut down drinking.

Deadline

Life areas this will improve

Brother

Next visit

Course tutor

End of October

• Employment and finance • Education and training • Tlouse and home • Love, inlimacy and sexual mailers • Spiritual matters

See Education staff Friday about Internet access

1 Co on prison alcohol groups.

Personal officer

2 Book Tai Kwon Do classes to keep me out of Ihe pub. 3 Co out with people who don't drink a lol and gel into trouble - my brother, my girlfriend., and sensible mates.

Tai Kwon Do instructor Brother, girlfriend

1 Decide what I want to say in a letter. 2 Talk it over with my girlfriend. 3 Write the letter.

4 Post the letter.

By end of this week Two weeks before release After release

• Health and medical • Leisure, hobbies, and pastimes • Family and relatives • Friends and acquaintances • Spiritual matters

This week

• Family and relatives

Girlfriend

Next visit

Joe on our wing who helps with letters Wing office staff

End of this month

F.nd of this month

Goal-focused interventions with offenders

Get back on good terms with Mum.

Contact

Transitions to Better Lives

relation to outcom e, thus influencing m otivation (Bandura 1986). Goals should be achievable but challenging. As long as the goal is valued, the achievem ent of a difficult goal is more satisfying (Locke 1996). Goals should be positive and rewarding. A pproach goals are usually com petency-based, and perm it individuals to focus on their successes and m astery of situations, w hereas avoidance goals require people to be vigilant for lapses in their good behaviour, and they require individuals to focus on their failures. Approach goals are positively related to w ell-being, w hereas avoidance goals are negatively related (Elliot et al. 1997), and approach goals appear to im prove offenders' engagem ent in treatm ent program m es (M ann et al. 2004). The tim escale for sub-goals should be fairly short, so that a sense of m astery and achievem ent is gained rapidly. Problem-solving

Successful goal-directed behaviour requires that the person has the com petencies required for success. The skills necessary include the abilities of em otion control, problem -solving and interpersonal com m unication, as well as a w hole range of practical skills relating to w ork, finance and the hom e. Treatment and training program m es to enhance these skills may be available to offenders, and they can be included as appropriate in sentence plans. H owever, problem -solving skills training is one generic intervention that is goal-based and aim s to provide individuals with the skills needed to identify and solve life's problem s. Problem -solving approaches m ay be seen as w ays of troubleshooting w hen the path to on e's life goals is thwarted. Social problem -solving is 'the self-directed cognitive-affectivebehavioral process by w hich an individual attem pts to identify or discover solutions to specific problem s encountered in everyday living' (D 'Zurilla and N ezu 2007: 11). The use of the descriptor 'social' identifies this as problem -solving applied to real-life problems. An effective solution is 'one that achieves the problem -solving goal (i.e., changes the situation for the better a n d /o r reduces the distress that it produces), w hile at the sam e tim e m axim izing other positive consequences and m inim izing negative consequences ... to others as w ell as oneself' (D 'Zurilla and N ezu 2007: 13). Thus, a solution that disregards the w elfare of other people is not an effective solution. A ccording to social problem -solving theory, problem -solving outcom es are determ ined b y tw o dim ensions: problem orientation, and problem -solving style (D 'Zurilla and N ezu 2007). Problem orientation (PO) is the set of cognitive-affective schem as that represent

190

Goal-focused interventions with offenders

a person's beliefs, attitudes and em otional reactions about problem s in living and ability to cope successfully with problems. Problem orientation can be either positive or negative. A positive problem orientation is the tendency to appraise problem s as a challenge and be optim istic about problem s being solvable if one applies some tim e and effort to the problem -solving process. A negative problem orientation is the tendency to view problem s as a threat, expect problem s to be unsolvable, doubt on e's ow n ability to solve problem s successfully, and becom e frustrated and upset w hen faced w ith problem s. Problem orientation serves a m otivational function, w ith a positive orientation facilitating adaptive problem -solving efforts and a negative orientation serving to inhibit problem -solving attempts. Problem -solving style refers to the cognitive behavioural activities that people engage in when attem pting to cope with problem s in living. Rational problem -solving is the constructive problem -solving style that involves the system atic application of specific skills, each of w hich m akes a distinct contribution towards the discovery of an adaptive solution or coping response. The specific skills are: defining a problem accurately; setting goals for change; generating a range of alternative solution ideas; considering the costs and benefits of each alternative; developing a solution plan; and evaluating the plan after it is im plem ented. An im p u lsivity/carelessness style is characterised by im pulsive, hurried and careless attem pts at problem resolution, and an avoidance style is characterised by procrastination, passivity and overdependence on others to provide solutions. Both such styles are dysfunctional in nature, usually leading to unsuccessful problem resolution. O ur studies have show n personality disordered offenders and vulnerable prisoners to be poorer at social problem -solving compared w ith a functioning adult sam ple (Hayward et al. 2008; M cM urran et al. 2002; M cM urran 2009b). Furtherm ore, poor social problem -solving has been show n to be associated w ith distress and depression in prisoners (Biggam and Pow er 1999a, 1999b; M cM urran and C hristopher 2009). Problem -solving skills training or problem -solving therapy aim s to teach people the skills for solving life's problem s, and has been used successfully in the treatm ent a range of problem s (Bell and D 'Zurilla 2009; M alouff et al. 2007). D 'Z u rilla and colleagues (D 'Zurilla and Goldfried 1971; D 'Zurilla and N ezu 1999 2007) describe six separate steps for successful problem -solving: (1) problem orientation, w hich is acknow ledging that problem s are a norm al part of life, recognising negative em otions as

191

Transitions to Better Lives

signals that a problem exists, and learning better to m anage and use their em otional experiences (for exam ple, view ing bad feelings as a cue that a problem exists); (2) problem definition, w hich is the ability to define a problem clearly and accurately; (3) goal setting, which is identification of the desired outcom e; (4) generation of alternatives, w hich is the creative generation of a range of possible w ays of achieving the goal; (5) decision-m aking, where after exam ining the likely positive and negative consequences of each potential solution to both self and others the best options are selected and arranged in logical sequence to form a m ean s-end action plan; and (6) evaluation, w hich is a review of the success or otherw ise of the action plan, either in progress or at its conclusion. In a problem -solving intervention called Stop and Think!, we have translated these steps into six key questions that guide the problem ­ solving process in clinical practice: (1) Feeling bad? (2) W hat's my problem ? (3) W hat do I w ant? (4) W hat are my options? (5) W hat is my plan? (6) H ow am I doing? These six key questions guide the Stop and Think! sessions, w ith a focus on each participant's current concerns, aim ing not only to solve existing problem s but also to teach people the problem -solving strategy. Stop and Think! group sessions have im proved problem -solving abilities w ith m entally disordered m ale offenders (M cM urran et al. 1999), personality disordered male offenders (M cM urran et al. 2001), and vulnerable m ale prisoners (Hayward et al. 2008). Recently, a random ised controlled trial of a com bination of 12 group sessions of Stop and Think! preceded by four individual psycho-education sessions improved social functioning in non-offenders diagnosed as having personality disorders (Huband et al. 2007). Problem orientation

A negative problem orientation is strongly associated with anxiety and depression in prisoners (M cM urran and C hristopher 2009), as well as non-offender populations (Bray et al. 2007; Kant et al. 1997). A negative problem orientation is w here problem s engender feelings of nervousness, threat and fear, there are feelings of frustration and upset when problem -solving efforts fail, and there is a lack of confidence in one's ability to solve problem s effectively. Clearly, this needs to be a focus in skills training or therapy if people are to be enabled to solve problem s effectively. In two m eta-analyses of treatm ent trials, a m ajor predictor of positive outcom e has been the inclusion of com ponents that encourage people

192

Goal-focused interventions with offenders

to becom e less negatively oriented and more positively oriented to problem -solving (Bell and D'Zurilla 2009; M alouff et al. 2007). In our ow n research w ith people diagnosed as having personality disorders, a reduction in negative problem orientation was the m ost significant predictor of im provem ents in social functioning (M cM urran et al. 2008b). W hen social problem -solving therapy for people diagnosed as having personality disorder w orks it does so by im proving social problem -solving ability, but specifically, by reducing negative problem orientation. Im proving problem orientation can be tackled by changing problem appraisal. Instead of view ing problem s as insurm ountable obstacles that get in the w ay of happiness, problem s are to be seen as a norm al part of life and, w ith a bit of effort, they can be tackled successfully. Problem s are normal - w e all have them a lot of the time - and they can be solved if you tackle them constructively. Helping a person to experience success in problem -solving is also important, and this often entails giving support in efforts to solve problems. In Stop and Think! participants are offered optional fortnightly individual support sessions, focusing on helping them carry out their problem -solving action plans. Throughout therapy, identifying the client's strengths enhances feelings of com petence. Identifying problem -solving successes and praising approxim ations to success is reinforcing. W hen problem -solving has not been successful, lack of success should be framed as a learning opportunity. This steers people aw ay from self-criticism and feelings of failure into a more positive approach of enquiry. W hy did that not work? How can I do it differently? The ability to be flexible and use alternative strategies when faced w ith obstacles is associated w ith good problem -solving skills in prisoners (Christopher and M cM urran 2009).

Conclusion Engagem ent w ith a coherent set of valued goals is w hat gives life m eaning, which is associated with happiness and w ell-being (Em mons and King 1998). The statem ent of purpose of HM Prison Service for England and Wales refers to its duty to help prisoners to lead lawabiding and useful lives in custody and after release. To lead a lawabiding and useful life captures the quality o f transcending self-interest that Emm ons (2005) associates w ith well-being. H elping the offender build a personal narrative or identity in w hich the self is defined as a law -abiding person w ho is useful to society is an em erging approach

193

Transitions to Better Lives

to offender resettlem ent. Draw ing upon principles of restorative justice, M aruna and LeBel (2003) have backed a strengths-based approach to the rehabilitation and resettlem ent of offenders, focusing on the positive contribution a person can m ake to society (see also the discussion of the Good Lives M odel in Chapter 4). Actively encouraging offenders to contribute to society - and encouraging society to accept offenders as valued m em bers - engenders feelings of belonging. This in turn facilitates the internalisation of w hat may have begun as externally directed goals, nam ely becom ing a law abiding and useful m em ber o f society.

194

Chapter 12

Treatment readiness and the therapeutic alliance Christina Kozar

This chapter discusses how the concept of therapeutic alliance can provide a fram ew ork upon which the strength of engagem ent between the client and the program m e provider can be understood, and how program m e providers might respond to clients who present with low levels of treatment readiness. M any offenders attend rehabilitation program m es because they are mandated to be there and, as a result, some may be either poorly motivated to attend, or attend unwillingly. It is suggested in this chapter that it is the skill, know ledge and attitude that a program m e facilitator or therapist brings to the programm e that can determ ine how w ell each individual participant engages with program m e content, how well group m em bers work together, and, ultimately, exert a profound influence on the extent to which participants benefit from the programm e. The argum ent that is advanced in this chapter, then, is that difficulties in form ing an alliance in the early stages of a rehabilitation program m e are more likely to occur in offenders who might be considered as having low levels of readiness, but it is the way in w hich treatment providers respond to this that will ultim ately determ ine how well offenders perform .

