Qualitative Research in Hospitality and Tourism 9781781901458, 9781781901441

Citation preview

ijchm cover (i).qxd

01/10/2012

09:44

Page 1

ISSN 0959-6119

Volume 24 Number 6 2012

International Journal of

Contemporary Hospitality Management Qualitative research in hospitality and tourism Guest Editors: Amit Sharma and Levent Altinay

www.emeraldinsight.com

International Journal of

ISSN 0959-6119

Contemporary Hospitality Management

Volume 24 Number 6 2012

Qualitative research in hospitality and tourism Guest Editors Amit Sharma and Levent Altinay

Access this journal online _______________________________ 815

CONTENTS

Editorial advisory board _________________________________ 816 Editorial ___________________________________________________ 817 Guest editorial ____________________________________________ 818 Use of qualitative research in foodservice organizations: a review of challenges, strategies, and applications Susan W. Arendt, Kevin R. Roberts, Catherine Strohbehn, Jason Ellis, Paola Paez and Janell Meyer ______________________________________

820

Investigative management and consumer research on the internet Peter Lugosi, Hania Janta and Pamela Watson _______________________

838

Structure and process modeling of seemingly unstructured leisure-travel decisions and behavior Drew Martin and Arch G. Woodside________________________________

855

Using focus groups as a tool to develop a hospitality work-life research study John W. O’Neill _________________________________________________

873

Consulting the oracle? Applications of modified Delphi technique to qualitative research in the hospitality industry Abu Elnasr E. Sobaih, Caroline Ritchie and Eleri Jones _________________

886

Beyond consensus: an alternative use of Delphi enquiry in hospitality research Alexandros Paraskevas and Mark N.K. Saunders _____________________

907

This journal is a member of and subscribes to the principles of the Committee on Publication Ethics

CONTENTS continued

How to promote cooperation in the hospitality industry: generating practitioner-relevant knowledge using the GABEK qualitative research strategy Harald Pechlaner and Michael Volgger ______________________________

925

RESEARCH IN BRIEF Reflecting on a phenomenological study of creativity and innovation in haute cuisine Marc B. Stierand and Viktor Do¨rfler _______________________________

946

Book review _______________________________________________ 958

www.emeraldinsight.com/ijchm.htm As a subscriber to this journal, you can benefit from instant, electronic access to this title via Emerald Management eJournals. Your access includes a variety of features that increase the value of your journal subscription.

Structured abstracts Emerald structured abstracts provide consistent, clear and informative summaries of the content of the articles, allowing faster evaluation of papers.

How to access this journal electronically

Additional complementary services available

Our liberal institution-wide licence allows everyone within your institution to access your journal electronically, making your subscription more cost-effective. Our web site has been designed to provide you with a comprehensive, simple system that needs only minimum administration. Access is available via IP authentication or username and password.

When you register your journal subscription online you will gain access to additional resources for Authors and Librarians, offering key information and support to subscribers. In addition, our dedicated Research, Teaching and Learning Zones provide specialist ‘‘How to guides’’, case studies, book reviews, management interviews and key readings.

To benefit from electronic access to this journal, please contact [email protected] A set of login details will then be provided to you. Should you wish to access via IP, please provide these details in your e-mail. Once registration is completed, your institution will have instant access to all articles through the journal’s Table of Contents page at www.emeraldinsight.com/ 0959-6119.htm More information about the journal is also available at www.emeraldinsight.com/ijchm.htm

E-mail alert services These services allow you to be kept up to date with the latest additions to the journal via e-mail, as soon as new material enters the database. Further information about the services available can be found at www.emeraldinsight.com/alerts

Emerald online training services Visit www.emeraldinsight.com/help and take an Emerald online tour to help you get the most from your subscription.

Emerald Research Connections An online meeting place for the world-wide research community, offering an opportunity for researchers to present their own work and find others to participate in future projects, or simply share ideas. Register yourself or search our database of researchers at www.emeraldinsight.com/connections

Key features of Emerald electronic journals

Choice of access

Automatic permission to make up to 25 copies of individual articles This facility can be used for training purposes, course notes, seminars etc. within the institution only. This only applies to articles of which Emerald owns copyright. For further details visit www.emeraldinsight.com/copyright Online publishing and archiving As well as current volumes of the journal, you can also gain access to past volumes on the internet via Emerald Management eJournals. You can browse or search these databases for relevant articles. Key readings This feature provides abstracts of related articles chosen by the journal editor, selected to provide readers with current awareness of interesting articles from other publications in the field. Non-article content Material in our journals such as product information, industry trends, company news, conferences, etc. is available online and can be accessed by users. Reference linking Direct links from the journal article references to abstracts of the most influential articles cited. Where possible, this link is to the full text of the article. E-mail an article Allows users to e-mail links only to relevant and interesting articles to another computer for later use, reference or printing purposes.

Electronic access to this journal is available via a number of channels. Our web site www.emeraldinsight.com is the recommended means of electronic access, as it provides fully searchable and value added access to the complete content of the journal. However, you can also access and search the article content of this journal through the following journal delivery services: EBSCOHost Electronic Journals Service ejournals.ebsco.com Informatics J-Gate www.j-gate.informindia.co.in Ingenta www.ingenta.com Minerva Electronic Online Services www.minerva.at OCLC FirstSearch www.oclc.org/firstsearch SilverLinker www.ovid.com SwetsWise www.swetswise.com

Emerald Customer Support For customer support and technical help contact: E-mail [email protected] Web http://info.emeraldinsight.com/products/ subs/customercharter.htm Tel +44 (0) 1274 785278 Fax +44 (0) 1274 785201

IJCHM 24,6

816

EDITORIAL ADVISORY BOARD Dr Albert Assaf University of Massachusetts, USA

Dr Johye Hwang Kyung Hee University, Korea

Dr Haemoon Oh Iowa State University, USA

Dr Billy Bai University of Nevada, Las Vegas, USA

Dr SooCheong (Shawn) Jang Purdue University, USA

Professor Michael C. Ottenbacher Heilbronn University, Germany

Dr Clayton W. Barrows University of New Hampshire, USA

Professor Vinnie Jauhari Institute for International Management and Technology, India

Professor Peter O’Connnor ESSEC Business School, France

Dr Soyoung Boo The George Washington University, USA Dr John T. Bowen University of Houston, USA Dr Wilco Chan Hong Kong Polytechnic University, Hong Kong Prakash K. Chathoth American University of Sharjah, United Arab Emirates Dr Huei-Ju (Margaret) Chen National Kaohsiung University of Hospitality and Tourism, Taiwan (Republic of China)

Dr Jorge Pereira-Moliner Universidad de Alicante, Spain Professor Nina Katrine Prebensen University of Tromso, Norway

Dr Lee Jolliffe University of the West Indies, Barbados

Professor Arie Reichel Ben-Gurion University of the Negev, Israel

Professor Jay Kandampully Ohio State University, USA Dr Osman M. Karatepe Eastern Mediterranean University, Turkey

Professor Cihan Cobanoglu University of South Florida, SarasotaManatee, USA

Professor Karolin Kokaz Ecole hoˆtelie`re de Lausanne, Switzerland

Professor John S.A. Edwards Bournemouth University, UK Dr Yuksel Ekinci Oxford Brookes University, UK Dr Erdogan Ekiz Al-Faisal University, Saudi Arabia Dr Mehmet Erdem University of Nevada, USA

Professor John Parnell University of North Carolina, USA

Professor Nick Johns CTF Consultants, UK

Professor Woody G. Kim Florida State University, USA

Dr Robin Dipietro University of South Carolina, USA

Professor A. Parasuraman University of Miami, USA

Dr Misty M. Johanson DePaul University, USA

Dr Po-Ju Chen University of Central Florida, USA

Dr Anne Crick University of the West Indies, Barbados

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 p. 816 # Emerald Group Publishing Limited 0959-6119

Dr Stuart A. Jauncey The Emirates Academy, United Arab Emirates

Professor Conrad Lashley Nottingham Trent University, UK Associate Professor Darren Lee-Ross James Cook University, Australia Associate Professor Xiang (Robert) Li University of South Carolina, USA Professor Stephen W. Litvin College of Charleston, USA Dr Peter Lugosi Oxford Brookes University, UK

Philippe Rossiter FIH Institute of Hospitality, UK Professor Chris Ryan Waikato Management School, New Zealand Dr Atul Sheel University of Massachusetts, USA Professor Stowe Shoemaker University of Houston, USA Professor Haiyan Song Hong Kong Polytechnic University, Hong Kong Professor Christine Sun National Taiwan Normal University, Taiwan (Republic of China) Dr Masako Shibata Taylor Osaka Gakuin University, Japan Professor Rhodri Thomas Leeds Metropolitan University, UK Professor Jean-Pierre van der Rest International University of Hospitality Management, The Netherlands

Dr Christina Geng-qing Chi Washington State University, USA

Dr Vincent P. Magnini Virginia Polytechnic Institute and State University, USA

Professor Christopher Guilding Griffith University, Australia

Dr Melih Madanoglu Florida Atlantic University, USA

Associate Professor Murat Hancer Oklahoma State University, USA

Professor Anna S. Mattila Pennsylvania State University, USA

Dr Robert J. Harrington University of Arkansas, USA

Dr Karl Mayer University of Nevada, USA

Professor Nigel R. Hemmington Auckland University of Technology, New Zealand

Professor Josef Mazanec Vienna University of Economics and Business Administration, Austria

Joan C. Henderson Nanyang Technological University, Singapore

Dr Asad Mohsin The University of Waikato Management School, New Zealand

Professor J.S. Perry Hobson Taylor’s University, Malaysia

Dr Marco Mongiello Imperial College, UK

Professor Lishan Xie Sun Yat-Sen University, People’s Republic of China

Dr Demian Hodari Ecole hoˆtelie`re de Lausanne, Switzerland

Professor Richard Nicholls Poznan University of Economics, Poland

Dr Atila Yuksel Adnan Menderes University, Turkey

Professor Cathy Hsu The Hong Kong Polytechnic University, Hong Kong

Dr Khaldoon Nusair University of Central Florida, USA

Dr Anit Sharma Penn State University, USA

Dr Nan Hua University of Houston, USA

Dr Kevin O’Gorman University of Strathclyde, UK

Dr Mathilda van Niekerk The University of Central Florida, USA Dr Serena Volo Free University of Bozen-Bolzano, Italy Dr Sandra Watson Edinburgh Napier University, UK Dr Karin Weber Hong Kong Polytechnic University, Hong Kong Professor Arch Woodside Boston College, Carroll School of Management, USA

Editorial

Editorial Welcome to IJCHM’s special issue on qualitative research in hospitality and tourism. I would like to pay special thanks to our Guest Editors Amit Sharma and Levent Altinay for putting together this very strong and timely special issue. The articles included in this special issue should be well received by scholars, students and practicing managers in our field.

817

Fevzi Okumus Editor-in-Chief

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 p. 817 q Emerald Group Publishing Limited 0959-6119

IJCHM 24,6

818

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 pp. 818-819 q Emerald Group Publishing Limited 0959-6119

Guest editorial This special issue of IJCHM on qualitative research methods includes seven full research articles, one research in brief, and a book review. The articles included in this special issue are all unique in their application and in-depth presentation of qualitative research methods. We are excited to present these articles, which aim to extensively and systematically elaborate the use of qualitative methods in hospitality and tourism research. In the first article, Susan Arendt and her colleagues present the challenges encountered when conducting qualitative research in foodservice operations. Strategies to overcome challenges are identified and discussed. The authors conducted food safety observations, interviews, and focus groups with foodservice employees and managers, and encountered multiple challenges including institutional review board approvals, and managements’ willingness to participate and organizational and cultural barriers. Authors point out that obtaining in-depth, credible information through observations, interviews, and focus groups adds depth and breadth to hospitality studies. However, given high industry turnover, recruitment and retention throughout a study is problematic. Moreover, authors encounter many barriers as they obtain data, such as establishing authenticity and overcoming Hawthorne and halo effects. In the second article Peter Lugosi and his colleagues introduce the notion of Investigative Research on the Internet (IRI) and conceptualises its processes through the principle of streaming. This paper discusses the similarities and differences between IRI and netnography and considers various aspects of the IRI process, including site selection, sampling, data collection and analysis. Investigative internet-based research uses the techniques of ethnography and netnography, including variations of participant observation and analysis of visual and textual material. The paper shows that IRI has a number of potential applications for hospitality management academics and practitioners. Streaming can help to understand the processes involved in conducting netnographic research, and streaming is a more appropriate way to conceptualise some internet-based studies that do not conform to netnographic or ethnographic ideals. In the next article Drew Martin and Arch Woodside introduce a structure and process model (SPM) as a framework for tourism decision making research. The authors show how SPM enables mapping and comparing visitors’ plans, motivations, choices, and consequences. The results demonstrate nuanced decision-behavior dynamics and complexities of visitors’ travel-related unconscious/conscious thinking and behavior. In the fourth article John O’Neill seeks to describe how the results of the qualitative research method of focus groups may be used as conceptual data at the onset of a research study to inform researchers regarding relevant issues for future more in-depth quantitative study. Seven focus group sessions with the homogeneous participants of new entrants to the hospitality industry were conducted. The paper suggests that long, unpredictable hours create both work-related and non-work stress. Further, there is general agreement regarding the stressors and benefits associated with working in the hospitality industry. In the fifth article Abu Elnasr Sobaih and his colleagues use the Delphi technique to achieve consensus among experts and/or gain judgment on

complex matters. This paper discusses the classical Delphi and its advantages/disadvantages in qualitative research, particularly in hospitality. This paper shows how the classical Delphi can be successfully modified to overcome its limitations. The authors find that identifying potential problems early in the research process enables critical design decisions to be made to eliminate them. Results of the study are presented in two cases. The first case uses a parallel expert group with similar experience to develop a research instrument for a limited number of prestigious experts well acquainted with each other who might have reached specious consensus through channels not accessible to the researcher. Case two enables the addition of new experts to an expert panel to overcome attrition in successive Delphi rounds. In the next article Alexandros Paraskevas and Mark Saunders evaluated the use of Delphi techniques in qualitative research for utilising “expert” opinions and explore how Policy Delphi can be used by hospitality researchers as an alternative to the more widely used Normative Delphi. This paper reflects on the research methodology of a project that explored organisational crisis signals detection using Policy Delphi with a criterion sample comprising senior hotel executives involved in crisis management. The paper presents a valuable research tool that has evaded the attention of many hospitality researchers offering an illustrative example of its use in exploratory research to deliver credible, transferable and confirmable findings. In the final article Harald Pechlaner and Michael Volgger inductively develop propositions regarding the promotion of local and regional cooperation in the hospitality industry at the network level in order to evaluate the relative importance of the two conflicting positions, and to demonstrate the suitability of GABEK to the development of these propositions. Following a qualitative case study design, the authors gather data by conducting open interviews in a South Tyrolean destination, and analyze with the aid of the GABEK technique. The results suggest that the structural and procedural conditions are relatively more critical to the promotion of interorganizational cooperation in the hospitality industry. The research note by Marc Stierand and Viktor Do¨rfler aims to present and reflect on a phenomenological research process used to elucidate the nature of creativity and innovation in haute cuisine. The authors employ in-depth unstructured interviews and field notes capturing subjective experiences to elucidate the experiences of top chefs from the UK, Spain, France, Austria and Germany with regards to creativity and innovation. The findings presented in this paper are twofold: first, an empirical sample finding is presented in order to contextualize the type of findings obtained; second, key methodological findings are presented explaining the process of elucidating the nature of creativity and innovation through iterative learning from the descriptions of the interviewees and the subjective experiences gathered. Alfonso Vargas-Sa´nchez reviews the book titled “Research Themes for Tourism”, Edited by Peter Robinson, Sine Heitmann and Peter Dieke, and published by CABI (Wallingford, UK). The reviewer highlights the 20 chapters in this book and 20 issues with some common characteristics: they are all contemporary and under-researched (some of them still “emerging”), with good potential for generating new knowledge and for inspiring both academics and practitioners to do more. We hope that our readers find all the articles published in this issue timely, relevant and useful. Amit Sharma and Levent Altinay Guest Editors

Guest editorial

819

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0959-6119.htm

IJCHM 24,6

820 Received 31 August 2010 Revised 14 February 2011 25 March 2011 Accepted 6 April 2011

Use of qualitative research in foodservice organizations A review of challenges, strategies, and applications Susan W. Arendt Department of Apparel, Educational Studies, and Hospitality Management, Iowa State University, Ames, Iowa, USA

Kevin R. Roberts Department of Hospitality Management and Dietetics, Kansas State University, Manhattan, Kansas, USA

Catherine Strohbehn Department of Apparel, Educational Studies, and Hospitality Management, Iowa State University, Ames, Iowa, USA

Jason Ellis Department of Agricultural Leadership, Education and Communication, University of Nebraska-Lincoln, Lincoln, Nebraska, USA

Paola Paez School of Nutrition, University of Costa Rica, Sabanilla Montes de Oca, Costa Rica, and

Janell Meyer Department of Apparel, Educational Studies, and Hospitality Management, Iowa State University, Ames, Iowa, USA Abstract

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 pp. 820-837 q Emerald Group Publishing Limited 0959-6119 DOI 10.1108/09596111211247182

Purpose – The purpose of this paper is to present the challenges encountered when conducting qualitative research in foodservice operations and to discuss the strategies to overcome the identified challenges. Design/methodology/approach – The researchers conducted food safety observations, interviews, and focus groups with more than 600 foodservice employees and managers. The researchers encountered multiple challenges including institutional review board approvals, managements’ willingness to participate, and organizational and cultural barriers. Findings – Obtaining in-depth, credible information through observations, interviews, and focus groups adds depth and breadth to hospitality studies. However, given high industry turnover, recruitment and retention throughout a study is problematic. Moreover, researchers encounter many barriers as they obtain data, such as establishing authenticity and overcoming Hawthorne and halo effects. Originality/value – Strategies to increase participation and thereby improve qualitative research have not been previously addressed in the hospitality literature Keywords Focus groups, Food safety, Foodservice, Interviews, Observations, Qualitative research, Service operations Paper type Research paper

Introduction The debate has continued in research circles whether qualitative research should be valued in the same manner as quantitative research (Gelo et al., 2008; Mays and Pope, 1995; Smith and Heshusius, 1986). Questions of rigor and worthiness generally surface during these discussions. Despite the criticisms, proper, rigorous, and ethically conducted qualitative research methods can be considered when: . investigating complex phenomena that are difficult to measure quantitatively; . generating data necessary for a comprehensive understanding of a problem; and . studying outcome variables as well as process variables.

Use of qualitative research 821

Investigating the context in which behaviors take place is also important in gaining insights into potential causal mechanisms, developing sound quantitative measurement instruments, and studying special populations (Achterberg and Arendt, 2008; Curry et al., 2009). The aim of this paper is to provide other researchers with some of our “lessons learned” in conducting qualitative research while ensuring appropriate rigor. Review of literature Qualitative and quantitative research Qualitative research is not about the numbers, meaning it is neither about sample size nor about data being represented graphically. Rather, the qualitative researcher seeks to understand the depth and breadth of a topic area through rigorous study of phenomena by critically selecting participants, studying those participants thoroughly, and continuing data collection until no new themes emerge during data analysis. Additionally, data are often expressed descriptively as verbiage compared to more quantitative means, numbers. Table I summarizes the major differences between qualitative and quantitative research. The lingo used in qualitative research helps define it. The terminology used to describe qualitative research and the processes used is different from that of quantitative research. For example, qualitative researchers will use terms such as participants or informants whereas quantitative researchers refer to a sample or subjects. Likewise, a qualitative researcher will address the trustworthiness of the data whereas a quantitative researcher will address validity. Proper use and application of qualitative terminology helps distinguish a novice from an expert qualitative researcher. Although some have viewed qualitative research as a distinct type of research, others view it along a continuum with purely qualitative (i.e. ethnographic study) on

Purpose Process Sample size Sample selection Setting data Analysis

Qualitative

Quantitative

Develop theory Inductive Smaller Purposeful Natural setting Words Hand coding/theming Computer assisted software programs

Apply theory Deductive Larger Random Laboratory setting Numbers Statistical tests Statistical software packages

Table I. Comparisons between qualitative and quantitative research

IJCHM 24,6

822

one end of the continuum and purely quantitative (i.e. controlled laboratory study) on the other end (Achterberg and Arendt, 2008). Figure 1 displays various qualitative approaches and data source examples along the more qualitative portion of this continuum. As can be seen, ethnography and observation are generally viewed as more qualitative methods while semi structured interviews and structured content analysis are viewed as less qualitative. Various data sources are utilized depending on which method is selected. For example in ethnographic work, researchers generally keep a diary, collect artifacts, and take photos or videos to use for analysis purposes. Qualitative research in hospitality Qualitative research has progressed in fields such as anthropology, education, and management (Marshall and Rossman, 2006). Medical and health services research have also been using qualitative methods and funding agencies have been supportive of such methods (Curry et al., 2009). The need for qualitative research and the appropriateness of utilizing specific qualitative methods for research in the hospitality field have been recognized (Kwortnik, 2003; Walsh, 2003). Hospitality and tourism researchers have utilized qualitative research as a single methodology (Dirks and Rice, 2004; Johnson et al., 2005; Paget et al., 2010; Papageorgiou, 2008) and as part of a mixed methods approach, utilizing both qualitative and quantitative research methods in the same study (Ineson et al., 2006; Parsa et al., 2005). Others have employed qualitative methods when more quantitative methodologies have failed to answer the research question (Mason et al., 2006). Riley and Love (2000) found that applied researchers used qualitative methods prior to employing quantitative methods whereas social science researchers generally used solely qualitative methods. The versatility of qualitative research is evident; qualitative research has been conducted in various hospitality and tourism contexts such as lodging, commercial and non-commercial foodservices, clubs, and tourism. Haiyan et al. (2010) used multiple qualitative methodologies to study human resources management aspects in China’s hotels whereas Lockyer and Roberts (2009) used only focus groups to study New Zealand hotel guests. Severt et al. (2008) employed a triangulation data collection process using structured visits, unstructured visits, and interviews to study organization philosophy in one hospital. Chapman et al. (2010) used video cameras to observe foodservice employees’ behaviors related to safe food handling techniques. Barrows (2000) studied the extent of training and challenges of training in private clubs; focus groups with club managers provided the data collection method. To study