W h a t is the therapeutic alliance? The therapeutic alliance, also know n as the w orking alliance, describes a therapist and client's m eaningful and collaborative w ork towards therapeutic change. Therapists have reflected on the nature of the therapeutic relationship since the days of Freud (1958), but 195

Transitions to Better Lives

in recent years it has been the w ork of Bordin (1979, 1994) that has probably been most influential in this area. Bordin suggests that the therapeutic alliance (TA) com prises three distinctive, but interrelated, elem ents: goals, tasks, and bond . First, the process by w hich therapist and client m utually agree on change goals is regarded as central to the form ation of a strong TA. The extent to w hich therapists can m otivate clients into undertaking the tasks o f therapy (on the basis that this will achieve the agreed goals) is thought to be an im portant determ inant of treatm ent effectiveness (Bordin 1979). The developm ent of a bond is also considered critical, but this is som ething that should naturally evolve as a part of the process of negotiating goals and com pleting the tasks required to achieve those goals. This aspect of the alliance is likely to be of particular relevance to forensic practitioners who seek to work with clients on behaviours that m ay be perceived as sham eful. This can only be done w ithin a context where a positive em otional bond has developed betw een the therapist and the client (Kear-Colw ell and Boer 2000). The bond thus describes the quality of the relationship required to w ork collaboratively on identified change goals (H atcher and Barends 2006), and for som e theorists not only provides a fram ew ork from w hich treatm ent can be delivered, but rather is the treatm ent (Bordin 1979; M iller, no date). In this respect, the ability of the clinician to m ake an em otional connection with the client through purposive goal-oriented therapy is regarded as a necessary, if not sufficient, condition for change. Disagreem ents on the goals a n d /o r tasks of treatm ent or strains in the bond are seen as an inevitable consequence of the therapeutic process. Genuine confrontation betw een the client and the therapist on their specific view s, needs and agendas is generally regarded as fundam ental to therapeutic change. Bordin (1979), for exam ple, suggests that in successful therapies both parties are required to work through difficulties that em erge in the relationship, given that the client brings things to the therapeutic process that parallel their experiences in other relationships. The resolution of these difficulties is seen by some as the most essential aspect of any therapy. Safran and colleagues (2002), for exam ple, argue that different clients w ill require different responses from therapists when the alliance ruptures. Ruptures occur when the client confronts the therapist about the therapy, or w hen the client com plies, defers or sim ply w ithdraw s when confronted w ith difficulties. Thus effective practice is not sim ply w orking collaboratively, but related to the w ay in w hich therapists respond to the varying problem s that routinely arise w ithin the therapeutic relationship. 196

Treatment readiness and the therapeutic alliance

The therapeutic alliance in forensic contexts Any exam ination of the therapeutic alliance w ithin a forensic context should consider a num ber o f factors over and above those that are thought to be associated w ith general psychotherapeutic work. A bove all, the level of coercion to which clients are subject w ithin prisons or com m unity corrections is likely to have a profound influence of the developm ent of the TA. Forensic clinicians have various obligations relating to the legal context in which they w ork (including reporting to parole boards, advising correctional case m an agers/p rison staff on the progress of clients). Correctional procedures typically require at least som e level of disclosure to other correctional w orkers around w hether clients attend their program m e sessions, and the quality of their participation. O ften, there is also som e expectation that inform ation relating to anti-social activities undertaken by clients while under the purview of correctional services (such as drugtaking, violent behaviour) will be reported if it com es to light during treatment. In a forensic context, then, the alliance will not only com prise those elem ents traditionally associated w ith the care of the client (tasks, goals, bond), but m ust also recognise issues of social control (Skeem et al. 2007). It is unavoidable that offenders will be at least wary, if not suspicious, about w hat clinicians will divulge to correctional staff about them and their participation in treatment. D eveloping an effective alliance w ith m andated clients therefore requires a reconciliation of these dual roles. A nother distinctive aspect of forensic or correctional program m e delivery is that it invariably occurs in a group form at. The alliance in group w ork is likely to be conceptually different from that in individual treatment (H orvath and Sym onds 1991), although, of course, there are fundam ental aspects of the therapeutic process that exist across both contexts. The therapist in group program m es m ust aim to (a) foster group cohesion so that group m em bers are convivial; (b) work through differences betw een group m em bers; and (c) ensure that group m em bers w ork together to assist each other during the course of the programm e. Serran and colleagues (2003) have argued that the level of group cohesion that develops provides an indication of the level of the TA w ithin offending behaviour program m es. This makes sense. In group interventions clients will benefit from the input of other group m em bers as w ell as from that of the therapists. Conversely, problem s in the relationship with either therapists or other group m em bers are also likely to im pinge on the strength of the alliance. Clients are likely to observe closely how therapists respond 197

Transitions to Better Lives

to ruptures with other group m em bers, and use this to inform their subsequent behaviour in the group. It is, however, a client's individual responses to the therapist's efforts at creating collaborative and purposeful w orking relationships that is discussed in this chapter. A myriad issues relating to internal and external treatment readiness factors will affect the course and outcom es of treatment for each individual client in an offending behaviour program m e. The com plexity of these interactions has been recently highlighted by Ross, Polaschek and Ward (2008), w ho revised Bordin's (1979) theory of the alliance to incorporate aspects of treatm ent readiness. They suggest that system ic issues can im pinge significantly on the therapeutic alliance: the more difficult the client (com plex needs, hostility) and the circum stances (workload, access to training and supervision), the m ore com prom ised therapists will be in their attem pts to develop an effective alliance. Therapist and client characteristics, including their personality, attachm ent style and interpersonal schem as, will also influence the interactions that occur w ithin offending behaviour program m es, and the quality of the alliance is more likely to be an outcom e of com plem entary transactions rather than separate actions by either party (Constantino et al. 2002). Indeed, Ross et al. suggest that therapists undertake a w ide search to exam ine w hich aspects of the model (client, organisational, or other contextual factors, for exam ple) to attend to w hen ruptures occur.

The therapeutic alliance and treatm en t outcomes A num ber of m eta-analytic review s investigating the im pact of the alliance on individual treatm ent outcom es have consistently shown that the therapeutic alliance has a m odest but robust positive impact. H orvath and Sym onds (1991), for exam ple, in their analysis of 24 studies relating to the quality of the alliance, found an effect size of 0.26, w hich they interpreted to m ean that at least a quarter of the therapeutic change observed could be directly attributed to the alliance. They observed that the alliance im pacts on outcom e across different types of therapy, lengths of treatment and sam ple sizes. Sim ilar observations were m ade by M artin, G arske and Davis (2000) in their m eta-analysis of 79 studies (w ith a com parable although slightly lower effect size of 0.22). Various studies have exam ined the relationship betw een the alliance and therapeutic outcom es in forensic populations, although few have involved sam ples solely com prising correctional services 198

Treatment readiness and the therapeutic alliance

clients, and m any have focused only on ind ividu al treatm ent. There is, nonetheless, a grow in g bod y o f ev id ence supporting the contention that the alliance plays an im portan t role in offend er treatm ent outcom es, although the findings o f som e studies are inconsistent. In the d rug and alcohol field, for exam ple, M eier, Barrow clough and D on m all's (2005) com prehen sive review o f studies conclud ed that early alliance seem s to be a good pred ictor of treatm ent retention, but a less con sisten t p red ictor o f treatm ent outcom e. B arber et al. (2001) also reported that the alliance successfu lly predicted retention across treatm ent con ditions in a sam ple of cocaine d ependent clients particip ating in a nu m ber o f different intervention s. In this study, stronger therapeutic alliances w ere associated w ith higher rates of retention, although su rp risingly this w as not the case in the cognitive therapy condition. A s in the M eier et al. (2005) review, the therapeu tic alliance w as not correlated w ith self-reported d rug use d uring the six m onths o f treatm ent for any treatm ent condition. These findings conflict, how ever, w ith those of an earlier study by C on nors and colleagues (1997), w ho found that a strong alliance in alcoholic com m u nity-based clients w as significantly associated w ith treatm ent p articipation as w ell as reduced d rinking d u ring 12 w eeks o f treatm ent and at 12-m onth follow -u p. G erstley et al. (1989) also found that alliance predicted decreased d rug use seven m onths post-treatm ent in a sam ple o f anti-social p erso nality disordered m ethadone-m aintained clients. In som e w ays, of course, w hether the alliance im pacts on retention o r ou tcom e is perhaps tw o sides of the sam e coin. C lients w ho do not exp erience a collaborative pu rposefu l therapeutic relationship eith er leave treatm ent or do not m ake as m any clinical gains com pared to oth er clients. A critical issue not explored in these studies is the m echanism s at w ork w ithin therapy that im pinge on the d ev elop m ent and m aintenance of the therapeutic alliance. Issues relatin g to client m otivation, and other aspects of treatm ent readiness, m ay have utility in exp loring how the alliance is form ed and developed. In their review, M eier et al. (2005) found a m od erate but robust relationship betw een the alliance, m otiv ation and treatm ent readiness. A m ore recent study by Brocato and W agner (2008) also found that alliance scores (as m easured by the W orking A lliance Inventory (W AI); H orvath and G reenberg 1989) for clien ts p articipating in an alternative-toprison residential d rug treatm ent program m e, w ere not associated w ith retention b u t w ere associated w ith m otivation to change and treatm ent readiness. C lients w ho scored high er on the 'B o n d ' scale o f the W AI (w hich explores the qu ality of the relation ship w ithin 199

Transitions to Better Lives

therapy), were also more likely to increase their m otivation to change during treatm ent. This suggests that the alliance both affects and is affected by a client's attitude to being in therapy, and their w illingness and capacity to positively change. In line w ith other chapters in this book, it also suggests that m otivation is only one, albeit im portant, facet of treatm ent readiness. A num ber of studies have exam ined the im pact of the TA in violence intervention program m es, although most of these have focused on treatm ent for dom estic violence or spousal abuse. Brown and O 'Leary (2000), for instance, adm inistered the observer-rated WAI in their research and found that while these scores predicted husband abuse at the end o f 14 sessions of cognitive behavioural treatm ent (CBT), they did not predict treatm ent retention. Similarly, Taft and colleagues (2003) found a relationship betw een therapist­ rated WAI (although not client ratings) and abuse up to six m onths after 16 CBT group sessions, but no relationship w ith retention. This latter study also revealed a positive relationship betw een WAI and readiness to change, psychopathy, borderline personality disorder features (although this was a w eaker relationship), and hostile-dom inant interpersonal problem s, perhaps em phasising the im portance of a range of personality and situational factors on the therapeutic alliance. It would seem , therefore, that not only does a client's preparedness and ability to enact change heavily influence the strength of the alliance, but so too do personality characteristics that prohibit the developm ent of close and intim ate relationships. Ross (2008), on the other hand, reported less com pelling support for the effects of the therapeutic alliance. In a study of a 36-w eek CBT prison-based treatm ent program m e for violent men, Ross found that the WAI predicted client com pletion but not outcom e (as measured by m ultiple m easures, including the Violence Risk Scale (Wong and Gordon 1999-2003)). WAI scores were also correlated with client m otivation, psychopathy and client attitude, but the only factor that remained significant in m ultiple regression equations was motivation. Structural equation m odelling showed that the alliance and m otivation m ediated both change and program m e com pletion, and that this was bi-directional. Based on these findings, Ross concluded that a strong alliance enhances m otivation for treatm ent (and that m otivation for treatm ent enhances the strength of the alliance), but that personality factors play a less significant role. O ne possible explanation for the differences in findings betw een R oss' study and those of Brown and O 'L eary and Taft et a l concerns the length of treatm ent. As for drug and alcohol studies, as described above, the greater the time betw een 200

Treatment readiness and the therapeutic alliance

alliance m easure and outcom e, the less likely a significant relationship is revealed. This speaks to the possibility of changes over time in the course of the alliance, and it m ay be that the strength of the alliance plateaus further along in the process o f group interventions.

Alliance formation The developm ent of a therapeutic alliance com m ences with assessm ent. A thorough assessm ent should be conducted to ensure that the learning style and abilities o f the client are elucidated prior to treatment so that therapists can be responsive to clients' specific needs (M arshall and Serran 2004). Part of this process should also involve the developm ent o f a detailed case form ulation to explain the m echanism s underlying offending behaviour and help clients develop greater insight into their ow n behaviour. The process of developing a case form ulation also appears to be associated w ith therapeutic responses to potential problem s in treatment, rather than responses that are sim ply about offender m anagem ent (that is the focus is on understanding offending behaviour and opportunities for bringing about change, rather than sim ply case m anaging offenders) (Kozar and Day 2009). Throughout this process, clients are oriented to the process of therapy and the nature and expectations of treatm ent, as this can often be of im m ense benefit (Constantino et al. 2002). Forensic clients in particular m ay benefit from this strategy, as they may be unfam iliar with therapeutic processes, have suspiciousness about w hat therapy in a correctional environm ent will involve, a n d / or will dem onstrate a num ber o f traits typically associated with poor alliance form ation, as outlined below. Throughout the negotiation of group activities, it is im portant that interactions betw een therapists and clients are respectful and work towards positive therapeutic outcom es. Bordin (1994) em phasised that the change goal elicited during treatm ent m ust capture som ething central to the client's concerns. In the forensic context, this will invariably be about the resolution of m echanism s that contribute to previous offending behaviour. Bordin suggested that the identification of these goals should in and o f itself have great therapeutic benefit. M arshall and Serran (2004) suggest that strategies that are m ost effectual include asking open-ended questions, behaving genuinely, offering encouragem ent, dem onstrating care and acceptance, and creating opportunities in group for behaviour to be rewarded. They suggest that directiveness, which involves suggesting possible 201

Transitions to Better Lives

directions or alternatives to observed behaviours, rather than 'telling' clients w hat to do, should be used judiciously. Luborsky et al. (1997) suggest very sim ilar processes for im proving the alliance based on L uborsky's previous w ork, particularly around providing support and guidance on the client's goals, and offering understanding and acceptance. They also em phasise the im portance of conveying realistic hopefulness about the client succeeding, the recognition of progress tow ards the goals, and finding w ays to encourage clients to express them selves on som e occasions. The im portance of developing a functional m eans of relating during treatm ent has also been em phasised by Ross et al. (2008), w ho posit that the therapeutic task for the therapist is to have an aw areness of their ow n schem as and how they interact with the clients' to inform w ays of responding in group that are helpful rather than harm ful. W here clients have experienced difficulties in relationship form ation previously, it stands to reason that difficulties in patterns o f relating will continue in the therapeutic context. Therapists m ust be prepared to work through these issues. The deeper the pathology, the more time needs to be spent on form ing the alliance (Bordin 1994). It is w hen therapist and client am icably negotiate a m eans of w orking together to effect positive change for the client that an adequate TA has been form ed. H orvath and Luborsky (1993) contend that this process not only requires the client approving the therapist's style but the therapist com m unicating the relevance of tasks to goals, and m aintaining an aw areness of a client's com m itm ent to therapy. They suggest that it is im portant to negotiate short- and m edium term expectations to foster a strong alliance but that the first phase of therapy is about developing trust and collaboration, and the second concerns challenging dysfunctional patterns. Exploratory strategies are required to undertake this w ork, but it is recom m ended that this should only be attem pted once a client's distress or other problem atic state has been resolved (Constantino et al. 2002).

W h o will have difficulties in forming an alliance? Because we know that som e client characteristics assist in the form ation of positive bonds (for exam ple, quality of object relations stem m ing from appropriate attachm ent and bonding with parental figures, expectations of change) w hile other characteristics do not (avoidance, interpersonal difficulties), it is possible to anticipate when therapists m ay need to adapt their approach to foster a strong alliance 202

Treatment readiness and the therapeutic alliance

(Castonguay et al. 2006). M any o f these internal treatm ent readiness factors can be easily determined pre-treatm ent and discussed during the assessm ent process. For exam ple, clients' poor expectations of im provem ent is one such factor that has been associated with poor alliance form ation (Constantino et al. 2002) and in a variety of treatment contexts (Connolly Gibbons et al. 2003; Constantino et al. 2005). In this situation, clients' goals are not aligned with those of the therapist (who should view therapeutic change as possible for the client). It m akes sense that if clients do not believe that therapeutic interventions will benefit them or are unfam iliar w ith the process of therapy and w hat it m ight offer (even perhaps suspicious), they are unlikely to enthusiastically enter into a therapeutic relationship and create a strong alliance. H ence, openness to being involved in therapy m ay significantly im pact on a client's ability to form a strong alliance (Constantino et al. 2002). The m ost profound w ay for clients to dem onstrate poor alliance form ation is to leave treatment. Brown, O 'L eary and Feldbau (1997) found that for a group of spousal abusers, a third of non-com pleters cited treatment-related reasons for ceasing treatm ent, w hich included dissatisfaction with the content or structure of the group intervention. So where a client's treatm ent goals are not aligned or there is disagreem ent w ith the tasks required for treatm ent on offer, a strong alliance is not possible and a decision to leave treatm ent m ay be m ade. O ther treatm ent readiness factors that potentially im pinge on a client's ability to form an alliance concern their mental health. In their review of factors that im pede the developm ent of a TA, Constantino et al. (2002) stated that severity of psychiatric sym ptom s equates to greater difficulty in form ing the alliance although again there is inconsistency in the research (see C onnolly Gibbons et al. 2003). Particularly in their more severe form s, however, it m akes sense that sym ptom s of psychosis, depression a n d /o r anxiety would impede clients' ability to relate to therapists and w illingness or capacity to disclose issues in treatm ent due to the range of difficulties these clients would experience in their psychological functioning and ability to relate to others. Brocato and W agner (2008), for exam ple, found that a high degree of psychological problem s and Axis I diagnoses (41 per cent) were found in offenders w ho prem aturely left residential drug and alcohol treatm ent. It m ay be, therefore, that if clients are psychologically distressed or have difficulties in controlling their em otions or relating to others, it w ill be m ore difficult for them to form an alliance and this will place them at risk of discontinuing treatment. 203

Transitions to Better Lives

There are a num ber of m ore pervasive client characteristics that seem to im pact on the capacity to form a therapeutic alliance. These include clients who have interpersonal difficulties, poor object relations, high levels of defensiveness, or are resistant or hostile (Constantino et al. 2002). It seem s likely that these traits would m ake the developm ent of trusting and intim ate relationships m ore difficult. H ostile-dom inant interpersonal problem s, which have been correlated in num erous studies with poor alliance form ation (see Beauford et al. 1997; C onnolly Gibbons et al. 2003; Skeem et al. 2007; Taft et al. 2004) seem particularly relevant to the forensic population. M any offenders have a history of violent behaviour a n d /o r anti-authoritarian attitudes that are likely to create difficulties in their interactions with others. A nother relevant factor to the forensic population concerns the difficulty in form ing an alliance w ith clients who experience paranoia. These clients seem to exhibit more difficulty than clients w ith any of the other personality disorder traits in alliance form ation, including those with anti-social personality disorder or borderline personality disorder (Lingiardi et al. 2005). In all, a high prevalence of anti-social or other types of personality disorder exist w ithin the forensic population, and these clients often exhibit other com plexities such as substance abuse and other A xis I diagnoses (Blackburn 2000). It should be expected, therefore, that m any clients will dem onstrate pervasive characteristics that w ill challenge alliance form ation (see Chapter 9). There are, however, various strategies that can be enacted to assist in avoiding difficulties both in the form ation of an alliance and in responding to ruptures in these situations (see below ). W hat is critical at the outset, however, is for therapists to identify which clients are likely to have m ost difficulty in form ing alliances during the assessm ent process, so that when treatm ent program m es com m ence therapists possess the requisite skills to assist clients.

Programme setting and therapist characteristics To date, there has been lim ited research on the im pact of organisational factors on treatm ent readiness and the therapeutic alliance. Luborsky et al. (1997) suggest that despite the paucity of research in this area, there is som e evidence in the substance-abuse field that the qualities of an organisation will strongly influence the alliance. Existing research suggests that providing reward for attendance, such as m oney or fo od /refreshm ents during treatment, is likely to foster the alliance, but further research on this is needed. 204

Treatment readiness and the therapeutic alliance

It could be hypothesised, however, that if clients perceive that an organisation acts respectfully and looks after their interests with som e generosity, they are more likely to com e to treatment with a positive attitude and be w illing to engage in collaborative w orking relationships. Organisations that value the provision of support and assistance to encourage program m e participation m ay be m ore likely to set the scene for therapeutic encounters that foster a strong alliance. Conversely, therefore, organisations w hose values are more punitive in nature, and lack sym pathy or respect for clients and their circum stances, m ight discourage trusting and intim ate relationships w ith clients. Som e program m es, for exam ple, m ake offenders pay to attend the sessions (Day et al. 2009b), and it m ay be that this alone leads to low er levels of treatm ent readiness and greater problem s in form ing effective alliances. A lternatively it may lead clients to believe that the service is in som e w ay im portant and valuable. It m akes sense that w hatever the underlying values an organisation holds towards its clients, these will play out in staff interactions as well as guide the choice of program m es and the m anner in w hich these are delivered. Serran et al. (2003) suggest that clients should be encouraged to w ork collaboratively not only w ith therapists but also w ith each other. They suggest that therapists need to deliver treatm ent according to standardised m anuals flexibly and sensitively, as opposed to strictly adhering to treatm ent protocols, in order to achieve this end. This highlights a tension that exists in program m e delivery. Som e correctional service treatm ent providers suggest that their staff deliver the m aterial in m anuals strictly as w ritten, diverging as little as possible if at all, to m aintain high levels of program m e integrity. Program m e integrity refers to attem pts to ensure that m aterials are delivered across a service consistently and w ith a high level of quality, but this m ust be balanced against being responsive to the characteristics of clients and the issues they bring to the group. It is likely to be particularly im portant to diverge from , or at least adapt, program m e m aterial and undertake additional activities or group processing w hen group cohesion is low and ruptures occur. Few em pirical studies have exam ined therapist skill as a variable that potentially im pacts on the alliance w ith forensic clients, although one study by Barber et al. (2008) exam ined therapist adherence to supportive-expressive (SE) therapy and its association with drug use in cocaine-dependent clients. SE therapy is a m anual-based treatment that initially focuses on the developm ent of the alliance in treatm ent and then exam ines interpersonal functioning and how this relates 205

Transitions to Better Lives

to drug abuse. A lm ost half o f this sam ple m et the criteria for anti­ social personality disorder, although not all of these clients had a history of conduct disorder. Strong alliance com bined with low levels of SE therapy adherence, rather than m oderate or high adherence, was associated with better outcom es. The authors concluded that m ore straightforw ard drug counselling techniques that target the here-and-now of drug clients' needs may be a better initial choice of therapy for these clients, rather than SE therapy that focuses on understanding m aladaptive relationships. So adherence to therapeutic protocols may, at times, reduce the strength of the alliance. Flexibility and responsivity appear to be key in this regard, to ensure that clients continue to be m otivated to engage in treatm ent by having their needs met. Therapist skill and style are central to alliance developm ent. Constantino et al.'s (2002) review of the developm ent of the alliance in general psychotherapeutic contexts suggests that therapist w arm th, support, acceptance, empathy, respect and directiveness are positive therapist attributes that contribute to its developm ent. They also noted that a balance m ust be struck betw een challenging clients' problem s and being attentive to their needs, and strong or rigid adherence to treatm ent m anuals can be detrim ental, particularly w hen there are relationship problem s w ithin the therapy because clients tend to then experience that their individual issues are not being addressed. They pointed out that ruptures may be caused by therapist behaviours, particularly when a technique is used at the w rong tim e in therapy. Furtherm ore, accurate interpretation is also likely to be im portant in the form ation of a strong alliance (Castonguay et al. 2006). Other characteristics associated with difficulties in the alliance caused by the therapist's style, such as rigidity, being tense, uncertain, self-focused, aloof, or critical, further suggest that therapists need to m ake clients feel safe and com fortable. A ckerm an and H ilsenroth (2001) also note that the w rong techniques used at the w rong times can create problem s in the alliance, such as the inappropriate use of disclosure (therapists discussing their ow n problem s) and transference interpretations (confronting clients about observed interpersonal characteristics too early in therapy). Ackerm an and H ilsenroth's review also identified a num ber of therapist behaviours that are likely to contribute to ruptures in the alliance. These include not being accepting, and rather being unresponsive, closed off, and unw illing to change their view despite client feedback. They suggest that exploratory interventions such as interpretation are best attem pted w hen the alliance is strong, to ensure that a high level of trust and m utual respect has been 206

Treatment readiness and the therapeutic alliance

developed to withstand the potential negative response a client may experience. Supportive interventions are best when the alliance is weak to enhance trust and ensure goals are aligned (Constantino et al. 2002). So not only do they reiterate the types of positive characteristics required of therapists, but they further em phasise that therapists need to use their skills flexibly to respond to client needs, which will vary across the treatm ent experience. H ighly skilled and trained staff are required to undertake the delivery of offending behaviour program m es. Successful program m e delivery requires enacting the right action at the right tim e to respond to both individual and group needs. N ot only is know ledge of offenders and offending essential b ut each therapist has to develop his or her ow n principles of program m e delivery based on training, experience, supervision and organisational practices. This skill and know ledge base in tandem with personal values and attitudes will inform a conceptualisation o f how clients can be assisted to change. Prim arily this com m ences with com m unicating a level of hope that change is possible. M arshall and Serran's (2004) review of their own research w ith sex offenders suggests that prom oting approach goals rather than avoidance goals assists in this end. They suggest that it is more effective to encourage clients to enact a new pro-social behaviour than have them avoid or cease an anti-social behaviour. The four m ost im portant therapist characteristics that correlated w ith clients' positive behaviour change were empathy, w arm th, rew ardingness and directiveness. This suggests that therapists need to balance hum ane and am icable approaches w ith skills largely employed by behaviourists, such as providing specific guidance around client behaviour and using praise when positive changes are observed. A bove all, therapists m ust be perceived as helpful (Serran et al. 2003). Respect is also integral to this process as it dem onstrates that therapists value and accept clients' strengths and interests. Training in the use of the four positive therapist characteristics also seemed to dem onstrate benefits to client outcom es. In summary, neither confrontational approaches nor unconditional positive regard appear to be effective approaches to w orking w ith offenders. A num ber of other behaviours should also be avoided, including blam ing, the therapist behaving as 'exp ert', or focusing on difficult issues too early. In w orking w ith clients w ho present w ith low levels of readiness and cautiousness about entering into a therapeutic relationship, the therapist needs to convey support and acceptance in order to foster respect and trust, w hile also insisting on change and that the client take risks in therapy to explore issues 207

Transitions to Better Lives

relating to their offending. There are a num ber of characteristics of clients, therapists, and the setting in w hich treatm ent is offered that are likely to im pact on the w ay in w hich the alliance develops. These include things that the client brings into the treatm ent environm ent (such as the personality of the client, his or her relationship history, and experiences of other program m es or psychological treatm ents) that will influence the m otivation and capacity to engage, as well as the w ays in w hich the therapist or program m e facilitator delivers the treatment. As the alliance begins to develop, it is then likely that readiness and the alliance will m utually interact and both will change over time. Central to this process, however, is how ruptures to the alliance or problem s in therapy are managed by the therapist. It is this area that is explored in the next section.