Figure 1. Qualitative methodology continuum identify methods and data sources used

tourists’ perceptions, Kivela and Johns (2003) interviewed Hong Kong tourists. As is evident from this brief summation and others not elaborated on, qualitative research methodologies have canvassed hospitality and tourism research topics including strategic management (Pouder and Clark, 2009), training (Anderson et al., 2001; Barrows, 2000), and customer’s perceptions (Kivela and Johns, 2003; MacKay and McVetty, 2002). Achterberg and Arendt (2008) indicated that qualitative approaches are used when there is an inability to understand the problem using quantitative approaches and when it is important to understand the process variables (e.g. motivators to following safe food handling) not just the outcome variables (e.g. answering the question, did the employee follow safe food handling practices?). Although the data collection and analysis processes can be tedious, technological advances are being made to facilitate using qualitative methods in hospitality research (Pullman et al., 2005). Approach/methodology The researchers have conducted multiple qualitative studies in foodservice operations and with foodservice employees. These have included observations, individual open-ended interviews, and focus groups. A brief critique of each method and the experiences researchers have had with each method will be covered in this section of the paper. A summation of the eight research projects may be found in Table II. All research protocols involving human subjects were approved by the institutional review board (IRB) at each respective university. Observations Ethnography and participant observation are terms associated with qualitative observational studies. In ethnographic studies, the researcher integrates into the culture or community being studied. This generally is done over a considerable amount of time so the observer can assimilate into their new surroundings. The observed become immune to having the observer around after a period of time and therefore behave naturally. Other approaches involve participant observations, which may occur over a shorter period of time. The observer might also videotape the participant(s) being observed and keep records through field notes and study-specific observational forms. Observations allow for greater depth of understanding than focus groups or interviews because phenomena are studied in the naturalistic setting (Harris et al., 2009). In Project 1, food safety practice observations were conducted with foodservice production employees in commercial restaurants. The research protocol involved 20-minute observations and ten-minute rest periods over three hours during a breakfast, lunch, or dinner shift. This schedule allowed for six total observation periods. Depending on the production system of the operation, observers could observe as many as four employees in one observation period. However, in the majority of operations, only three employees were able to be observed at one time by each observer. A total of 242 employees, representing 33 different foodservice operations, participated in the study. In this study, managers were called to recruit potential participants (Roberts et al., 2008). Results of this study showed that training can improve behavior, but does not ensure behaviors will change. Conducting actual observations was a key element in this study, rather than relying on self-reported data as other studies have done.

Use of qualitative research 823

Table II. Summary of qualitative food safety studies conducted by authors Observations Hourly foodservice employees in commercial operations

Individual interviews Upper level managers, commercial and noncommercial foodservice operations Individual interviews Restaurant managers Focus groups Hourly restaurant employees Focus groups Hourly retail foodservice employees, based on age cohort Focus groups Retail foodservice supervisors, based on age cohort

Project 4 (Arendt et al., 2011a)

Project 5 (Roberts and Barrett, 2009)

Project 6 (Howells et al., 2008)

Project 7 (Arendt et al., 2011b)

Project 8 (Roberts et al., 2012)

Total number of participants

Observations and interviews Non-supervisory employees at commercial and non-commercial operations, based on age cohort

Project 3 (Arendt et al., 2011a)

Observations Project 2 (Paez et al., 2007; Strohbehn Food preparers at deli operations Hourly employees in commercial et al., 2008) and non-commercial retail foodservices

Project 1 (Roberts et al., 2008)

Methodology Selection criteria

36 613

32

159

20

4

25

80

15

242

Number of participants

Research observation, independent theme identification, and consensus

Research observation and consensus

Research observation and consensus Motivators to following safe food Research observation and handling practices consensus

Perceptions of foods safety

Barriers to following proper food Research observation and safety practices consensus

Barriers to offering food safety training to employees

Motivators to following safe food Research observation and handling practices consensus

Behavioral compliance with identified food safety standards

Frequencies; qualitative evaluation of methods

T-tests; simple linear regression

Behavioral compliance with identified food safety standards Hand washing behaviors

Analysis methods

Variable of interest

824

Project (Reference)

IJCHM 24,6

For Project 2, two subsequent observational studies focusing on handwashing frequencies and methods were conducted. In the first study, Paez et al. (2007) developed a data collection tool and protocols to track hand washing behaviors in deli-type restaurants. Observations were conducted during phases of production and service in a limited menu setting for a total of 30 hours with 15 participants. The reliable data collection tool and standardized protocols were used in a subsequent study to assess compliance with hand washing behaviors described in Food Code 2005 (Strohbehn et al., 2008). Sixteen venues representing four sectors of retail foodservice (assisted living for the elderly, child care, restaurants, and schools) and a range of scope of menu offerings, number served, service styles, amount of from scratch production, and employee tenure with the specific foodservice. Observations of 80 hourly employees were made at the beginning of the three-year project in the 16 operations with a data collection period of 15 hours in each facility during phases of production, service, and cleaning for a total of 240 hours. From these observations, compliance rates with Food Code recommendations of when and how hands were washed were collected for each sector of the industry, with training programs tailored to meet specific sector characteristics. Post-intervention observations were collected over a three-hour period in each of the 16 units (48 hours total) at the end of the study. Both studies within Project 2 showed that actual handwashing practices were not consistent with Food Code recommendations (Food and Drug Administration, 2005). In another study by Arendt and others (2011a), Project 3, foodservice employees of different age groups were observed. As compared to the Roberts et al. study (2008), each researcher observed one employee for a three-hour period on two different days for a total of six hours of observational time for each employee. A total of 25 employees were observed for the entire six hours resulting in 150 hours of data collection. Nine different foodservice operations participated in the study. Both commercial and noncommercial foodservice organizations were involved; hospital, university dining, school, long-term care facility, quick-serve restaurant, casual dining restaurant, catering, and deli retail. Observations indicated a gap between food safety knowledge and safe food behaviors. Individual interviews In contrast to focus groups, individual interviews are used when one-on-one questioning is desired. Sensitive topics or certain interviewee characteristics lend themselves to individual interviews rather than focus group interviews. For example, if studying foodservice employees’ specific food handling behaviors that may contaminate food, individual interviews may be more effective as the participants may not be willing to share their thoughts and behaviors freely when other foodservice employees are present. Kaplowitz (2000), when studying the socially sensitive topic of ecosystems with residents of Mexico, found participants were more likely to bring forth socially sensitive topics in an individual interview format as compared to focus group. The author noted complementary data from the two methods, not substitutive. In Project 4, researchers conducted individual interviews with four upper-level managers of commercial and noncommercial foodservice operations (Arendt et al., 2011a). The goal of the project was to further inform the data collected from employees and supervisors about food safety knowledge, behaviors, and training during focus groups. Data were analyzed by four researchers experienced in qualitative data

Use of qualitative research 825

IJCHM 24,6

826

analysis, coded and themed. Independently developed themes were then compared and discussed until consensus was reached. Common themes were then identified among the researchers. For Project 5, individual interviews were utilized in a study to explore barriers restaurant managers have to offering food safety training to their employees within the commercial foodservice environment (Roberts and Barrett, 2009). Twenty restaurant managers were interviewed and each interview lasted approximately 45 minutes to one hour. As one researcher focused on the discussion, a second researcher was present to record notes. Conversations were also audio-recorded and transcribed. Transcriptions were then compared to researchers’ notes to ensure accuracy. Data were coded by the researchers to develop themes for further analysis. Examples of themes identified in the analysis included lack of time, costs associated with training, and high employee turnover. Results were used to build a questionnaire that was utilized with a larger sample. Focus groups Focus groups, sometimes referred to as group interviews, allow the researcher to assemble a group of individuals with specific qualifications and/or characteristics. The focus group moderator poses open-ended questions to the members and allows the members to have a “coffee table” discussion about each question. One benefit of a focus group, as compared to individual interviews, is that members of the focus group can build on one another’s ideas. A drawback is that certain members might dominate the discussion or conversely, not participate in the discussion. An experienced moderator and assistant moderator are recommended to assure focus group success (Krueger and Casey, 2009). Like individual interviews, data analysis involves the process of coding and theming transcripts. Some researchers suggest that using computer-assisted qualitative software may enhance rigor of analysis (Kidd and Parshall, 2000). Focus groups were used to elicit barriers among hourly employees to following proper food safety practices within the commercial restaurant environment in Project 6 (Howells et al., 2008). The research protocol involved a focus group led by one of the authors; an experienced moderator, with the assistance of an experienced assistant moderator who observed the session and took field notes during the conversation. Thirty focus groups were completed with a total of 159 employees. Each focus group lasted approximately one hour and had an average of six employees per focus group. A multitude of barriers were identified by participants, such as time constraints, inadequate knowledge, not having reminders, and lack of resources. Results were used to develop and test various interventions within the foodservice industry. In a similar study (Project 7), focus groups were used to assess hourly employees’ perceptions and develop practice recommendations for supervisors in foodservice operations working with multigenerational employees (Arendt et al., 2011b). Again, an experienced moderator facilitated the focus groups and an experienced assistant moderator observed the sessions and recorded detailed notes. Four focus groups were conducted with different age cohorts of foodservice employees (18-25, 26-40, 41-60, and over 60 years of age). A total of 32 participants were included in this study. Project 8 focused on supervisors’ roles in motivating employees to follow proper food safety practices (Roberts et al., 2012). Participants were recruited from two

university towns in the Midwest. Future supervisors were recruited in hospitality management classes at the two universities and supervisors were recruited from foodservice operations in the two towns. As in the other focus group studies, an experienced moderator facilitated the focus groups and an experienced assistant moderator observed the sessions and recorded detailed notes. Three focus groups were conducted with an average of twelve participants per focus group. A total of 36 participants were included in this study. Challenges Challenges abound when doing any type of research, whether using a qualitative, quantitative, or mixed methods approach. Specific qualitative approach challenges related to the Institutional Review Board approval, recruitment, selection, retention, and data collection are addressed here. IRB approval Before recruitment and data collection can begin, most researchers must go through the IRB approval process at their respective universities. This requires scrupulous planning and preparation of multiple documents for the approval process. In one observational study, whereby researchers (Arendt et al., 2011a) were proposing to observe foodservice employees in their work setting and then interview them after completing observations, eight attachments were required for the IRB submission process. Examples of attachment documents for this study included a recruitment script, observations form, debriefing script, informed consent, and interview guide. Because many employees in the region where foodservice organizations were recruited spoke Spanish rather than English, some materials written in English were translated into Spanish after approval and then resubmitted. Although the time involved in documentation is one of the challenges, it is not unique to qualitative research as quantitative researchers working with human subjects must go through the same time-intensive process. The challenges lie in the advanced planning process and layout of specific observation and interview forms. Due to the emergent nature of qualitative research, it is not always feasible to anticipate what will happen in the field, in this case foodservice operations. Thus, the researcher is presented with a conundrum: it is unknown, which documents or procedures to specify until one is in the field; yet data collection in the field cannot begin until approval is received. The ability to adapt research protocol based on results and/or participant feedback is vital to the success of qualitative studies; this adaptability includes flexibility in the types of questions asked, observational time, methods of recruitment, and number of sites recruited. The lag time between submission and acceptance of IRB forms is a challenge that most researchers face. With universities experiencing limited resources and staff experiencing increased work expectations, the review period can often take six weeks to three months. IRB committees may schedule meetings on a weekly or monthly basis. Denzin and Lincoln (2005, p. 38) addressed the challenges with IRBs when presenting qualitative research for approval. They provided explanation to these challenges as: IRBs being understaffed; members who are uninformed about qualitative research; lack of proper procedures; and methods unavailable for expediting exempt forms of research. In one study (Arendt et al., 2011a), the principal investigator was asked to

Use of qualitative research 827

IJCHM 24,6

828

attend an IRB meeting to clarify and defend the submitted work. Two questions posed by the IRB committee at that meeting were: (1) “What would you do if you saw an employee intentionally put something in the food?” (2) “How will you assure the supervisor/manager will not find out how the employee performed during your observations?” Recruitment and selection The overall purpose in qualitative research is to search out the depth and breadth of an identified topic area. Qualitative research is used to develop theory, explore phenomena, and understand individuals in their natural setting (Achterberg and Arendt, 2008). Therefore, sample selection is generally done by recruiting participants who have unique knowledge about the topic being studied. Selection criteria are generally set prior to recruitment efforts. For the researcher to be successful in recruiting he/she must build rapport and trust with the potential participant; these relationships must be nurtured and developed over time. Once a participant agrees to join the study, it is important to continue relationship building. This means being responsive to questions in a timely manner and adhering to the parameters set in the recruitment phase. Without these inputs, the informant will not share openly and honestly, thereby impacting the overall trustworthiness of the study. Morgan (1998) noted that recruitment problems are the main reason focus groups go awry. Unlike quantitative research that may use a rented list of contacts for survey distribution, qualitative researchers can struggle to identify potential participants (employees, supervisors/managers). Getting access to names and contact information is often only the first step to getting consenting participants. In some cases, potential participants also must gain consent from a higher authority, such as a manager, district manager, and/or owner. Getting past the gatekeeper, that person in charge of granting access to an organization, is essential (Taylor and Bogdan, 1998). As Taylor and Bogdan wrote, “You want to convince the gatekeeper that you are a nonthreatening person who will not harm their organization in any way” (p. 29). Another concern expressed by participants is the paperwork or documentation of consent that is required for observations, focus groups, or interviews. The question of “how will this information be used” must be answered satisfactorily by the researcher. Show rates and retention Getting participants into the project is followed by the challenge of retaining participants throughout the study duration. Many studies involve repeated observations of the same participants for various reasons (e.g. changes in behaviors). Participant drop-out creates missing data and results in an unusable case. In one study (Arendt et al., 2011a), multiple observations of 28 employees at different operations had a loss of 11 percent; three employees left the study after the first observation due to termination of employment or concerns continuing the study. In a three-year behavioral-based study exploring food safety practices (York et al., 2009), which was an extension of Project 1 (Roberts et al., 2008), 86 percent of participants dropped out of the study, yielding 33 participants completing the entire project. Retention considerations also exist for other qualitative data collection sessions such as individual interviews and focus groups, even if multiple contacts with the same

individuals are not part of the protocol. If an individual agrees to an interview or focus group, it is unknown until the session whether he/she will actually be present. Show rates for interviews and focus group work described in Table II ranged from 13 percent to 140 percent for employees and 100 percent for supervisors, potential supervisors, and upper-level managers. Once a participant shows up for the interview or focus group, the researcher is then challenged with making certain the participant stays through the entire session and provides thoughtful and truthful information. For example, in Project 6, half of the participants in one of the focus groups had to leave early. One participant had to leave early to pick up her son from school and four other participants had ridden with her. In another focus group, where researchers traveled two hours to get to the focus group location, none of the employees showed up for the session. Upon calling the operation’s manager for an explanation, researchers were told employees left work early to go to the focus group together in one car, however, “they must have gotten distracted along the way”. An additional issue that has been identified is the challenge associated with gifts or compensation for participants. Due to the in-depth nature of qualitative research and the time it requires on the part of the participant, it is common to provide a “thank you gift” or some type of compensation for participants in a study. For most researchers, university protocol requires participants who receive compensation to sign a form acknowledging receipt of the money. Participants start to grow leery of all the forms and papers, particularly those who are not legal to work in the United States. In some organizations, employees were prohibited from accepting gifts because of organization policy and procedures. Data collection approaches During the data collection phase, challenges abound. The moderator of a focus group is particularly challenged with keeping all participants engaged and making certain no one dominates the discussion. Researchers are challenged with maintaining a neutral position when collecting data and at times, this may be difficult to do given the passion researchers have for their chosen field of work. The horns and halo effects may be challenges researchers need to overcome when performing qualitative data collection. The halo effect occurs when the researcher sees one thing that a participant does well or an aspect the researcher admires and this clouds the researcher’s ability to objectively collect data. In contrast, the horns effect occurs when the researcher sees something he/she dislikes about the participant and this overshadows the researcher’s abilities to objectively collect data. Researcher fatigue, when conducting focus groups, interviews, and observations becomes another hurdle to overcome. The hospitality industry is diverse and therefore researchers conducting research in the industry are challenged to collect data from this diverse group. Language is just one area where diversity is manifested. Data collection in the field is similar to the position of an “embedded reporter” in that researchers must be in the food production, service, or cleaning areas. It requires knowledge of operations to anticipate where best to stand, and where and when to move. Many foodservices have limited space in the back of the house, thus it is imperative the researcher be cognizant of potential risks from hot production equipment or employees’ traffic routes. Another challenge in data collection is establishing rapport with the foodservice staff under observation. It is

Use of qualitative research 829

IJCHM 24,6

imperative the researcher “blend in” with the operational activities. Initially, there will be some Hawthorne effect demonstrated, but as the employees become accustomed to the researcher’s presence, typical behaviors will resurface. In some cases, the manager or owner may not discuss the project with employees; resentment could be a factor.

830

Publishing qualitative research After completing qualitative research, dissemination through peer-reviewed journals is desired but sometimes challenging. Often times manuscripts may be rejected outright or reviewers may not understand qualitative research and reject the manuscript due to small sample size or nonrandom sampling techniques. Table III summarizes the challenges elaborated in this section and potential strategies. An in-depth discussion of strategies that can be used to overcome these challenges follows.

Approaches

Challenges

Strategies

Observations

Advanced planning for IRB approval

Educate and work with IRB Develop timeline so ample time for meticulous planning and material development. Add in as much flexibility as possible Set selection criteria, establish a trusting relationship, follow chain of command, and nurture relationship over time. Make contacts at non-peak business hours, the earlier in the day the better Over recruitment, build a comfortable trusting relationship with those observed, researchers experience in field, clothing similar to those being observed, understand climate of the organization Use native speaker or interpreter Speak truthfully yet vaguely about the project Let participants know forms will not be with the data collected Set the dollar amount at a level that compensates for time, but isn’t an incentive to participation

Recruitment and selection

Retention Thank you gifts and forms

Table III. Challenges and strategies when conducting observations, individual interviews, and focus groups

Individual interviews

Show rate/retention Diverse languages

Reminder calls, build relationship/common bond Use native speaker or interpreter

Focus groups

Recruitment – getting past the gate keeper Show rate/retention

Establish and nurture trusting relationships Follow chain of command Over recruitment, convenient locations, collect preferred contact information, reminder “calls”, develop personal connection Develop comfort and trust, small talk, common bond, light food and beverage, blend in clothing, set rules for the group Experienced moderator, relaxed and organized moderator, extensive preplanning, set protocols and checklists

Participants’ engagement and domination Moderator concerns Horn and halo effect Researcher influence

Strategies IRB approval Strategies to help cope with the many challenges presented before and during IRB approval are as follows: . educate and work with the IRB; and . develop adequate time into the research plan for meticulous planning and development of all IRB materials. Researchers should work with IRB members and provide them with an explanation of the hospitality industry. Experience has shown that committee members often have limited experiences with hospitality organizations, other than as customers, and therefore do not understand the hierarchical structure and workplace challenges such as turnover. Anticipating this limited understanding and providing background information about general procedures in the proposed data collection methods will help IRB members better understand the uncertainties when working with hospitality organizations and lead to project approval. Developing a meticulous, standard protocol is essential and will assist with both the IRB process and data collection phase of the research. For work involving inherent threat to safety, such as the food safety work described, actions needed by the researchers must be well-thought out and understood. For example, if observers see or think they see an inherent threat to the safety of the food, actions must be taken. Ethically, researchers cannot condone intentional sabotage of food. Recruitment and selection Qualitative research necessitates researchers to over-recruit to assure sufficient show rates. This is important for data collection methods that require participants to travel to a designated location, such as focus groups. However, it also is important for observation studies that involve multiple data collection periods. Missing data points for a participant yields them unusable in most study designs. Once selection criteria for the foodservice organizations or employees have been established, contacts should be identified. When selecting organizations to contact, it is best to build upon a trusting relationship, particularly when researching sensitive and potentially libelous topics, such as food safety. As with quantitative research, typically surveys, alignment with the university is usually viewed positively. In foodservice-based research, this was particularly evident, perhaps given the recognition and educational materials each of the researcher’s institutions has developed. Some of the research projects include Extension personnel who have a history of working with professional organizations to which some of the foodservice decision makers may belong, such as School Nutrition Association. The chain of command cannot be ignored when recruiting potential participants. When searching for non-supervisory employees, managers or those at higher levels in the organizations were contacted to get approval for posting recruitment flyers in the operation. Strategizing the time and mode of contact is essential. For recruitment of foodservice operations, contact via phone was most effective and calling during the morning worked best. Although obvious, contact during peak business hours (e.g. meal times) is not well received. For some types of operations, e-mail worked well and gave managers/owners more information that they could respond to at their convenience.

Use of qualitative research 831

IJCHM 24,6

832

The hospitality industry is still divided by those that embrace technology and those that do not see a need for it. Thus, multiple modes of communication will ensure messages are received. In addressing gatekeepers and informants, these researchers’ approaches are consistent with that of Taylor and Bogdan (1998, p. 33): truthful, but vague and imprecise so that when an informant asks about the work during the recruitment or data collection phase, specific explanation is avoided to mitigate effects of bias. For example, when employees in one observational study (Arendt et al., 2011a) asked what observers were doing, researchers responded with a truthful but vague response such as, “We want to better understand how best to communicate information about work tasks”. Okumus et al. (2007) in their work with international hotel groups noted the importance of negotiating with more than one member of the organization and building personal relationships. They also noted that it takes a substantial amount of time to get potential participants to commit, four months in their case. Compensation should be appropriate for offsetting the time to participate, but not at a level at which it will serve as an incentive. For example, in some research the compensation has ranged from $15USD for multiple observations of an employee and a 15-minute interview (Arendt et al., 2011a), recognizing the employee would already be doing their job, to $40USD for participation in a 90-120 minute focus group session outside of regular work hours that might have required travel (Arendt et al., 2011b). Show rates and retention Despite best efforts in recruitment, if employees do not show up for the observation, individual interview, or focus group, then all of the resources put forth have been lost without recognition of any benefit. Although there may not be much a researcher can do to make certain an employee shows up for their work shift or scheduled meeting, there are some specific strategies to ensure the best show rate possible. These strategies include careful selection of site for the focus group or observation, timely reminders about the upcoming data collection session, and developing a personal connection with the potential participant so they understand they will be missed if they do not show up when expected. When selecting a site for the data collection session, consideration should be given to the distance a potential participant must travel and comfort level the potential participant will have in coming to the location. Some specific strategies when selecting a location include: one with access to public transit, one where potential participants feel comfort, one with adequate parking, and one that is easy to find. The following example illustrates care in location selection. A focus group was being established for foodservice employees over the age of 60 years. A conference room in the local library, available by public transportation, was used for this focus group. This older group of employees was familiar with the library, the library was easily accessible to them (within a five minute car or bus ride), and had adequate, well lit parking. More than 100 percent of those recruited and confirmed were present; one individual showed up for the session but had not confirmed attendance at the focus group. Frequent reminders should be used up until and including the day of the data collection session, but not to the extent that they become obnoxious. The communication mechanism preferred by the potential participant should be considered. For example, phone call reminders were made to foodservice employees