W orking through ruptures in the therapeutic alliance Identifying ruptures

There is no particular course of rupture and repair patterns that is considered ideal for any particular client group, but there should be an expectation that ruptures w ill occur during the course of a therapeutic episode w hen dysfunctional relationship patterns are challenged. H orvath and Luborsky (1993) suggest that if ruptures do not em anate during later therapy, it is perhaps a sign that treatment is 'coasting', and that dysfunctional behaviours are not being challenged or that the client is responding to the therapist in an idealised w'ay. This is a particular issue in som e offending behaviour program m es, where there is pressure on both facilitators and clients to avoid confrontation such that participants get through the program m e or 'pass the course'. This m eans that continual reassessm ent of the therapeutic approach through supervision or peer review processes is required to ensure therapeutic integrity. A wide variety of client behaviours may indicate w hen a rupture has occurred. Safran and Muran (2006) distinguish betw een two different types of ruptures: confrontation ruptures, where the client confronts the therapist about the therapy, and w ithdraw al ruptures, in which the client com plies, defers or w ithdraw s w hen they are confronted w ith difficulties. They argue that it is negotiation of needs (rather than collaboration) that m ost aptly describes the constantly shifting properties of therapeutic interactions, done at both a conscious and an unconscious level. For exam ple, unconscious influences may

208

Treatment readiness and the therapeutic alliance

take the form of projecting onto the therapist unresolved issues stem m ing from attachm ent style and replayed w ithout aw areness in the therapeutic relationship. Ruptures are defined as a breakdow n in or failure to develop collaboration, or periods of poor relatedness betw een therapist and client. Safran and M uran also add that even the m ost subtle fluctuation in the quality of therapeutic interactions is w orth exploring as it may assist in revealing and resolving client's relational schem as and self-defeating patterns. In addition, failure to explore more dram atic ruptures can lead to treatm ent failure and drop-out. Because ruptures m ay be dem onstrated in very subtle ways, it is im portant to carefully m onitor clients' experiences of the therapeutic alliance. Therapists m ay assum e that they are being received very differently from how the client actually perceives them , w hether positively or negatively (Constantino et al. 2002). A sking frequently for feedback from clients is essential, and having a structured efficient process to im plem ent this ensures both therapist and client com pliance. Duncan et al. (2003) have devised a brief m easure of the alliance - the Session Rating Scale (SRS) - for use as a clinical tool. This m easure was based largely on Bordin's (1979) concept of the TA as w ell as G aston's (1990) notion about the im portance of therapist and client having com m on beliefs around how people change. The SRS requires that the client rate four visual analogue scales: the therapeutic relationship, goals and topics, approach or m ethod, and an overall rating of the session. This measure has dem onstrated good test-retest reliability and adequate construct validity, and is moderately, and significantly, correlated w ith the O utcom e Rating Scale (ORS). The O RS (M iller et al. 2003) consists of four analogue scales requiring the client to identify their experience over the previous w eek. Ratings are m ade overall, individually, interpersonally and socially. These measures provide im m ediate feedback to therapists regarding the quality of the experience that clients had during their session, as well as the outcom es they are achieving outside of treatment. They can also becom e im portant clinical tools for discussing discrepancies in the therapist's perception of a client's group experience and progress, and how the client rates these experiences. The use of these tools is likely to assist in both the identification of ruptures and ensuring that progress is being m ade outside o f therapy. Clients are likely to experience difficulties at certain stages w ithin therapy. M arshall and Serran (2004) point out that w ithin forensic settings clients are m istrustful of the professionals running treatment program m es, so it m ay be that at the outset clients are disgruntled 209

Transitions to Better Lives

w ith having to attend treatment. M arshall and Serran suggest that the therapist requires great skill to overcom e this m istrust so that they can w ork effectively w ith the client, develop group cohesiveness and reduce symptom ology. They suggest that the therapist m ust model how to address others in the group and adjust their style to the needs of clients. Engaging coerced clients requires skills in identifying w hat client factors m ight be contributing to an unw illingness to engage in a therapeutic program m e, a level of self-aw areness to understand their ow n reactions to a client's behaviour, and an ability to accept and w ork with the client despite the challenges that arise. Although there is lim ited research in this area, current research suggests that when faced with anger and hostility, therapists who elicit som e self­ disclosure from the client in response to the anger, rather than ignore it or avoid responding to it, do better (Castonguay et al. 2006). This suggestion is also consistent w ith the attributes of being genuine and transparent w ithin treatm ent program m es. Therapists who attem pt to anticipate problem s, provide clients w ith a m eans of com m unicating dissatisfaction, and are open to dealing w ith ruptures as they occur are more likely to foster a strong alliance. A client's difficulties, of course, m ay not alw ays be expressed through anger or hostility. Safran et al.'s (2002) review of research suggests that m any clients do not express their dissatisfaction with treatm ent, w hich em phasises the im portance of asking clients for feedback on their experience of the alliance in treatm ent. Therapists w'ho becom e aware of clients' negative reactions may stick rigidly to their treatm ent m odel rather than respond to the rupture, or they m ay express their ow n negative feelings defensively. Conversely, therapists w ho respond non-defensively and shift their behaviours to respond to the rupture tend to im prove the alliance. An alliance is negotiated w hen there is a w illingness and ability to stay in tune w ith a client, while also accepting and responding to their difficulties. Responding to ruptures

The m anner in w hich ruptures are dealt w ith is critical to the course that a program m e takes. It is unlikely that taking a punitive approach to difficulties that arise in session will endear therapists to clients. Taking an interest in w hy clients are unhappy w ith the treatm ent process and seeking am icable m eans of shifting negative behaviour in a group will likely enhance the possibility of resolving a rupture. Serran et al. (2003) note that offenders m ay m inim ise their offending in an effort to protect them selves, and this will be

210

Treatment readiness and the therapeutic alliance

dem onstrated in a num ber of w ays during the treatm ent process. They see it as the responsibility of the therapist to create a safe and com fortable environm ent and build self-esteem to w ork w ith these defences. Thus therapist w arm th and em pathy can assist in reducing resistance w hen difficulties arise. Treatment providers, therefore, m ust prom ote program m e practices that seek to resolve difficulties through the use of collaboration and encouragem ent to ultim ately assist in group cohesion and alliance form ation. Problem s cannot, however, be avoided if ruptures are significant. Taft and M urphy (2007) caution against the use of confrontational techniques in dom estic violence program m es, arguing that they too are not likely to foster a TA and hence may ham per treatm ent gains. They go on to argue that if a client view s relationships as being based on pow er and control a confrontational approach m ay affirm this schem a, resulting in clients feeling angered if they perceive they are being belittled - clients may respond aggressively in an effort to feel em pow ered w ithin this type o f interaction. They suggest that m otivational and other therapeutic strategies to challenge client m inim isations and justifications are likely to be more effective at enhancing engagem ent and m aking treatm ent gains. It is encouraging also that the Good Lives M odel (see Chapter 4; Ward and Stew art 2003), which prom otes the perspective that clients have human needs that can be achieved in pro-social w ays using a strength-based approach, is gaining greater acceptance as an approach that can be incorporated into correctional services practices. This approach values clients as hum ans requiring autonom y and respect, rather than having problem s that m ust be eradicated in therapy. Enacting appropriate responses to ruptures in the therapeutic relationship is considered key to ensuring a successful therapeutic relationship and m axim ising clients' positive behaviour change outside of therapy. There is, however, relatively little em pirical work exam ining the im pact of ruptures (and attem pts to repair ruptures) on treatment outcom es. One study by Strauss et al. (2006) showed that stronger alliances and rupture-repair episodes predicted im provem ent in both depression and sym ptom s of personality disorder m ore than clients w ho experienced a rupture that w as not resolved or those w ho did not experience a rupture at all. In the context of offending behaviour program m es, it is inevitable that clients will bring to the group those characteristics and behaviours that played a part in their offences, and it is the role of the therapist to identify and intervene at this level during the treatm ent process. The ability of the therapist to foster the developm ent of the alliance, anticipate ruptures and repair

Transitions to Better Lives

these throughout the therapeutic process should, therefore, also be regarded as a treatm ent readiness factor in and of itself. Bennett, Parry and Ryle (2006) exam ined therapist responses to ruptures in a task analysis of cognitive analytic therapy sessions with borderline personality disordered clients, com paring good and poor client outcom es. Therapists in good outcom e cases identified and acknow ledged 84 per cent of the ruptures exam ined in their therapy sessions com pared to 34 per cent in poor outcom e cases. In addition, therapists in good outcom e cases either fully resolved or partially resolved alm ost all (87 per cent) identified ruptures with their clients, com pared to only a third of ruptures identified in poor outcom e cases. This study suggests that the process of identifying and having useful w ays o f responding to ruptures m ight be a central task relating to therapeutic im provem ent. A num ber of different approaches exist to respond to ruptures, and various com m onalities are central to these. Identification and acknow ledgem ent are the first tw o critical tasks followed by allow ing the client to reflect on the nature o f the rupture (Bennett et al. 2006; Safran et al. 2002). Constantino et al. (2002) also urge clinicians to 'avoid avoiding', so that if a client is angry it is best to allow' a freedom to express that, or the anger is likely to rem ain present and interfere with treatm ent. Bennett et al. (2006) then suggest invoking processes around negotiating and explaining the threat to the alliance and linking it to the dysfunctional patterns of responding previously identified in the client's case form ulation. This should ultim ately lead to a revised understanding of the rupture and new w ays of relating. This contrasts w ith Safran et al.'s m odel, which focuses more on the exam ination of core relational them es derived w ithin the therapeutic process, rather than early form ulations of the client's problem overtly discussed w ithin an assessm ent phase. They also em phasise the im portance of having clients express their feelings and identify underlying w ishes and needs dem onstrated by the ruptures. W ithin the forensic setting, where previous assessm ent of the function of offending and the m echanism s underlying this should be elucidated, Bennett et al.'s (2006) m odel has greatest utility. It is com m ensurate w ith the notion of exploring offence-paralleling behaviours at appropriate junctures w ithin group treatm ent (Jones 2004). The relative m erits of encouraging self-expression and self­ exploration of w ants and needs, however, should not be discounted as part of this process. Consideration of the im portance o f tim ing of techniques should again be em phasised w ithin the context of rupture repair processes. 212

Treatment readiness and the therapeutic alliance

Safran et al. (2002) suggest allying with resistance may at times be appropriate in ensuring that clients can use their defences. Similarly, Bennett et al. (2006) found that therapists who were involved in good outcome cases were also more likely to collude knowingly with a client at times in order to maintain the relationship when difficulties arose, but they did this with an understanding of what was occurring. This level of insight while concurring with client dysfunction during critical stages of relationship formation contrasted with therapists in poor outcome cases, who colluded without knowing that the therapy may be compromised. Safran et al. (2002) also discuss the possibility of dealing with ruptures at times indirectly, such as by shifting tasks or goals, or directly, say by providing an explanation for the use of particular activities, responding to complaints, reframing the meaning of tasks or goals in a manner that the client can relate to, and clarifying misunderstandings.

Conclusion This chapter has reviewed the construct of the therapeutic alliance and how it might inform how' therapists delivering offending behaviour programmes understand and respond to low levels of treatment readiness. It is suggested that the ability to form a therapeutic alliance with offenders and respond appropriately to ruptures when they occur are skills that therapists should possess in working with offenders. They are, however, complex skills that require much organisational support and training. An in-depth knowledge of the client, techniques to assist in alliance formation (particularly with those clients who have demonstrated difficulties in this regard), an ability to present material in a responsive manner, and the flexibility to achieve this while responding to potential ruptures in a group context are all required. There is a plethora of requisite skills and qualities, including openness to feedback when things are not going so well, to ensure this occurs. Those who have limited training and experience seem most likely to struggle with the challenges posed in achieving a strong TA in offending behaviour programmes, and often experience the therapeutic process as something to survive (Kozar and Day 2009). In these situations in particular, correctional providers should ensure adequate resourcing of expert supervision, training and observational feedback so that staff are adequately supported throughout this highly complex process.

213

Chapter 13

Readiness and risk: a case illustration

The purpose of this chapter is to illustrate som e of the concepts presented in this book through a detailed case description. The case is based on m aterial drawn from the authors' clinical experience, com bined and altered sufficiently to protect the anonym ity of individual offenders. In describing the case, a convicted child sexual offender w ho is due to appear before a parole board, a num ber of factors are considered that have the potential to influence the offend er's ability to benefit from a treatm ent program m e, such that his level of risk of reoffending is reduced. The chapter starts with a review of the history of the case, current crim inal justice and clinical context in which the offender was assessed, before m oving on to a consideration of the findings of the initial assessm ent, case form ulation and, finally, treatm ent recom m endations. This is followed by a discussion of treatm ent readiness, and how this m ight inform the developm ent of an expanded case form ulation and new treatm ent recom m endations.

The case of M r Jones Current context and index offending

M r Jones is a 41-year-old w hite m ale prisoner w ho has been referred to a psychologist by the parole board to ascertain his level of risk of reoffending and current treatm ent needs. He has been in prison for 12 years. A m ong the decisions to be m ade by the board are w hether M r

214

Readiness and risk: a case illustration

Jones presents sufficient risk o f reoffending to w arrant his continued im prisonm ent; w hether additional custodial treatm ent program m es are needed (and likely to be beneficial); and w hat treatm ent, support and supervision conditions will be required if he is to be released back into the community. These are all questions that forensic psychologists are often invited to provide an expert opinion on. M r Jones is currently serving a sentence of preventive detention follow ing his conviction for kidnapping, sexual violation and indecent assault of a 15-year-old boy. The offence occurred w hen M r Jones w as driving in a suburb of the city in which he lives and noticed the victim w aiting near a bus stop. H e induced the victim to get into the car (rather than w ait for a bus), and then drove to a secluded area where he becam e physically assaultive, striking the boy repeatedly across the face. It is reported in the Police Sum m ary of Facts that M r Jones kept the victim in the car for approxim ately 45 m inutes, continuing his verbal threats and intim idation, before driving him to another location where he fondled the victim 's genitals and, despite the bo y's resistance, inserted a foreign object into his anus. M r Jones then drove the victim back to the bus stop and released him , threatening to have him killed if the incident was reported to the police. Previous offending

Mr Jones has a history o f previous sexual offences of this nature. H e was convicted of the indecent assault of a 15-year-old boy som e four years prior to the index offence. Before this, Mr Jones's first detected sexual offending was against an eight-year-old boy, for which he was sentenced to 18 m onths' probation. At age 22 he was again convicted for indecent assault against two boys, aged 10 and 12 years, and sentenced to 19 m onths' probation and seven m onths' non-residential periodic detention. This offence also involved a conviction for abducting a child, although no further inform ation on the specifics of this offence is available in the file docum entation. At age 25 (four years previous to the current offence), he was convicted on three charges of sexual offending against the 15-year-old male, for which he was sentenced to six m onths im prisonm ent. M r Jones thus has a pattern of offending sexually against pre­ adolescent and adolescent m ales, with a total of five previous convictions for indecent assaults. There would appear to be a significant increase in the level of violence involved in these offences, although M r Jones disputes the degree of force and coercion used in

215

Transitions to Better Lives

som e of the earlier offences. All of the convictions were as an adult offender, from the age of 18 until the current (or index) offence which w as com m itted when M r Jones w as aged 29. In previous reports, however, it has been noted that Mr. Jones has also acknowledged an earlier onset of offending and a more extensive num ber and type of offences than those reflected in his crim inal convictions. In addition to his sexual offences, M r Jones has previous juvenile convictions for vehicle conversion, false statem ent and theft. Offender background and offence précipitants

M r Jones has been assessed on a num ber of different occasions in the course of his involvem ent with the crim inal justice system , and there are several com prehensive accounts of his social and fam ilial background contained in previous reports. To sum m arise these, M r Jones grew up in a suburb of a large industrial city, with his mother, father and younger sister, and reported being doted on and m aterialistically over­ indulged as a child by both his m other and grandm other. H is father becam e ill with some form of progressive dem entia w hen he was eight years of age, and this led to a reduction in parental discipline. There are suggestions that M r Jones resented the fam ily's focus on his fath er's illness, and reports indicate that by the age of ten he had begun to behave disrespectfully, or even abusively, towards his father. Indeed, it appears that M r Jones becam e exceptionally self-focused and developed a sense o f entitlem ent in relation to his apparently passive, dependent and over-indulgent m other and ineffectual father. As an adolescent he developed the capacity to lie and m anipulate his m other in order to get w hat he wanted: he is described as having poor im pulse control, poor ability to delay gratification, and deficits in social skills appropriate for his age, including a lack of em pathy for others. These problem s indicate difficulties in the pursuit of the goods of relatedness, agency and em otional com petency (inner peace), or more specifically, in the internal capabilities required to secure these goods in socially acceptable and personally m eaningful w ays (Ward and M aruna 2007) M r Jones has described him self as a 'slow learner' w ho preferred to stay at hom e rather than attend school. H e reports being severely teased by his peers regarding his sm all stature, eczem a and chronic asthm a. He seem s not to have been w ell accepted by his peers, leading to feelings of social inadequacy and the developm ent of m aladaptive strategies to align him self socially, such as telling exaggerated stories and engaging in farcical behaviour, but such attem pts ultim ately served only to com pound his rejection and isolation. These strategies 216

Readiness and risk: a case illustration

could be viewed as w ays of establishing a sense of connectedness to his peers and thus reflect the prim ary goods o f com m unity and relatedness. M r Jones would socially withdraw and rum inate over others' responses, leading to anger and resentm ent (problem s w ith em otional com petency). These rum inations would escalate his anger to the point that he would som etim es bully more physically and em otionally vulnerable peers. D uring this period of adolescence he also appears to have developed a pattern of alleviating his em otional distress through com pulsive sexual stim ulation and fantasy relating to them es of aggression and dom inance. Such sexual stim ulation strongly reinforced the fantasies of dom inance over others as a way of defending against his feelings of rejection and inadequacy, w hile sim ultaneously sexualising his violent im pulses. This repeated early pairing of sexual and aggressive retributional drives has been identified as one of the prim ary risk factors in M r Jones's sexual offending. In more constructive terms, this issue points to an em erging practical identity revolving around them es of agency and em pow erm ent, albeit translated into destructive w ays of relating to others and the broader social world. A t approxim ately age 15, M r Jones began teaching m artial arts to younger children w ithin an established Jud o Dojo. This experience is reported to have contributed to a sense of social efficacy, as he felt admired and respected by the younger students. H owever, it is also reported that he would fantasise about offending against his pupils and used his position as teacher to gain sexual stim ulation from physical contact w ith both m ale and fem ale children during classes. It is reported that he would later m asturbate to fantasies of abusing children, again further reinforcing these deviant fantasies. M r Jones's father died when he w as 16 years old and it has been reported that he continues to experience feelings of guilt, sham e and anger in relation to his behaviour towards him. M r Jones left school shortly afterw ards to join the w orkforce, but he dem onstrated an unstable w ork history, losing several jobs in his first year out of school. H e has reported that he had difficulty interacting effectively w ith adults, but that another factor in his em ploym ent problem s was his sense of entitlem ent, such that he did not like to be told w hat to do. Previous formulations

Previous psychological reports have indicated that M r Jones has continued to experience difficulties developing and m aintaining

217

Transitions to Better Lives

satisfactory adult relationships, resulting in subjective distress and feelings of isolation. It appears that M r Jones utilised his previous pattern of sexual preoccupation to m ediate stress and regulate his affect, and that fantasies of control and dom inance served to soothe his sense of inadequacy w hile expressing his anger and resentm ent over the perceived rejections o f others. They have noted that M r Jones's sexual offences against children have becom e more opportunistic and predatory in nature over time. Specifically, in his most recent offences, M r Jones drove around in his car to locate a potential victim , an adolescent male that he believed he could intim idate and control. He then physically and sexually abused the victim at different intervals during an extended period o f time, and induced a significant level of fear both to gain the victim 's com pliance and to prevent him from reporting the incident. Such behaviours are thought to m anifest M r Jones's sexualised fantasies o f dom inance, control and aggression, w hich then serve to facilitate tem porary feelings of pow er and efficacy that Mr. Jones otherw ise lacks.

Previous treatm ent provided M r Jones has now com pleted a specialised treatm ent program m e for sex offenders on three previous occasions. H is response at the com pletion of his first program m e experience was judged to be 'unsatisfactory', and he undertook the program m e a second tim e prior to his release from prison (for his first offence) after he becam e eligible for parole. At the end of this second attem pt at the program m e, it was stated that although he was m otivated to address treatm ent issues, 'his inability to be consistent in his application of therapy and an interm ittent return to m anipulative and dishonest behaviour restricted the progress he m ade'. M r Jones sexually reoffended after a short period of time in the com munity, and was returned to prison on a sentence of preventive detention. He was again referred to a treatment program m e, when it was noted that he had undergone 'a m ajor attitudinal change' and was now 'ready to benefit from treatm ent'. He subsequently successfully com pleted the sex offender treatm ent program m e, and although it was noted that w hile he appeared to have m ade further progress in the cognitive understanding of the factors contributing to his offences, he was yet to dem onstrate a consistent change in his behaviour. Prior reports go on to point out that a particular concern is his pattern of dishonesty, w hich is described as increasingly more 218

Readiness and risk: a case illustration

refined and difficult to detect. A lthough he successfully com pleted the program m e, there were incidents of rule-breaking related to his use of a com puter and attem pting to com m unicate inappropriately by letter w ith another inm ate. Follow ing these incidents he was discharged from the program m e for short periods, but subsequently allowed to return. An obvious problem in the previous treatment is its tendency to focus on negative treatm ent goals and failure to engage with M r Jones's im portant values and an identity that is focused on a com bination of perceived vulnerability and need to dom inate or, more accurately, to disem pow er others. His dishonesty and lack of investm ent in com m unity norm s all point to problem s of social consolidation and a view of him self as an outsider: som eone w ho had to fight to survive and be acknow ledged by others. In fact, he is eager to be accepted by other people but is unsure of how best to achieve this. It seem s obvious that any m eaningful treatm ent plan would need to pick up on these them es and provide constructive w ays for him to achieve the needs that underpin them.

C urrent assessment Mr Jones had been assessed on a num ber of occasions for various sentencing and parole board hearings, and has been through a sex offender treatm ent program m e on tw'o separate occasions, with all of the accom panying assessm ent procedures that are involved in that process. Thus there is an abundance of assessm ent inform ation available in his file. These are reviewed here relative to the assessm ent of the current level of risk for sexual reoffending and related barriers to treatment readiness. Intellectual functioning

O ne area of previous assessm ent involved M r Jones's level of intellectual functioning. This issue is im portant not as a risk factor per se, but rather as a factor that m ight influence M r Jones's ability to respond to cognitive behavioural treatm ent m odalities, and as a result lim it his ability to successfully im plem ent the relapse prevention strategies that such m odalities typically em phasise. A full scale IQ of 86 w as reported follow ing an initial assessm ent of intellectual functioning using the Revised W echsler A dult Intelligence Scale over ten years ago. This places M r Jones in the Tow average' range. H is verbal IQ score w as recorded as 81, w hile his perform ance

219

Transitions to Better Lives

IQ was recorded as 98. It w as noted that his lower verbal score could be understood in part as related to his poor perform ance in and early departure from form al education. H owever, a subsequent adm inistration of the W echsler Abbreviated Scale of Intelligence three years ago showed a full scale IQ of 106, w'ith a verbal IQ of 104 and a perform ance IQ of 106, all o f w'hich are solidly in the 'average' range of intellectual functioning. Despite the apparent discrepancy in testing results over time, it seem s safe to conclude that M r Jones is of at least 'low average', and possibly 'average' intelligence, so that form al intellectual deficits do not appear to pose a substantial im pedim ent to his capacity to understand what m ight be required in participation in a treatment programm e.