over the age of 60 years, while e-mails were used when reminding employees in the youngest age category. As technology and social networking continues to evolve, consideration could be given to text message reminders and use of social networking methods, such as Twitter, to improve show rates. The keys to retention are participant comfort and trust in the researcher. Researchers can employ several strategies to achieve trust and comfort. Having experience in foodservice operations appears to help as the researcher can use terminology common to the participant, relate to the participant’s work experiences, and demonstrate understanding of the situation being described. All researchers involved in Projects 1-8 had foodservice work experiences at both hourly and managerial levels of employment, which allowed them to build rapport with participants. Researchers in observational studies must demonstrate emotional intelligence and become accepted within the setting quickly in order to minimize observer bias. The ability to chat or put employees at ease is critical. Likewise, appearance is important. A researcher does not want to stand out from the participants but be perceived as “one of them”. For this reason, great care should be taken when selecting attire for observations, interviews, and focus groups. Compliance with organizational and regulatory policies is important. For example, when observing in a foodservice production area, researchers should be mindful to wear clothing and follow practices consistent with the Food Code (e.g. close-toed shoes and hair restraints). For work with foodservice personnel, it is important not to appear in clothing similar to foodservice inspectors as this may conjure up feelings of fear and distrust. In earlier work, not described in this article, white lab coats were worn while conducting observations in food production areas. While this was acceptable for non-commercial types of operations, researchers have since concluded a more casual look is less intimidating. In subsequent on-site observations, researchers wore collared polo shirts (with university and project logos) and khaki slacks; similar attire to the uniforms worn in many foodservices. Data collection approaches Experience and extensive preplanning is necessary so that researchers can achieve objectives and participants can feel at ease with one another and/or with the researchers. One such preplanning activity is development of a checklist to make certain all necessary equipment is along when traveling to observation sites. Researcher’s anxiety, stress, and disorganization are evident to participants and therefore will affect willingness of participants to contribute fully and truthfully, thereby affecting overall data quality. Recording of observations can create some anxiety among participants. In more recent observations, researchers have used small (three inch, pocket size) spiral notebooks to record specific behaviors (such as hand washing or checking temperatures) rather than a standard size pad of paper and clipboard. This also allows the researcher to protect confidentiality of observations as the notebook can be slipped in and out of pocket when engaged with participants and supervisors. When conducting observations in operations, it is of utmost importance to understand the climate and culture of the organization. If researchers have work experience in like operations, this is helpful. There may be opportunities for participant observation, whereby the researcher takes part in the activities of the organization. For

Use of qualitative research 833

IJCHM 24,6

example, it may be important to taste the food if offered in order to fit in and not offend those observed. Should participants’ primary language be one other than that of the researchers’, it is essential that a trustworthy interpreter be hired. Ideally, the research team would include someone who spoke the primary language of the participants, even for observational studies.

834

Publishing qualitative research The best strategy for getting qualitative research published is to target journals accepting of qualitative work, explaining qualitative terminology that may not be familiar to reviewers/readers, and preparing a thorough methodology section. Certain journals, such as Journal of the American Dietetic Association, have developed specific authorship guidelines for publishing qualitative research (Van Horn, 2009). Likewise, Gibbert et al. (2008) have offered suggestions on how to write rigorous case studies and where to publish these. Conclusion This paper has addressed the challenges researchers face when conducting qualitative research with hospitality employees and managers using observations, individual interviews, and focus groups. Specifically, the challenges in obtaining institutional approval of qualitative research work, recruitment and show rates of potential participants, and retention of participants were explained. Strategies were identified to overcome or minimize these challenges. Working with IRBs, building trust with gatekeepers and participants, over-recruiting and most importantly, having experienced qualitative researchers are some of the strategies suggested. The use of qualitative research methods such as observations, focus groups, and interviews was effective when conducting food safety research within the hospitality industry. The use of qualitative research methods has allowed the researchers to identify many complex variables that survey research alone would not have permitted. These findings include compliance of behavior with identified food safety standards, hand washing behaviors, motivators and barriers to follow food safety practices, and manager and employee perceptions of food safety. There is a place for qualitative methods in hospitality research. Selection of this type of methodology should be in alignment with the research purpose and questions. Qualitative research should be used if depth and breadth of understanding is sought. However, as with any research method used, whether it is qualitative, quantitative, or a combination, rigorous standards must be maintained during the entire research process. By anticipating potential challenges in advance, researchers can take a proactive approach to implementing strategies and assure that rigorous standards are upheld throughout the qualitative research process. References Achterberg, C. and Arendt, S. (2008), “The philosophy, role, and methods of qualitative inquiry in research”, in Monsen, E.R. (Ed.), Research: Successful Approaches, 3rd ed., American Dietetic Association, Chicago, IL, pp. 65-74. Anderson, N., Groves, D.L., Lengfelder, J. and Timothy, D. (2001), “A research approach to training: a case study of mystery guest methodology”, International Journal of Contemporary Hospitality Management, Vol. 13 No. 2, pp. 93-102.

Arendt, S., Roberts, K.R., Strohbehn, C., Ellis, J. and Paez, P. (2011a), “Manager’s impact on motivating foodservice employees to follow safe food handling practices: perspectives of multigenerational employees”, working paper, Iowa State University, Ames, IA, 15 April. Arendt, S., Strohbehn, C., Paez, P., Correia, A., Shelley, M. and Meyer, J. (2011b), “Barriers and motivators for following safe food practices: age cohort qualitative analysis”, working paper, Iowa State University, Ames, IA, 1 August. Barrows, C.W. (2000), “An exploratory study of food and beverage training in private clubs”, International Journal of Contemporary Hospitality Management, Vol. 12 No. 3, pp. 190-7. Chapman, B., Eversley, T., Fillion, K., MacLaurin, T. and Powell, D. (2010), “Assessment of food safety practices of food service food handlers (risk assessment data): testing a communication intervention (evaluation of tools)”, Journal of Food Protection, Vol. 73 No. 6, pp. 1101-7. Curry, L., Nembhard, I. and Bradley, E. (2009), “Qualitative and mixed methods provide unique contributions to outcomes research”, Circulation, Vol. 119 No. 10, pp. 1442-52. Denzin, N. and Lincoln, Y. (2005), The Sage Handbook of Qualitative Research, Sage Publications, Thousand Oaks, CA, p. 38. Dirks, D. and Rice, S.K. (2004), “Dining while black: tipping as social artifact”, Cornell Hotel and Restaurant Administration Quarterly, Vol. 45 No. 1, pp. 30-47. Food and Drug Administration (2005), Food Code (Report No. PB 2005-102200), US Department of Commerce Technology Administration, National Technical Information Service, Virginia. Gelo, O., Braakmann, D. and Benetka, G. (2008), “Quantitative and qualitative research: beyond the debate”, Integrative Psychological and Behavioral Science, Vol. 42 No. 3, pp. 266-90. Gibbert, M., Ruigrok, W. and Wicki, B. (2008), “What passes as a rigorous case study?”, Strategic Management Journal, Vol. 29 No. 13, pp. 1465-74. Haiyan, K., Cheung, C. and Hanqin, Z. (2010), “Career management systems: what are China’s state-owned hotels practicing?”, International Journal of Contemporary Hospitality Management, Vol. 22 No. 4, pp. 467-82. Harris, J., Gleason, P., Sheean, P., Boushey, C., Beto, J. and Bruemmer, B. (2009), “An introduction to qualitative research for food and nutrition professionals”, Journal of the American Dietetic Association, Vol. 109 No. 1, pp. 80-90. Howells, A.D., Roberts, K.R., Shanklin, C.W., Pilling, V.K., Brannon, L.A. and Barrett, B. (2008), “Restaurant employees’ perceptions of barriers to three food safety practices”, Journal of the American Dietetic Association, Vol. 108 No. 8, pp. 1345-9. Ineson, E.M., Lyons, A. and Branston, C. (2006), “Cross cultural change, adjustment and culture shock: UK to USA”, Tourism, Vol. 51 No. 1, pp. 3-20. Johnson, C., Surlemont, B., Nicold, P. and Revaz, F. (2005), “Behind the stars: a concise typology of Michelin restaurants in Europe”, Cornell Hotel and Restaurant Administration Quarterly, Vol. 46 No. 2, pp. 170-87. Kaplowitz, M.D. (2000), “Statistical analysis of sensitive topics in group and individual interviews”, Quality & Quantity: International Journal of Methodology, Vol. 34 No. 4, pp. 419-31. Kidd, P.S. and Parshall, M.B. (2000), “Getting the focus and the group: enhancing analytical rigor in focus group research”, Qualitative Health Research, Vol. 10 No. 3, pp. 293-308. Kivela, J. and Johns, N. (2003), “Restaurants, gastronomy and tourists: a novel method for investigating tourists’ dining out experiences”, Tourism, Vol. 51 No. 1, pp. 3-20.

Use of qualitative research 835

IJCHM 24,6

836

Krueger, R.A. and Casey, M.A. (2009), Focus Groups: A Practical Guide for Applied Research, 4th ed., Sage Publications, Thousand Oaks, CA. Kwortnik, R.J. (2003), “Clarifying ‘fuzzy’ hospitality-management problems with depth interview and qualitative analysis”, Cornell Hotel and Restaurant Administration Quarterly, Vol. 44 No. 2, pp. 117-29. Lockyer, T. and Roberts, L. (2009), “Motel accommodation: trigger points to guest accommodation selection”, International Journal of Contemporary Hospitality Management, Vol. 21 No. 1, pp. 24-37. MacKay, K.J. and McVetty, D. (2002), “Images of first-time visitors to Queen Charlotte Islands and Gwaii Haanas National Park Reserve”, Journal of Park & Recreation Administration, Vol. 20 No. 2, pp. 11-30. Marshall, C. and Rossman, G.B. (2006), Designing Qualitative Research, Sage Publications, Thousand Oaks, CA. Mason, D., Tideswell, C. and Roberts, E. (2006), “Guest perceptions of hotel loyalty”, Journal of Hospitality & Tourism Research, Vol. 30 No. 2, pp. 191-206. Mays, N. and Pope, C. (1995), “Education and debate – qualitative research: rigour and qualitative research”, British Medical Journal, Vol. 311 No. 6997, pp. 109-12. Morgan, D. (1998), Planning Focus Groups, Sage Publications, Thousand Oaks, CA. Okumus, F., Altinay, L. and Roper, A. (2007), “Gaining access for research – reflections from experience”, Annals of Tourism Research, Vol. 34 No. 1, pp. 7-26. Paez, P., Strohbehn, C. and Sneed, J. (2007), “Handwashing frequencies and methods used in deli-types of foodservice operations”, Food Protection Trends, Vol. 27 No. 11, pp. 903-8. Paget, E., Dimanche, F. and Mounet, J. (2010), “A tourism innovation case: an actor-network approach”, Annals of Tourism Research, Vol. 37 No. 3, pp. 828-47. Papageorgiou, G.C. (2008), “The human dimension of tourism – supply-side perspectives”, Annals of Tourism Research, Vol. 35 No. 1, pp. 211-32. Parsa, H.G., Self, J.T., Nijite, D. and King, T. (2005), “Why restaurants fail”, Cornell Hotel and Restaurant Administration Quarterly, Vol. 46 No. 3, pp. 304-22. Pouder, R.W. and Clark, J.D. (2009), “Formulating strategic direction for a gated residential community”, Property Management, Vol. 27 No. 4, pp. 216-27. Pullman, M., McGuire, K. and Cleveland, C. (2005), “Let me count the words: quantifying open-ended interactions with guests”, Cornell Hotel and Restaurant Administration Quarterly, Vol. 46 No. 3, pp. 323-43. Riley, R.W. and Love, L.L. (2000), “The state of qualitative tourism research”, Annals of Tourism Research, Vol. 27 No. 1, pp. 164-87. Roberts, K.R., Arendt, S., Strohbehn, C., Ellis, J. and Paez, P. (2012), “Educating future managers to motivate employees to follow food safety practices”, Journal of Foodservice Management Education (forthcoming). Roberts, K.R. and Barrett, B. (2009), “Behavioral, normative, and control beliefs impact on the intention to offer food safety training to employees”, Food Protection Trends, Vol. 29 No. 1, pp. 21-30. Roberts, K.R., Barrett, B.B., Howells, A.D., Shanklin, C.W., Pilling, V.K. and Brannon, L.A. (2008), “Food safety training and foodservice employees’ knowledge and behavior”, Food Protection Trends, Vol. 28 No. 4, pp. 252-60. Severt, D., Aiello, T., Elswick, S. and Cyr, C. (2008), “Hospitality in hospitals?”, International Journal of Contemporary Hospitality Management, Vol. 20 No. 6, pp. 664-78.