Assessment of risk to reoffend In this assessm ent, M r Jones's potential to reoffend w as evaluated using actuarial risk assessm ent m easures and considering dynam ic risk factors. Specifically, he was evaluated using the STATIC-99 (H anson and Thornton 2000), a risk assessm ent measure utilising historical offence related variables, as w ell as the Psychopathy Checklist Revised (PCL-R) (Hare 2003), and the STABLE-2007 (H anson et al. 2007), a measure of dynam ic risk factors for sexual reoffending. In relation to the STATIC-99, M r Jones's score placed him in the 'h igh' range of risk for sexual reoffending, based on his prior sexual offences, his choice of unrelated m ale strangers as victim s, and his lack of long-term intim ate adult relationships. Large sam ples of other sexual offenders scoring in this range on the STATIC-99 have shown sexual recidivism rates of 39 per cent at five years, 45 per cent at ten years, and 52 per cent at 15 years follow ing release to the com m unity (H arris et al. 2003). M ore recent norm s for the STATIC-99 indicate that M r Jones's score places him in a group that has a five-year sexual recidivism rate of 28.2 per cent for routine correctional sam ples and 44.0 per cent for high-risk sam ples, and ten-year sexual recidivism rates of 39.8 per cent for routine sam ples and 54.3 per cent for highrisk sam ples (H elm us et al. 2009). A nother w ay of describing risk of reoffence for som eone w ith M r Jones's STATIC-99 score is that his relative risk of sexual recidivism is four tim es higher than the average sex offender. Stable dynam ic risk factors are defined by H anson et al. (2007) as 'personal skill deficits, predilections, and learned behaviors 220

Readiness and risk: a case illustration

that correlate w ith sexual recidivism but that can be changed' through intervention (2007: i). The STABLE-2007 assesses 13 dynam ic risk factors. M r Jones scored in the high range on this m easure, including high scores on those factors involving lack of positive social influences, capacity for relationship stability, general social rejection, lack of concern for others, impulsivity, poor problem -solving skills, negative emotionality, using sex as coping, and deviant sexual preferences. It should also be noted that risk assessm ent w ith sex offenders also calls for consideration o f acute dynam ic risk factors, defined as highly transient conditions that only last hours or days. These factors include 'rapidly changing environm ental and intrapersonal stresses, conditions, or events that have been show n by previous research to be related to im m inent sexual re-offence' (H anson et a l 2007: i). The A CU TE-2007 (H anson et al. 2007) is designed to assess these factors, including item s such as victim access, rejection of supervision, collapse of social supports, and substance abuse, am ong others. However, because these factors exert an influence in the im m ediate environm ent in close tem poral proxim ity to potential sexual offending, it is not possible to assess these factors m eaningfully w hile an offender rem ains incarcerated (unless the concern is over sexual reoffending in the custodial environm ent). O ne may attem pt to anticipate the likely factors that m ay be present in the post-release environm ent and extrapolate from current functioning, but any attem pt to do this is ultim ately speculative and does not reflect the application for which the A CU TE-2007 was designed: to provide an ongoing assessm ent measure for m onitoring the risk of sex offenders under supervision in the community. Psychopathy, which represents a severe form of personality disorder w ith a strong em pirical association w ith a variety of clinical and crim inal justice outcom es, including recidivism , was also reassessed using the revised Psychopathy C hecklist (PCL-R). M r Jones was scored in the 'm oderately high' range of the PCL-R. Thus w hile he dem onstrates m oderately high levels of psychopathic traits, he does not m eet the criteria to be classified as a psychopath and does not reach the level of psychopathy considered to be in the sam e class of offenders as those w ho have consistently dem onstrated the worst outcom es for treatm ent failure and serious reoffending. The com bination of deviant arousal and psychopathy have been associated w ith som e of the highest observed rates of sexual reoffending. The strongest predictor of sexual reoffending is the com bination of high levels of psychopathy and the presence of 221

Transitions to Better Lives

deviant sexual interest or arousal, defined as arousal to sexual activities with children or coercive sex with non-consenting adults. In a recent exam ple of such findings, H ildebrand, de Ruiter and de Vodel (2004) exam ined the sexual recidivism rates am ong a sample of treated rapists. They reported a sexual reconviction rate of 82 per cent over an average follow -up of 11.8 years for offenders w ho were both psychopathic and sexually deviant, com pared to 18 per cent for offenders w ho were both non-psychopathic and non-deviant. Sim ilar outcom es have been observed w ith other sam ples including child m olesters (Rice and Harris 1997). The assessm ent of sexual deviance thus w arrants close consideration. It appears that risk assessm ent experts som etim es conclude that deviant sexual arousal m ust be present based solely on the presence of convictions for sexual offences. Yet sexual offending alone is not sufficient evidence for the presence of sexual deviance. International experts such as H art and Kropp (2009) have stated: M ental health professionals should attem pt a direct and com prehensive evaluation o f sexual deviance, gathering inform ation about norm al and abnorm al sexual thoughts, urges, im ages, fantasies, behavior, and physiological arousal. An im portant corollary of this standard is that assessm ents of sexual deviance should avoid over-focusing on convictions for sexual offenses. Sexual offenses are neither necessary nor sufficient for a diagnosis of sexual deviance. M any people w ith sexual deviance never act on their thoughts, im ages, urges, or fantasies; and m any of those w ho act in a m anner consistent w ith their sexual deviance do so in a way that m ay be perfectly legal. Also, many - perhaps the m ajority - of people w ho com m it sexual offenses do not suffer from sexual deviance. Sexual offenses m ay be the result of other causal factors, including such things as anger, generalized negative attitudes toward w om en, poor impulse control, poor heterosexual skills, and inappropriate sexualization of nonsexual needs. A ssum ing that all sexual offenders have sexual deviance is as illogical as assum ing that all thieves have kleptom ania or that all arsonists have pyrom ania. (2009: 560). In the case of Mr Jones, there is sufficient evidence that he experiences deviant sexual arousal. The assessm ents conducted as part of his treatm ent program m e participation included the Penile Plethysm ograph, w hich indicated significant levels of arousal to 222

Readiness and risk: a case illustration

coercive sexual stim uli w ith pre- and post-pubescent m ales and females. H e also has self-reported sexual arousal to fantasies of dom inance and physically coercive sex w ith adolescent boys. In light of the results of the current assessm ent of static and dynam ic factors, Mr Jones is still considered to present a 'high risk' for sexual reoffending.

Assessment of needs Considering personality features m ore broadly as they relate to treatm ent considerations, the results of a M illon Clinical M ultiaxial Inventory, third edition (M CM I-III) provide a profile of M r Jones's personality features. As far as can be assessed from scores on this m easure, he appears to display prom inent avoidant, dependent and self-defeating personality features, m arked by a significant level of reported anxiety. D espite rem arks regarding a sense of entitlem ent presented in various earlier reports, M r Jones had a particularly low score on the scale m easuring narcissistic traits, w hich are often associated w ith a sense of entitlem ent. Case formulation

Identification of the factors contributing to and m aintaining Mr Jones's offending behaviour have been presented in previous reports, and will be briefly integrated into an explanatory form ulation here. This could be summ arised as follows. M r Jones's early life experiences included rejection and ridicule by his peers, com bined w ith over-indulgence by his m other and ineffectual discipline related to his progressively disabled father. These conditions led to the developm ent of a profound sense of inadequacy and anxiety, social incom petence and isolation, along with a sense of entitlem ent and the use o f m anipulation and lying to get his w ay w ithin the family. A lthough he had a strong desire to connect interpersonally w ith others, he did not develop the social skills or interpersonal sensitivity to achieve lasting relationships. W hen his dependency needs w ere not m et and he experienced, instead, the teasing and ridicule of his peers, he appears to have begun to com fort him self through sexual stim ulation, with fantasies of control, aggression and retribution. H is offending thus results from a highly sexualised ideation o f others, particularly physically and em otionally less m ature m ales w'ith wrhom he can act out the deviant

223

Transitions to Better Lives

sexual fantasies that boost his sense o f efficacy and power. It also appears that M r Jones displays entitlem ent stem m ing from his over-indulgence as a child and his limited capacity to recognise the im pact of his behaviour on others or the consequences to himself. U nlike the overtly grandiose sense of self-w orth associated w ith narcissism , M r Jones in fact struggles w ith a profound sense of inadequacy and social incom petence, contributing to his anxiety and occasional depression. The potentially positive side of those w ith essentially avoidant personality patterns is that they can be extrem ely sensitive to the needs and perspective of others. They can potentially show substantial com passion and understanding and be em otionally responsive, if they can set aside their m istrust and expectation of rejection or ridicule to develop more direct and less reactive interactions w ith others. Such a form ulation is consistent w ith the results of the M innesota M ultiphasic Personality Inventory (M M PI) personality assessm ent. Individuals with the personality profile of M r Jones desperately w ant to be accepted and involved w ith other people, but this desire is blocked by intense fear of being rejected or ridiculed. They scan their environm ent for threats to their self-esteem and try to present them selves in a favourable manner, but are seldom successful, as they feel a continual sense of uneasiness and anxiety, tending to overreact to m inor events. They typically perceive them selves as socially inept, inferior and inadequate. Because o f a fear of social situations and close relationships, such individuals may rely heavily on fantasy to gratify their needs for affection and to cope w ith their anger. This has the potential to restrict them to a solitary life where they are m ore likely to reactivate m em ories of past social rejections, rather than risk form ing new, m ore satisfactory relationships. However, passive aggressive elem ents may also be present, as indicated by m oodiness and resentm ent, w ith significant difficulty trusting others. This is consistent with previous descriptions of M r Jones's vacillating betw een being friendly and cooperative and then being hostile, followed by apologies. Conceptualising M r Jones from a Good Lives M odel (see Chapter 4) perspective, it would appear that his 'practical identity' was based on the prim ary goods o f agency and related ness and he sees him self as a person w ho needs to fight back and gain respect from others through dom inating and sexually abusive behaviour. His perceived vulnerability and trouble articulating his needs to others also highlights difficulties w ith em otional competency. Taking all this into account, a good lives plan needs to em phasise the im portance of 224

Readiness and risk: a case illustration

establishing m eaningful relationships and to find a w ay of enhancing his sense of agency and personal control. This should be done by carefully considering his social ecology and the level of resources available w ithin the com m unity he is likely to be released into.

Typical treatm ent recommendations Treatment goals for som eone with M r Jo n es's history and presentation typically involve the continuation of the treatment that has already been started. There are som e indications of progress: M r Jones appears, for exam ple, to have a solid cognitive understanding of his offence cycle, including the precursors and high-risk situations associated with his sexual offences. Furtherm ore, he states that he continues to practise the m asturbatory reconditioning techniques that he has been taught in order to strengthen his arousal to appropriate adult partners and decrease his interest in young m ales like those he has victim ised. The areas w here he needs to dem onstrate continuing progress involve applying the skills and insights that he has learned in his current everyday interpersonal relationships. H e also accurately identifies that he needs to w ork on being consistently open and honest in his dealing w ith others and elim inate his use of deceit, m anipulation and secrecy, as these behaviours have contributed to the developm ent of situations in which he has offended, as well as interfered w ith his capacity to benefit from interventions. Specific treatm ent goals for M r Jones have been consistently identified in previous psychological reports. These have often been listed as follows: • Develop an understanding of how he cam e to offend. • Learn how to m aintain healthy adult relationships. • Replace sexual thoughts of children w ith appropriate adult fantasies. • Replace coercive sexual thoughts with appropriate intim ate fantasies. • Learn appropriate interpersonal boundaries. • Develop alternative coping strategies during tim e of negative em otions. • Learn not to see children as sex objects. • U nderstand the effects o f offending on victim s and develop empathy. • Challenge thinking errors. 225

Transitions to Better Lives

• Learn w arning signs and high-risk situ ations and how to m anage these. • Take full responsibility for his offend in g (this is an ethical issue as w ell). • Integrate all the above w ith in a good lives plan and his practical identities and associated values.