Smith, J.K. and Heshusius, L. (1986), “Closing down the conversation: the end of the quantitative-qualitative debate among educational inquirers”, Educational Research, Vol. 15 No. 1, pp. 4-12. Strohbehn, C.H., Sneed, J., Paez, P. and Meyer, J. (2008), “Handwashing frequencies and procedures used in retail food services”, Journal of Food Protection, Vol. 71 No. 8, pp. 1641-50. Taylor, S.J. and Bogdan, R. (1998), Introduction to Qualitative Research Methods, 3rd ed., John Wiley & Sons, New York, NY. Van Horn, L. (2009), “New author guidelines include qualitative research section”, Journal of the American Dietetic Association, Vol. 109 No. 1, p. 17. Walsh, K. (2003), “Qualitative research: advancing the science and practice of hospitality”, Cornell Hotel and Restaurant Administration Quarterly, Vol. 44 No. 2, pp. 66-74. York, V.K., Brannon, L.A., Shanklin, C.W., Roberts, K.R., Howells, A.D. and Barrett, E.B. (2009), “Foodservice employees benefit from interventions targeting barriers to food safety”, Journal of the American Dietetic Association, Vol. 109 No. 9, pp. 1576-81. Corresponding author Susan W. Arendt can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Use of qualitative research 837

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0959-6119.htm

IJCHM 24,6

Investigative management and consumer research on the internet

838 Received 14 August 2010 Revised 21 December 2010 8 April 2011 18 October 2011 1 November 2011 Accepted 1 November 2011

Peter Lugosi Oxford School of Hospitality Management, Oxford Brookes University, Oxford, UK, and

Hania Janta and Pamela Watson School of Tourism, Bournemouth University, Bournemouth, UK Abstract Purpose – This paper aims to introduce the notion of investigative research on the internet (IRI) and conceptualise its processes through the principle of streaming. It seeks to discuss the similarities and differences between IRI and netnography and considers various aspects of the IRI process, including site selection, sampling, data collection and analysis. Design/methodology/approach – Investigative internet-based research uses the techniques of ethnography and netnography, including variations of participant observation and analysis of visual and textual material. Three international empirical cases are used to illustrate the application of IRI and streaming in research on international workers, consumer cultures and on emerging business phenomena. Findings – IRI has a number of potential applications for hospitality management academics and practitioners. Streaming can help to understand the processes involved in conducting netnographic research, and streaming is a more appropriate way to conceptualise some internet-based studies that do not conform to netnographic or ethnographic ideals. Research limitations/implications – The three empirical cases highlight the processes of streaming in practice, which can be applied elsewhere. Principal limitations are the ethical dimensions of conducting undisclosed research and the sampling bias resulting from adopting an unobtrusive role and focusing on active internet users. Practical implications – The paper highlights several issues, identified through streaming, that can be used to design human resource, marketing and operational strategies. Originality/value – The paper demonstrates the application of streaming. Streaming can help researchers conduct netnographic studies; it is also a more appropriate way to describe broader types of investigative internet research. Moreover, it demonstrates the applicability of streaming in research on hospitality management and public policy issues. Keywords Internet, Online, Virtual, Ethnography, Netnography, Qualitative research, Streaming analysis, Hospitality management Paper type Research paper

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 pp. 838-854 q Emerald Group Publishing Limited 0959-6119 DOI 10.1108/09596111211247191

1. Introduction The development of web 2.0 and interactive technologies offer numerous opportunities for qualitative management and consumer research. Researchers may use instant messaging services and chat rooms (virtual spaces facilitating synchronous communication) to conduct interviews and focus groups (Fielding et al., 2008; Grossnickle and Raskin, 2001). These enable fieldworkers to collect data through

real-time interaction, although there are various other sources that facilitate asynchronous data collection, focusing on archived material (Hewson et al., 2003; Dholakia and Zhang, 2004; Poria and Oppewal, 2003). This paper advances existing knowledge on online methods by introducing the notion of investigative internet research, or Investigative Research on the Internet (IRI), and distinguishing such qualitative, exploratory studies from other forms of online research. It is argued that IRI shares many of the characteristics of netnography (Kozinets, 2002) and other forms of online ethnography (Hine, 2000, 2005), but that not all IRI can or should be called netnography. We conceptualise the processes involved in such internet-based research using the notion of streaming, which can inform future netnographic studies as well as other IRI that does not conform to ethnographic ideals. Finally, the paper demonstrates, through three international empirical cases, how IRI and streaming can be applied in research on international hospitality workers, consumer cultures and networks, and more broadly on emerging hospitality business phenomena. We concentrate on three sources of data: (1) message boards or forums, which enable users to engage in asynchronous exchanges or postings on different topic threads; (2) blogs, narratives and opinion pieces written by individuals, to which others may respond and which may also provoke ongoing asynchronous exchanges; and (3) websites provided by commercial and non-commercial organisations or individuals, offering news, reviews and either lay or professional commentary. We discuss the investigative processes, including navigation in and across multiple “field sites”, sampling, analysis and interpretation, through which researchers can exploit these three sources of data in their studies. The paper contributes to knowledge in several ways. First, we illustrate how internet research and streaming can be applied to address issues relevant to hospitality managers, academics working in broader disciplines, alongside policy makers. Second, and more broadly, it uses research on hospitality to conceptualise qualitative, investigative, internet-focused research that is applicable to other intellectual contexts. Consequently, rather than simply importing concepts from other disciplines, hospitality research creates opportunities to contribute to wider fields of enquiry. 2. Conceptualising investigative research on the internet It is important to begin by distinguishing IRI from other types of internet-based methods of data collection, for example, text mining. These approaches use software that searches the internet and attempts to identify patterns and relationships between textual information (Miller, 2005). IRI may use specialist software, and it will inevitably employ search engines; however, it does not rely on textual material alone, and the researcher has a much greater role in sourcing, analysing and making connections between sources of information. IRI has a number of features. First, as the term investigative implies, it is inherently analytical, exploratory and seeks to simultaneously discover and construct. This implies that the research process attempts to unearth existing pieces of information, which emerge in a variety of ways, for example through objects, actions, images, texts, relationships, interactions, spaces and institutions. Moreover, the researcher or

Investigative research on the internet 839

IJCHM 24,6

840

investigator connects these to other pieces of data that may not seem immediately related. Second, the research act is a dynamic process, responsive to changes in the fieldwork “site”, but also requiring sensitivity towards the individuals who provide information, towards relationships and interactions during the fieldwork, and towards the various sources of data, which may not be predetermined at the study’s outset. Third, because the research is dynamic, it also has to be context sensitive, taking account of the subtle idiosyncrasies of the virtual spaces through which researchers navigate. Researchers also have to be mindful of the social, cultural and sub-cultural factors entangled in the offline worlds in which data emerges. Finally, as noted above, in IRI the researcher is not a neutral entity, but is central to the research process, making decisions at every moment about what is relevant, what is excluded and which lines of enquiry are pursued. IRI thus shares many of the features of online ethnography (Hine, 2000; Kozinets, 2002; Wittel, 2000); consequently, it is useful to discuss these research approaches in further detail, before introducing the notion of streaming to conceptualise the processes of conducting IRI. 3. Ethnography and netnography Our discussion primarily draws on and engages with netnography, rather than other conceptualisations of internet-based ethnography (e.g. Hine, 2000; 2005), partly because Kozinets (2002, 2010) has attempted to define its processes in detail, but also because the concept clearly distinguishes it from traditional ethnography. Moreover, netnography, which was originally developed for marketing and consumer research, is rapidly becoming an accepted concept in business research, increasingly being applied to the study of hospitality (e.g. Watson et al., 2008). Netnography, like its antecedent, ethnography, is not a single method or approach; it involves a range of investigative strategies. The lack of space prevents a detailed discussion of ethnography, but it is useful to briefly consider some of its key features, which helps to understand netnography’s underpinning principles. Lugosi (2009), synthesising earlier discussions, suggests ethnography seeks to gain an emic understanding of groups and individuals through immersion in their activities over extended periods of time. The emerging insights are context sensitive and embrace the complexity of human phenomena. The researcher, rather than being a detached neutral entity in the research process, has a central role in knowledge generation. This means fieldworkers have to be critically sensitive about their relationships with informants and the data they develop. Ethnographers often use some variation of participant observation, although this is often complemented by interviews, the analysis of written material, and, increasingly, by visual methods (see, e.g. Pink, 2007). Historically, the insights gained through ethnography have been communicated through rich textual accounts, although these, too, have been enhanced by the use of technology and visual methods. Netnography also attempts to understand, rather than measure, but Kozinets (2010, p. 5) highlights several features that distinguish it from ethnography. First, interaction is different than in “traditional”, face-to-face ethnography. Engagement can be with a much wider range of geographically dispersed people, and participation can take a number of different forms in online environments. Data can be gathered without any interactions but engagement may also range from brief exchanges to longer and deeper interactions, using a variety of communications, e.g. text, sounds, images and other “virtual” expressions such as emoticons. Second, rather than relying on field notes and