Assessment of tre a tm e n t readiness T his is w here trad itional assessm en t and treatm ent recom m endation reports often end. H ow ever, to b e o f m axim um utility, assessm ent and treatm ent p lanning reports should also, in o u r view, inclu de an explicit con sid eration o f treatm ent readiness. There is quite a lot of inform ation already av ailable from this assessm en t that is directly relevan t to un derstand ing M r Jo n es's read iness for further treatm ent, and his likely prognosis. It m ay also be helpful, how ever, to ad m inister som e m easures o f treatm ent read iness (described in C h ap ter 5). For exam ple, the C orrections Victoria Treatm ent R ead iness Q uestion naire (C V TR Q ) (C asey et al. 2007), although not validated specifically for use w ith sexual offenders, p u rports to assess read iness to engage in a treatm ent program m e through four subscales: attitu d es and m otivation, em otional reactions, offend ing beliefs, and efficacy. T his m easure d oes n ot provide cut-off scores, but had it been ad m inistered , M r Jon es w ould have an overall score that w ould place him in the m od erate ran ge of treatm ent read iness (a total score o f around 70) w hen com pared w ith other o ffend ers.1 H ow ever, closer exam in ation o f his subscale scores m ay be m ore inform ative. They w ould reveal, for exam ple, that he scored in the high range of the attitu d es and m otivation subscale and the em otional reactions subscale (high er scores ind icate high er levels o f treatm ent read iness), but in the low range of the O ffend in g Beliefs and Efficacy subscales. This w ould suggest that M r Jo n es has a relatively positive attitude tow ards treatm ent p rogram m es, bu t is not happ y b ein g identified as an offender. H e regrets his past offences and w ants to stop offending. These are all positive ind icators o f his read iness for treatm ent. O n the other hand, his responses ind icate that he blam es others rather 'A C VTRQ score of over 72 suggests that an individual displays a capacity to engage in a treatm ent program m e, but this m easure w as only validated for prisoners attending cognitive skills programmes. 226

Readiness and risk: a case illustration

than him self for his offending, feels that he d oesn't deserve his current sentence, and is angry w ith others regarding his sentence (offending beliefs). H e also does not see him self as well organised, hates being told w hat to do, and generally does not trust other people (efficacy). These factors present barriers to his current treatment readiness. An assessm ent of this type can then be linked back to the form ulation of his offending. It identifies core beliefs about him self and other people (for exam ple, his sense of entitlem ent and injustice) as representing significant barriers to any positive and genuine engagem ent in a program m e. Perhaps even m ore im portant from a readiness perspective is the detection of his core values and associated identity. It is likely that M r Jones has a num ber of potentially m aladaptive schem as that will influence both his treatm ent readiness and his treatment needs. These include schem as related to disconnection and rejection, such as aband onm ent/instability, which lead him to believe that others will not be able to provide em otional support and practical protection. Related schem as of m istru st/ab u se and em otional deprivation may also be present, with the expectation that one's desire for norm al degrees o f em otional support will not be met. A d efectiv eness/sham e schem a m ay also be present in som e form , with the feeling that he is defective, inferior and unwanted, or that he would be found to be unlovable by significant others if his true self is exposed. Related to all of these m ay be a schem a involving social isolation /alienation, in which M r Jones feels isolated from the rest of the world, and different from other people. O ther potential schem as for M r Jones include those related to impaired autonomy, such as d ep en den cy/in com p eten ce schem as (the belief that one is unable to handle one's everyday responsibilities and take care of oneself); failure (the belief that one has failed and will inevitably fail, one is fundam entally inadequate relative to others in achievem ent, that one is stupid, inept and untalented); and an insufficient self-control/self-d iscipline schem a (a pervasive difficulty in exercising self-control and frustration tolerance, or to restrain one's im pulses). Given this, it m ight be possible to predict that Mr Jones would w illingly attend treatm ent sessions, and present him self as som eone w ho is genuinely interested in changing his behaviour, and w illing to learn from others in the treatm ent group. However, he is likely to experience significant difficulties in full and frank self-disclosure (of the type required in offender treatm ent), and this m ay be understood in terms of a pervasive lack of confidence that others will be w illing 227

Transitions to Better Lives

to support him emotionally, or perhaps will judge him negatively. It may be expected, therefore, that w hile he m ay be able to form an apparently strong therapeutic alliance w ith program m e staff in the early stages of treatm ent, significant ruptures will occur as soon as he is challenged about those aspects of his offending that he finds uncom fortable or sham eful. A t this tim e he m ight experience program m e staff as punitive and uncaring, and becom e angry about the requirem ent for him to attend. M r Jones is also a person w ho is likely to vacillate betw een being friendly and cooperative and then being hostile, followed by apologies to becom e accepted again. A s he is often uncom fortable w ith others' anger, he m ay also resort to covert expressions of hostility, such as passive obstructionism . It is not difficult to predict that as a direct consequence of this interpersonal style he will experience problem s in his relationships w ith other group m em bers, although he may also seek to intim idate m em bers o f the group w hom he considers to be less powerful. In addition, while his PCL-R scores were not sufficiently elevated to w arrant his categorisation as highly psychopathic, M r Jones did score highly on item s that will affect his treatm ent readiness. The personality and interpersonal characteristics that Mr. Jones has been assessed as dem onstrating, such as lying, m anipulation, lack of rem orse and empathy, im pulsivity and irresponsibility, m ay m ake it m ore difficult for M r Jones to internalise and consistently utilise the gains he has m ade in treatment. H e m ay also believe that ultimately, treatm ent is likely to be ineffective, given that he experiences so little control over his behaviour. These considerations m ake it im perative that things that m atter to him are identified and m ade a focus of any intervention plan. It is m uch easier to m otivate people to engage in the process of behaviour change if they believe their needs and values have been sufficiently addressed and are built into an intervention plan (Ward and M aruna 2007). Traditional risk reduction treatment plans are unlikely to appeal to som eone like M r Jones.

External readiness factors There are a num ber of external factors that m ight also influence M r Jones's readiness to effectively engage in treatment at this stage. O ne set of factors relates to the circum stances in which treatment will occur. Although treatment is not legally mandated for M r Jones, it is apparent that the parole board will not seriously consider his 228

Readiness and risk: a case illustration

release until he has more satisfactorily com pleted a specialised sex offender treatment programm e. Thus there is considerable pressure, if not outright coercion, to engage in treatment. To the degree that treatment goals are congruent w ith M r Jones's personal goals, and the recognition that successful com pletion of a treatment programm e will be beneficial in personally m eaningful ways, his perception of coercion and the attendant im pedim ents to treatment readiness may be minimised. This means that clinicians need to take the time to sit down w ith M r Jones and enquire about his interests and aspirations, and how the personal concerns evident in his past offending can be translated into treatment goals and form the basis of a good lives plan. A nother external readiness factor relates to the availability of adequate treatm ent resources. Engaging offenders such as Mr Jones in treatm ent is a task that requires great therapeutic skill. Therefore, one of the resource factors that will influence M r Jones's treatment readiness will be the availability of well-trained and highly experienced treatm ent staff w ith the skills to w ork w ith offenders who present with his particular challenging risk factors, personality features and prior treatm ent experiences. The ability to engage w ith M r Jones in a w arm , em pathic, respectful and rew arding manner, especially w hen he becom es aggressively defensive, evasive or m anipulative, will have a substantial im pact on his initial levels of treatment readiness; as will his ability to rem ain engaged in treatm ent when it becom es more directly focused on the personal factors that contribute to his risk of reoffending. These therapeutic skills will, of course, also interact with other, internal readiness factors, such as M r Jones's enduring m istrust of others. There are also program m e factors that will influence M r Jones's treatment readiness. Particularly in light of his previous treatment experiences with lim ited success, there will be a need for flexibility in the identification of treatment goals and the application of suitable treatment approaches for M r Jones. If the program m e attem pts to im pose a rigidly defined set of avoidance goals broadly associated w ith reducing risk, and im plem ents treatm ent w ith a strict adherence to treatment m anuals, it is less likely that Mr Jones will becom e engaged in the treatment process. Conversely, to the degree that the program m e staff can collaborate w ith M r Jones to define treatment goals that are individually relevant and valued, and tailor the pace, content and m odality of treatm ent delivery to m atch his individual needs, treatm ent readiness is likely to be m axim ised. A final set of external readiness factors relate to opportunity, 229

Transitions to Better Lives

location and support for treatm ent. O pportu nities for M r Jones to d em onstrate and p ractise the treatm ent gains he desires to m ake in an ecologically valid m an ner w ill be severely lim ited w hile he rem ains incarcerated. This is one o f the d ilem m as facing those responsible for the treatm ent and supervision o f sex offenders: the effectiveness o f treatm ent in custod ial setting s m ay be low er than treatm ent in com m u nity settings (perhaps p artly because o f the m ore coercive nature of the custodial environm en t and the lack o f ecological v alid ity to practising treatm ent gain s in that setting), b u t offend ers are often not considered safe to release to the com m u nity until they hav e d em onstrated sufficient treatm ent gains. It w ould probably be beneficial if M r Jones is provided w ith the o pportu nity to m ove into a supervised com m u nity settin g at the earliest stage of the treatm ent process, ideally through a series o f progressive tem porary com m unity v isits resultin g in som e form o f con d itional release to a supported com m u nity treatm ent setting. Finally, M r Jones, like m any o ffend ers, is largely estranged from family, m ost of w hom now live som e distance from the prison w here he resid es, and he has few pro-social friend s. Support for his treatm ent m ust, therefore, com e p rim arily from treatm ent and custod ial staff. To the extent that there is a unified and con sisten tly supportive approach taken by those in his prison environm ent, treatm ent read in ess is likely to b e enhanced. C onversely, if he gets m ixed m essages from different staff, som e of w hom express doubt or even overt d enigration o f the treatm ent program m e, treatm ent read in ess is likely to be d im in ished .

Implications for tre a tm e n t readiness So, how m ight this consideration of treatm ent readiness inform answ ers to the questions raised by the parole board? In light of M r Jo nes's extensive treatm ent history, it is anticipated that little m ore is to be gained from additional intensive residential treatm ent for his sexual offending. H e w ould appear to have acquired, at a cognitive level, an adequate understand ing o f the precursors to his offend ing and the factors associated w ith increased risk o f reoffending. H e has also been taught a variety of cognitive and behaviou ral techniques for m anaging this risk. W hat rem ains is the need for M r Jones to further internalise these treatm ent gains and d em onstrate a consistent adherence to candid, honest, pro-social behaviou r in his daily interactions.

230

Readiness and risk: a case illustration

To assist w ith this, treatm ent providers will need to attend much more closely to the way in which M r Jones relates to others, and provide clear, consistent and supportive feedback about how this m ight reflect beliefs about him self and the world that also contribute to the causes of his offending. They w ill need to expect significant ruptures in the therapeutic alliance to occur, and respond to these in w ays that are both therapeutic and inform ed by the case form ulation. It may, of course, be that this style of w orking is beyond the skills or resources of som e program m es (w hich may, for exam ple, be psychoeducational in nature, involving large groups of offenders), and if this is the case there would be little reason to expect that further participation would be particularly beneficial. It m ay also, of course, com prom ise the treatm ent of other participants. There are other ways in w hich risk can be m anaged outside of a form al treatm ent program m e. A key factor in m inim ising risk in this case, for exam ple, will be the adequacy of his close relationships with appropriate adult partners. The distress, isolation and frustration that have resulted from previous situations in w hich M r Jones has failed to establish adequate relationships w ith other adults have directly contributed to his sexual offences. Therefore, it is im portant for M r Jones to establish and m aintain a strong social netw ork of individuals fam iliar w ith his offence cycle w ho can be actively involved in his transition to an offence-free lifestyle in the community. Setting such a goal is, of course, much m ore com patible with M r Jones's professed desire to form age-appropriate and fulfilling relationships than goals that relate to him not behaving in particular ways. An adequate release or good lives plan for M r Jones could also include consideration of his living situation, with particular attention to the nature of his close interpersonal relationships. The lack of stable, healthy relationships w ith appropriate adult friends or partners has been a significant factor to the isolation and distressing em otions that contribute to his sexual offending. M r Jones also recognises that placing him self in situations w here he is alone, especially if he is driving around in a car with access to potential victim s, will put him at particularly high risk of reoffending. H aving a supportive social network available, and suitable activities accessed through reliable m odes of public transportation, will correspondingly reduce his im m ediate level of risk, especially if these are areas that can be adequately monitored in the com m unity on an ongoing basis.

231

Transitions to Better Lives

Conclusion In this chapter we have attem pted to discuss some of the w ays in which a consideration of treatm ent readiness can not only inform decisions about w hether or not to refer an offender to a particular rehabilitation program m e, but also how an understanding of readiness can also assist in the process o f determ ining treatm ent needs and m anage risk. This is not to suggest that this chapter provides a tem plate for how to w ork w ith offenders, as any assessm ent will be determ ined by the particular context in w hich it occurs and the questions that it seeks to address. The case we developed for use in this chapter was of som eone with a long history of treatm ent experiences, and for w hom significant concerns rem ained about his future risk, despite participation in program m es that was at tim es adequate. H e was referred for assessm ent because of an upcom ing parole board hearing, the results of which would have significant im plications for his future. W hile there are a num ber of com plicating factors in relation to assessing offenders such as Mr Jones, we chose this exam ple because it is not an uncom m on presentation. O ffenders typically arrive for assessm ent w ith histories of attendance at a range of different program m es, beliefs about w hat is reasonable and to be expected from them in treatm ent, and often a desire to negotiate the easiest pathw ay out of the crim inal justice system . Our conclusion here is that any assessm ent of readiness of treatm ent should be closely and inextricably linked to the assessm ent of risk of reoffending and treatment needs, and undertaken with the goal of developing com prehensive case form ulations that allow program m e content to be individualised to m eet the needs of each and every participant. Had treatm ent readiness been considered in the original assessm ent and treatm ent planning process, it is possible that these previous attem pts at rehabilitation m ay have proven more successful. In our view this entails taking M r Jones's personal concerns and values seriously and attem pting to understand his offending and his associated lifestyle from the view point of w hat m atters to him. Strength-based approaches aim first to identify an individual's core values and then to ascertain ju st w hat internal and external resources he requires to secure them in w'ays that also reduce risk. The prospect of a future life that revolves prim arily around risk reduction and avoidance is arguably empty, and unlikely to m otivate an offender to desist from further sexually abusive actions. Focusing instead on the

232

Readiness and risk: a case illustration

pursuit of p ersonally valued goals, and on con structive, collaborative approaches to their achiev em en t, m ay offer a m ore prom ising transition to a better life.