creating textual accounts, netnography gathers and uses a wider range of data, which varies in terms of focus and quality. This means the researcher has to broaden her or his repertoire in gathering, interpreting and integrating vastly different data. These are themes echoed by others distinguishing between traditional and virtual ethnographies (e.g. Wittel, 2000). Kozinets (2010, p. 89) outlines six criteria for selecting sites for study, arguing that they should be: (1) relevant to the research focus and questions; (2) active, with recent and regular communications; (3) interactive, having flows of communication between participants; (4) substantial, in terms of numbers of users; (5) heterogeneous, involving different participants; and (6) data-rich. Kozinets (2002) also proposes a number of procedures to facilitate the process of conducting online research and he identifies several stages: entry into the “community”, data collection utilising a variety of textual and visual material, analysis and interpretation, as well as member feedback, which requires an overt approach. However, this was challenged by Langer and Beckman (2005) who suggest adopting a covert approach and Beaven and Laws (2007) who propose conducting observational or passive netnography. Informed consent is seen to be unnecessary when there is an openly available archive of online material or when forums are publicly accessible (Beaven and Laws, 2007). Kozinets’ clear guidance helps to shape the direction and processes of online research, but there remain several areas of debate for fieldworkers, which make it difficult to think of all IRI as netnography. The first concerns the centrality of community in netnographic research. The notion of community has continued to evoke critical debate, and there have been extensive discussions about its appropriateness in understanding the multiple relationships that exist in cyberspace (see Jones, 1998). Contemporary debates have shifted away from traditional conceptions of community, based, for example, on structures and stable roles, to more fluid ones, suggesting that people continually (re)construct and perform notions of shared identity and belonging (Guimara˜es, 2005). Kozinets (2010) acknowledges the problematic nature of “community”, but maintains that the focus of netnography is on “online communities”, i.e. groups constructed and maintained in virtual worlds. He also advocates the increased use of netnography, in conjunction with traditional forms of ethnography, to study “communities online”, i.e. groups that exist offline, but also interact through the web. According to Kozinets (2010), communities involve a minimum number of people, who identify with others in the community and engage in ongoing, meaningful interactions with other members. This can be used to describe some of the social configurations and interactions relevant to hospitality management research, but it is problematic to designate individuals with shared interests or involvement in particular work and leisure activities as being part of a community. Contemporary ethnographers, engaged in “multi-sited” fieldwork, have maintained that ethnography continues to offer insights into the complex sets of actions through which individuals construct and articulate relationships and notions of self/other (see,

Investigative research on the internet 841

IJCHM 24,6

842

e.g. Falzon, 2009; Guimara˜es, 2005). Extended engagement with multiple “informants” provides opportunities to understand behaviours, values and norms that enables researchers to define and thus construct individuals as belonging to a community. Netnography, and ethnography in and of virtual environments, therefore becomes a productive as well as an analytical act. Consequently, if we acknowledge the central role of the researcher in constructing a notion of community, then we have to recognise that the fieldworker’s version of community may not be shared by the people being represented. A final challenge, which relates to the previous point, concerns the implicit assumptions about the nature of knowledge and understanding gained through netnographic research. The central role of the researcher, and her or his extended engagement with individuals offering insights into the nuances of interactions, norms, values and identities, is one of the defining characteristics of both netnography and ethnography (Hine, 2000, 2005; Leander and McKim, 2003). Netnography, like ethnography, thus stresses a particular epistemological ambition in assuming that immersion into a group and its culture can help fieldworkers develop emic perspectives. However, if notions of community are fluid and highly contestable, and engagement with informants may be fragmented and asymmetric, not all the interactions and relationships can provide detailed ethnographic data, a point acknowledged by its practitioners (Hine, 2000, 2005; Markham and Baym, 2009). Moreover, not all IRI may set out to or claim to have gained such ethnographic insights, and describing such research as netnography is therefore potentially problematic. Calling all IRI “netnography” is an example of concept creep – a simultaneous extension and reinterpretation of netnography. Kozinets (2002), Hine (2000) and others recognise that netnography, and other forms of online ethnography, are flexible approaches, involving a range of methods and techniques, and are thus open to critical reconsideration. This paper engages in the reinterpretation of netnography by proposing the notion of streaming to help understand the processes involved in such internet research. However, IRI and streaming offer a broader conceptualisation of qualitative, exploratory internet-based studies. This is necessary and useful for two reasons: firstly, rather than simply, and uncritically, inflating the netnography concept, this paper defines a key aspect of such online research, thus helping researchers conduct netnographic studies. Secondly, distinguishing the processes involved in IRI from the concept of “netnography” helps researchers understand and conduct research via the internet, which may not conform to ethnographic ideals. 4. Investigative research on the internet as streaming We use “streaming” as a metaphorical device to conceptualise the data gathering and analysis processes in which researchers trace relationships between actors and agencies through the internet. The notion of streaming may initially seem to be limited, as is any other metaphor: it may create an image of a defined path to be followed and imply a singular direction, something which does not reflect well the multi-spatial and multi-directional nature of the internet and the movement of its users. Streaming echoes Hine (2000) and Leander and McKim’s (2003) focus on mobility and stresses the fluidity of relationships and interactions. Streaming should be thought of as a dynamic, creative process, when the researcher opens new directions for data gathering and analysis as he or she moves through and across different sites. As ethnographers

working in multi-sited environments have argued (Hine, 2005; Leander and McKim, 2003; Falzon, 2009; Wittel, 2000), fieldworkers connect, perform and effectively construct the field site. Even if they establish one set of connections, thus creating a “stream” through which they move, they can retrace their movement, move backwards, identify new divergences and create alternative connections – streams. At the core of streaming is the notion of connectedness and the importance of mobility. Urry (2007) is one of many commentators to stress that contemporary society is characterised by increased movement both as people move around physically but also as information circulates rapidly, facilitated by the growing prevalence of technology (see also Hannam et al., 2006). Consequently, individuals are less likely to be members of stable social groups, based on strong social ties; instead, people are connected, often through loose ties, to a wide range of other people through networks that stretch across geographical boundaries (Urry, 2007). Streaming can thus be thought of as a research aim to follow the networks of relationships through which information about people, places and experiences flows. Just as Hine (2000, p. 35) noted, rather than remaining on a single site or focusing on one context, internet research should use movement between sites as an analytical device. The key issue is that researchers attempt to understand how relationships and flows of information operate through virtual media. Movement across different virtual spaces in data gathering and analysis also implies that streaming is more likely to examine interactions and representations during extended periods and across several time periods, rather than only at a single point in time. A further feature of streaming is its inductive nature, involving simultaneous data collection and analysis (Lofland and Lofland, 1995). Concurrent data gathering and open, inductive analysis are necessary as representations and interactions occur across multiple discussion threads or even websites. In many cases, even when interactions take place on a single forum, they include links to other sites. Examining relationships and interactions through virtual networks inevitably leads to choices that researchers make about which sites, texts, discussions or representations of people, places, events or individuals should be traced further. These have the potential to become what we refer to as key streams, which lead to new lines of enquiry. Within traditional ethnography, “key informants” have often played a crucial role in gaining access, mediating on fieldworkers’ behalf and providing information. The notion of key streams avoids distinguishing between people, sites and texts, as they all have the potential to be crucial reference points researchers may draw upon once, but may return to repeatedly, to gain insights. There may also emerge more minor streams, located elsewhere, that are also useful but less so than key ones. Streams have an elaborative function, providing illuminating illustrations of behaviours, events, beliefs, values etc., thus helping to understand the dimensions and dynamics of particular phenomena. They also have a connective function, linking to further sites, documents, representations etc. where issues can be explored further. It is also useful to distinguish between key streams, which are the primary sources from which streaming begins and sub-streams that may exist within, or extend from, these key streams. These may be the discussion threads on a site, or representations by individuals, e.g. in the form of personal narratives, informal commentaries, or lengthier, formal discourses, identified by the researcher as relevant based on the examination of another site or text.

Investigative research on the internet 843

IJCHM 24,6

844

Finally, the process of following posts or threads and moving between different websites also helps to identify a final key feature of streaming, which is that it involves snowball sampling that may be complemented by other forms of criterion sampling, within a broader purposive sampling approach (see Patton, 2002). Internet interactions are driven by user generated material, and users can create new linkages between other users, sites and threads of discussion; therefore, it is important, in principle, to let the data and the emerging themes shape the sample size, rather than it being defined too narrowly at the outset of the research. Ideally, data collection will stop when the researcher arrives at a point of theoretical saturation, when no new themes emerge from the data (Strauss and Corbin, 1998). However, the specificity of the study’s aim and objectives and resource constraints, for example in terms of funding, time, project life etc. will often impose other limits on the study. Consequently, the research may come to focus principally on a specific forum, a limited number of threads – based on some research question driven criteria, a fixed time period, a limited number or network of users or even an individual, which is used as a particular case. 5. Streaming in practice 5.1 Streaming in research on employee experiences Background and purpose. Migrant workers make up a large part of the hospitality workforce and their experiences are frequently shaped by their immigrant status, language skills and their ability to mobilise social capital: they are frequently confined to low paid, low status jobs and hospitality offers particular career pathways, while denying others (Janta, 2011; Janta et al., 2012). The first study used to illustrate streaming is part of a wider project examining Polish migrant employees’ experiences in the hospitality and tourism sector (Janta et al., 2011). The study applied a mixed-methods approach to allow an understanding of respondents’ lives at work and beyond it, including their interactions, aspirations, work conditions, job scopes and demographic profiles. However, we concentrate here on the qualitative internet research dimension of the study. A key objective of the research was to examine migrant networks and how networking, in virtual environments in particular, mediated their employment experiences. More specifically, it sought to understand how migrants supported each other and represented their experiences of mobility and hospitality employment to other current and future migrants. Data collection and streaming. There is a high usage of social networking sites and discussion forums among Poles in the UK (Janta et al., 2012), and the public nature of these forums prompted the researcher to engage in observational or passive netnography (Beaven and Laws, 2007). Streaming utilised purposive sampling and began by identifying the key streams, using the six criteria, outlined earlier, to select appropriate sites. Initially, the site’s traffic and activity were prominent criteria. Three important sites were identified: (1) mojawyspa.pl (my island) with 23,000 threads; (2) ang.pl, a site dedicated to those learning English, with more than 72,000 threads; and (3) gazeta.pl – the biggest forum with more than 7,400 different forums on it and more than 102 million written posts.