233

Chapter 14

Ways forward and conclusions

W orm with et al. (2007) have suggested that m uch of the future understanding about correctional intervention 'is likely to com e from m oving beyond the sim ple, but im portant, treatm ent-recidivism study, to the exam ination of in-program issues' (2007: 882). They identify non-engagem ent, non-com pletion and drop-out as som e of the m ost critical factors that influence rehabilitative outcom es. In this book we have described and discussed how treatm ent readiness is likely to be an im portant determ inant of program m e engagem ent and com pletion, and offered a framew'ork from which to understand and assess readiness, as well as to inform interventions that m ight be used to m odify or address low levels of readiness. It is evident to us, in w riting this book, that w hile m anaging offenders who are not ready for rehabilitation is clearly a difficult and challenging task for those involved in program m e delivery, there is a very limited evidence base from w hich to inform their practice. A lthough m uch has been w ritten about m anaging resistance, and responding to low levels o f m otivation, few em pirical studies have been conducted w ith offenders w ho are referred to offence-specific rehabilitation program m es. This includes research on both the level of readiness (and how this m ight inform referral to program m es), and on the im pact of attem pts to im prove readiness on program m e outcomes. A prerequisite for such research, however, is the developm ent of m easures that can reliably m easure treatm ent readiness, that have established norm s and cut-off points, and that can be used to assess changes in readiness over time. Som e prom ising approaches to

234

Ways forward and conclusions

assessing readiness have been developed, but m uch more w ork is needed to validate these measures. The challenge for practitioners is to w ork in w ays that most closely m atch the needs and aspirations o f offenders to those of the crim inal justice system , and by extension the broader community. It is reasonable to expect those w ho have com m itted serious offences to consider seriously the w ays in which they m ight be able to change their behaviour. It is also to be expected that m any will not see the need to do this, or have confidence that the program m es and services that are offered to them w ill be able to meet these needs. In our view finding new w ays to understand the barriers that stand in the way of effective rehabilitation, and allow offenders to becom e more m eaningfully engaged in a process of behaviour change, can only lead to better program m e outcom es and improved com m unity safety.

235

Appendix

Measures ofTreatment Readiness

U niversity of Rhode Island C hange A ssessm ent Citation: M cC onnaughy, F. A., Prochaska, J. O. and Velicer, W. F. (1983) 'Stages of change in psychotherapy: M easurem ent and sam ple profiles', Psychotherapy: Theory, Research and Practice, 20: 368-75. Instructions: Please read each statem ent belo w carefully and then decide w hether you agree or disagree w ith each statem ent. C ircle the num ber that best represents how you feel. Please circle one and only one num ber for every statem ent. -o 60 m C O

M) n

2. 3.

4. 5.

236

As far as I'm concerned, 1 d on't have any problem s that need changing. 1 think I m ight be ready for som e self-im provem ent. I am d oing som ething about the problem s that had been bothering me. It m ight b e w orthw hile to w ork on m y problem . I'm not the problem one. It d oesn't m ake m u ch sense for m e to be here.

3

a.

S Q

* S


(/) «

0 © © ©

© ©

© ©

© ©

© ©

© ©

© ©

© ©

©

©

©

©

© ©

£ 2

tt>

" "O

“ “ (/} T3 1.

tj,

® £

C o

4» £

Appendix

6.

7. 8.

9.

10. 11.

12. 13.

14. 15. 16.

17.

18.

19. 20. 21.

Ii w orries m e that I m ight slip back on a problem I have already changed, so I am here to seek help. I am finally d oing som e w ork on my problem . I'v e been thinking that I m ight w ant to change som ething about m yself. I have been successful in w orking on m y problem b u t I'm not sure I can keep up the effort on m y own. At tim es m y problem is difficult, but I'm w orkin g on it. Being here is pretty m uch a w aste of tim e for m e becau se the problem d oesn't have to do w ith me. I'm hoping this place will help me to better understand m yself. I guess I have faults, but there's nothing that I really need to change. I am really w orking hard to change. I have a problem and I really think I should w ork at it. I'm not follow ing through with w hat I had already changed as w ell as I had hoped, and I'm here to prevent a relapse o f the problem . F.ven though I'm not alw ays successfu l in changing, 1 am at least w orking on m y problem . 1 thought once I had resolved my problem I w ould be free of it, but som etim es I still find m yself struggling w ith it. I w ish I had m ore ideas on how to solve the problem . I have started w orking on m y problem s bu t I w ould like help. M aybe this place w ill be able to help me.

© ©

© ©

© ©

© ©

© ©

©

©

©

©

©

© ©

© ©

© ©

© ©

© ©

© ©

© ©

© ©

© ©

© ©

© © ©

© © ©

© © ©

© © ©

© © ©

©

©

©

©

©

©

©

©

©

©

© © © ©

© © © ©

© © © ©

© © © ©

© © © © 237

Transitions to B e tte r Lives

22. 1 m ay need a bo ost right now to help m e m aintain the changes I've already m ade. 23. I m ay be part of the problem , but I d on't really think I am . 24. I hope that som eone here will have som e good advice for m e. 25. A nyone can talk about changing; I'm actually doing som ething about it. 26. All this talk about psychology is boring. W hy c a n 't people just forget about their problem s? 27. I'm here to prevent m yself from having a relapse of m y problem . 28. It is frustrating, but I feel I m ight be having a recurrence of a problem I thought I had resolved. 29. 1 have w orries bu t so does the next guy. W hy spend tim e thinking abou t them ? 30. I am actively w orkin g on m y problem . 31. I w ould rather cope w ith m y faults than try to change them. 32. A fter all I had done to try to change m y problem , every now and again it com es back to haunt me.

© ® ©

© © ©

© © ©

© © ©

© © ©

©

©

©

©

©

© ©

© ©

© ©

© ©

© ©

©

©

©

©

©

© © ©

© © ©

© © ©

© © ©

© © ©

©

©

©

©

©

Scoring p ro to co l

P recontem plation item s:

1, 5, 11, 13, 23, 26, 29, 31

C ontem plation item s:

2, 4, 8, 12, 15, 19, 21, 24

A ction item s:

3, 7, 10, 14, 17, 20, 25, 30

M aintenance item s:

6, 9, 16, 18, 22, 27, 28, 32

To obtain a Readiness to C hange score, first sum item s from each subscale and divide b y 7 to obtain the m ean for each subscale. Then sum the m eans from the C ontem plation, A ction, and M aintenance subscales and subtract the P recontem plation m ean (C + A + M - PC = Readiness). Scores of 8 and less are classified as Precontem plators, 8 - > l l are Contem plators, and 11->14 are P reparation to A ction Takers. (For further details see w w w .u ri.e d u /re se arch /cp rc/m e asu res.h tm )

238

Appendix

Readiness to Change Q uestionnaire Citation: Rollnick, S., H eather, N ., G old, R. and H all, W. (1992) 'D evelopm ent of a short readiness to change questionnaire for use in b rief opportunistic interventions am ong excessive d rinkers', British Journal of Addiction, 87: 7 4 3 54. Instructions: Please read each question below carefully and then decide w hether you agree or disagree w ith the statem ents. Please circle the answ er of your choice to each question. If you have any problem s please ask the questionnaire adm inistrator. Key: S D = S tro n g ly S A = S tro n g ly a g re e

d isa g re e

D

=

I d on't think that I drink too m uch. 2. I am trying to d rink less than I used to. 3. 1 enjoy m y drinking, but som etim es I drink too much. 4. Som etim es I think I should cut dow n on m y drinking. 5. It's a w aste of tim e thinking about m y drinking. 6. I have ju st recently changed m y drinking habits. 7. Anyone can talk about w anting to do som ething abou t drinking, but I am actually d oing som ething about it. 8. 1 am at the stage w here I should think about drinking less alcohol. 9. M y drinking is a problem som etim es. 10. T here is no need for m e to think about changing m y drinking. 11. I am actually changing my drinking habits right now. 12. D rinking less alcohol w ould be pointless for me.

D isa g re e

U =

U n su re

A =

A g ree

1.

SD

D

U

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

239

Transitions to Better Lives

Scoring protocol P recontem plation item s: 1, 5, 10, 12 C ontem plation item s:

3, 4, 8, 9

Action item s:

2, 6, 7, 11

To calculate the score for each scale, sim ply add the item s score for the scale in question. T h e ran ge on each scale is - 8 through 0 to +8. A negative scale score reflects an overall disagreem ent w ith item s m easuring the stage of change, w hereas a positive score represents overall agreem ent. T h e highest score represents the Stage o f C hange Designation. Note: If tw o or m ore scale scores are equal, then the scale farthest along the continuum of change (P recontem p latio n-C on tem p lation-A ctio n) represents the Stage of C h ange Designation. For exam ple, if a som eone scores 6 on the Precontem plation scale, 6 on the C ontem plation scale, and - 2 on the A ction scale, then that person is assigned to the C ontem plation stage. If one o f the four item s on a scale is m issing, the ind ividu al's score for that scale should be pro-rated (i.e. m ultiplied by 4 / 3 or 1.33). If tw o or more item s are m issing, the scale score cannot be calculated. In this case, the Stage of C hange D esignation w ill be invalid.

240

Appendix

Revised Readiness to Change Q uestionnaire [T re a tm e n t V ersio n ]

Citation: H eather, N. and H onekopp, J. (2008) 'A revised edition of the Readiness to C hange Q uestionnaire [Treatment Version]', Addiction Research and Theory, 16: 421-33. Instructions: The follow ing questions are designed to identify how you personally feel about your drinking right now. Please think about your current situation and drinking habits, even if you given up drinking com pletely. Read each question below carefully and then decide w hether you agree or disagree w ith the statem ents. Please circle the answ er o f your choice to each question. If you have any problem s please ask the questionnaire adm inistrator. K ey : S D

=

S tro n g ly d isa g re e

D =

D isa g re e U =

U n s u re A

=

A g re e

S A = S tro n g ly a g re e

1.

It's a w aste o f tim e thinking about m y drinking becau se I do not have a problem . 2. I enjoy m y drinking but som etim es I drink too much. 3. There is nothing seriously w rong w ith m y drinking. 4. Som etim es I think I should quit or cut dow n on my drinking. 5. Anyone can talk about w anting to do som ething abou t their drinking, but I'm actually doing som ething about. 6. 1 am a fairly norm al drinker. 7. M y drinking is a problem som etim es. 8. I am actually changing my drinking habits right now (either cutting dow n or quitting). 9. I have started to carry out a plan to cut dow n or quit drinking. 10. There is nothing I really need to change abou t m y drinking. 11. Som etim es I w ond er if my drinking is out of control. 12. I am actively w orking on m y drinking problem s.

SD

D

U

A

SA

SD

D

U

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

SA

D

u u

A

SD

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

SD

D

u

A

SA

241

Transitions to Better Lives

Scoring protocol Precontem plation item s:

1, 3,

6,

10

Contem plation item s:

1, 4,

7,

11

Action item s:

5, 8,

9,

12

Follow the scoring protocol for the Readiness to C hange Q uestionnaire.

242

Appendix

Personal D rinking Q uestionnaire (S O C R A T E S 8A )

Citation: M iller, W. R. and Tonigan, J. S. (1996) 'A ssessing d rinkers' m otivation for change: The Stages o f C hange Readiness and Treatm ent Eagerness Scale (SO CRA TES)', Psychology of Addictive Behaviors, 10: 81 -9 . Instructions: Please read the follow ing statem ents carefully. E ach one describes a w ay that you m ight (or m ight not) feel about your drinking. For each statem ent, circle one num ber from 1 to 5, to indicate how m uch you agree or disagree w ith it righ t now. Please circle one and only one num ber for every statem ent. -o Ji' v HO h e ® O n £ .2

tr> -o 1. I really w ant to m ake changes in m y drinking. 2. Som etim es I w ond er if I am an alcoholic. 3. If I d on't change m y drinking soon, m y problem s are going to get w orse. 4. I have already started m aking som e changes in m y drinking. 5. I w as drinking too m uch at one lim e, b u l I'v e m anaged to change m y drinking. 6. Som etim es I w ond er if my d rinking is hu rlin g other people. 7. I am a problem drinker. 8. I'm not ju st thinking about changin g m y drinking, I'm already d oing som ething about it. 9. I have already changed my d rinking, and I am lookin g for w ays to keep from slipping back to m y old pattern. 10. I have serious problem s w ith drinking. 11. Som etim es I w ond er if I am in control of m y drinking.

ai ¡3, ® $ « Q

'C

v c

"O D

«.

£ Sb S