These became the key streams, and the starting point for streaming was a specific virtual space – i.e. the forum, from which networks of relationship and interactions were traced. Within these key streams it was then necessary to identify sub-streams, which were the relevant message boards and threads, found by analysing headings and, where possible, by using the sites’ internal search engines. At the level of sub-streams, essential sampling criteria were increasingly the relevance of posts to the research objectives rather than just the traffic. Examples of topics included “job seeking” and “applying for positions”, “hotel work” and “other jobs in hospitality”. The relevance of some streams was obvious, for example: “Working in a hotel? Discuss it!” or “Why are educated Poles washing pots in the UK??” on gazeta.pl or “Manchester-waiter-a few questions:)” on mojawyspa.pl. However, other titles were less obvious, for example “Are you planning to return to Poland one day?” on gazeta.pl and others discussing migrants’ future plans, but these made repeated references to hospitality work. Approximately 100 existing threads, launched between 2004 and 2008, were downloaded, printed and analysed in this research. Some continued for only a few days while others until 2010. Key emerging questions that drove the processes of streaming, as noted in the previous paragraph, concerned the thread topics and subsequent interactions. Examining the threads led to other relevant issues and helped answer such questions as who uses the space and for what purpose. For example, a thread launched on gazeta.pl, entitled “How is it getting a job after returning to Poland?”, had an intensive beginning, with 95 posts within three days (11-14 April 2006). Tracing this stream helped to understand the development of migrant careers over extended periods, including the experience of return and the possibility to utilise the human capital developed during work in the UK. Other streams showed how migrants interacted, building their social networks. For example, a thread “Anyone going to Bournemouth?” on ang.pl was launched in 2005 and the last post was written in January 2010. Indicative emerging themes identified through streaming. Streams of enquiry helped to understand specific aspects of instrumental and emotional support directly related to hospitality employment. For example, one reply to a posting entitled “Interview – waitress – on Friday help me” (ang.pl, 27 May 2008) provided twenty examples of possible interview questions such as “What do you like about waitressing?” and “Are you good at multi-tasking?” Other aspects of support revealed through tracing key sub-streams incorporated reflections on hospitality work. Examples included streams entitled “Why are qualified Poles washing pots in the UK?” (2 August 2006), which discussed the experience of having to take jobs for which they were overqualified. Others were less philosophical and more prescriptive, such as the stream entitled “Don’t come! No work here!”, where posts revealed the bleak situation of Polish migrants: “Only those who have worked in the UK before and those who speak English well will find a job – even such as ‘on the sink’ [i.e. low status kitchen porter positions] – this is what it looks like now!’ (gazeta.pl, 30 January 2007). Not all the postings were so negative; some sub-streams also revealed information about migrants’ positive work-related learning experiences. For example, a thread entitled: “A waiter in England” (gazeta.pl, 25 November 2004), started by a user

Investigative research on the internet 845

IJCHM 24,6

846

seeking advice on learning English while working in foodservice, included a number of constructive postings by other users: It is hard work, especially when it gets busy and you need to deal with a number of things at the same time. But it gets better, for example when they ask you where you come from and they don’t want to believe that you come from a non-English speaking country and they appreciate your English:)

Another reply, posted on the same day, also revealed themes of emotional support from fellow migrants and positive work-based encounters. It also helped identify further streams of enquiry concerning the role of formal education in migrants’ experiences: Don’t give up! I don’t know why [the other site users] are scaring you! I work behind a bar, maybe it’s not the same as being a waiter but it is not simple either! (sometimes) I have been learning English for 4 years and I don’t have a FCE [First Certificate in English] (yet) but I don’t have problems in communicating. So don’t put him off! People are really understanding!

Significance, applications and contributions to knowledge. Streaming, using particular online forums as initial points of focus, was important in meeting the study’s objectives because these were spaces of interaction dedicated to a particular social group. However, these spaces were used by individuals from a wide range of geographical areas and it would have been very difficult to reach such a disparate network of individuals through conventional methods. In some cases, sites, and threads within them, concentrated on specific geographical areas or activities, e.g. hospitality work or language learning, which helped to focus in on specific subgroups, based on location or activity. Finally, streaming was useful because it helped to follow particular lines of enquiry extending from these key streams. By conducting internet-based research on this and related topics, through streaming, hospitality academics can contribute to knowledge outside their subject areas, thus helping to raise the profile and influence of the research community. More specifically, migration has become a sensitive political topic, especially in the current economic climate, so understanding their experiences, for example of adjustment, learning, access to work and social inclusion/exclusion, has clear implications for policy makers regarding the type of migration that should be encouraged and how migrants can be supported to make positive contributions to society. Furthermore, understanding how migrants perceive their work, understanding their career aspirations, appreciating their processes of learning, particularly through hospitality work has clear relevance to practitioners who devise human resource strategies concerning recruitment, training and development as well as retention. 5.2 Streaming in research on consumer experiences Background and purpose. Recent years have seen the development of “foodie” culture and the rise of enthusiastic consumers using the internet to communicate their meal experiences and their passion for food (see Watson et al., 2008). Blogs describing restaurant visits are narratives that weave together a complex set of events into a coherent story and thus help to understand the consumer experiences restaurants offer. Such forms of consumption and their representation through blogging can be considered a form of serious leisure (Stebbins, 1992), which is used by people to construct their identities and build communities of consumption (Watson et al., 2008). Many food bloggers are examples of Stebbins’ (1992) skilled amateurs: they become

self-aware participants in their chosen activity and develop the ability to articulate their emotions and express their opinions. The second case illustrating streaming is a study of blogging, with particular reference to meal and restaurant experiences. Given the central role of identity construction and communities of consumption within food blogging, a key objective of the research was to understand how internet users articulate their identities and engage with networks to mediate their food related experiences. Data collection and streaming. This study also adopted a passive or observational form of online research, although permission was sought from the blogger to study the site. Streaming and the identification of key streams began with a broad search of the internet using “food blogs” as the core search term. From the millions of returns, irrelevant ones, for example those focusing primarily on personal cooking, were rejected. As in the previous case, the sampling for key and sub-streams was purposive, and a specific blog was selected because it was substantial, relevant and active. The key stream, Grab Your Fork (grabyourfork.blogspot.com/, referred to hereafter as GYF) was established in 2004. The blogger, Helen Yee, posts regularly and responds to each comment individually. GYF won of the “Best Restaurant Review Blog” award in 2006, and was a finalist in 2004, 2005, 2007 and 2008 in various categories for food blog awards. In addition, it was listed fourth in a worldwide top ten food blogs by The Times in 2008. Yee is based in Sydney, Australia but comments are made by people from all around the world. The study focused on posts made over 24 months during 2006 and 2007. In total this covers 460 posts plus 2,774 comments by other visitors, but 15 key sub-streams discussing “memorable meals” were identified within this process of streaming. One example of these key sub-streams is her description of a visit to Sydney’s famous Tetsuya’s restaurant in April 2005, which received 51 comments, the latest as recent as September 2009. In this study the starting point for streaming was the individual, Helen Yee, and the study traced her experiences, interactions and relationships extending from her blog. Key themes and emerging issues driving, and emerging from, streaming included presentations of self, particularly as they are articulated through her descriptions of food and eating. Presentations of self were also important issues in other users’ posts. Additional important themes were her extending networks, the reciprocal transactions involved, and the shared codes and meanings in interactions. Indicative emerging themes identified through streaming. The construction of identity and articulations of self were central features of food blogging and Yee was candid about her passion for food and why she blogs: “An unbridled commitment to share, to comment and to espouse the joys of food are the main reasons why many of us blog” (Helen Yee, GYF, 27 January 2010). Her commentaries also revealed other aspects of her identity, for example, her use of English points to her constructed class status: “The roast suckling pig embodies all three in glorious gormandizing hedonistic glory” (Helen Yee, GYF, 2 February 2006). Moreover, her descriptions of eating offal and other animal parts reveal information about her ethnicity and how it defines food habits, for example, “if there are three particular idiosyncrasies about Asian palates it’s our love of bones, skin and fat” (Helen Yee, GYF, 2 February 2006). Beyond Yee’s commentaries, it was useful to trace her interactions through her blog to better understand how communities of consumption were articulated through online interaction. There are numerous visitors to Yee’s blog, some simply “lurking”

Investigative research on the internet 847

IJCHM 24,6

848

(passively observing rather than participating) or making transient visits, but she has developed relationships with several posters. Between them they have established interactional repertoires and codes. For example, they appear to be familiar with technical culinary terms, including “galette”, “confit” and “quenelle”, which are used without explanation. Moreover, Yee and followers of the blog regularly perpetuate interactional routines with comments and descriptions of their meal experiences or of reading other people’s meal experiences: Sorry I meant “yee sang” in my earlier comment. It’s a bit hard to type and drool at the same time as you know (Oslofoodie, GYF, 2 February 2006).

Tracing the interactions, which occur in physical and virtual settings, also reveals the potential role of Yee and her extending network as informed critics who influence purchasing decision making: [. . .] next time try the wat tan hor and fried radish cake which are both very good and if having the ice kacang make sure you ask for them to put enough rose water. you’ll see they put corn in there’s [sic]. some pics of the dishes i’ve had are available to look at [link attached] in my food slideshow (SimonLeong, GYF, 16 October 2006).

Comments such as this, from “SimonLeong”, not only identify subtle details of the meal experience, they also draw upon, and reveal, aspects of their own identity. References to heritage and personal experience are thus entangled with commentary on various aspects of the service operation: I am Burmese and new to Sydney. I have been to several other Burmese restaurants in Asia, UK and US, and I must stay I was rather impressed with the food here. It is authentic indeed and the service from Victor and his staff were [sic] impeccable (Cho, GYF, 16 October 2006).

Furthermore, as the quotes from “SimonLeong” above and “fooDcrazEE” below highlight, interactions between Yee and other blog followers provide links to further sites and discussion threads, which helps the researcher illustrate key points, identify other users and trace further streams. What a feast! The YEE SANG is different from Malaysia though. Take a look here for a post of Malaysian Yee Sang. [Link attached] (fooDcrazEE, GYF, 2 February 2006).

Significance, applications and contributions to knowledge. A streaming approach, in this case taking the individual blogger as the starting point from which to extend investigations, was central to meeting the study’s objectives for several reasons. First, focusing on a committed skilled amateur, such as Yee, provides a rich starting point for examining how individuals express notions of identity through serious leisure. Yee is an extreme case because of her nuanced appreciation of foodways and her commitment to expressing her engagement. Following Yee thus helps to generate a wide set of conceptual themes that are useful when examining the behaviour of other skilled amateurs. Second, Yee is a common point of reference, which other similar enthusiasts use when articulating their own identities and expertise. Linked to this, Yee’s blog is an expressive space where performances of self are mediated and others can develop interactional routines and connect to a broader social network of likeminded individuals. Applying this technique to such social activities also has a number of implications for academia and practice. An analysis of food blogs, again, provides opportunities for

hospitality academics to make contributions to knowledge in other fields, for example in leisure studies about emerging forms of serious leisure, or to social scientific debates about the interactions of identity, networked sociality and food. Researching food blogging also has implications for hospitality marketing and operations management. Key individuals in networks disseminate information, shape tastes and influence purchasing decisions. Therefore, identifying influential individuals, like Helen Yee, understanding their motivations, modus operandi and relationships with other consumers provides clear marketing opportunities. Moreover, their critical commentary on decor, food and service gives managers insights into consumer perceptions of their commercial propositions. 5.3 Streaming in research on emerging hospitality trends and phenomena Background and purpose. The final illustration of streaming takes as its focus an emerging trend or phenomenon in commercial hospitality. This is the most ambiguous of all three types of streaming because trends may actually lack clear definition, and exploratory research offers a way to clarify them and their dynamics. The use of streaming to conceptualise the research process is particularly appropriate in these cases because, although it may be informed by an ethnographic sensibility and desire to gain emic insights, it is problematic to call it netnography. Emerging phenomena are amorphous in nature and there may not be a clear community to which they can be ascribed. Some hospitality phenomena such as “homestay” (Lynch et al., 2009) have evolved to a stage where their manifestations in different international contexts are easy to identify. However, there are many other emerging ones that are only now being investigated and understood. One such phenomenon are “rom” and “kert” venues, which developed in Budapest, Hungary in the beginning of the twenty-first century. The term “rom”, meaning “ruin” in Hungarian, refers to hospitality venues operating in dilapidated buildings in Budapest. “Kert” means “garden” and was used in reference to outdoor venues. The two terms were frequently used in combination because “rom-kert” were located in the open courtyards of buildings, although the two should, in principle, be treated as separate categories of hospitality operation. No research had been conducted on these types of operation so a key objective of this study was to understand what the “propositions of hospitality” were in these venues and how this genre of hospitality emerged. Data collection and streaming. Tracing the evolution of the rom/kert venues helps to illustrate streaming that may, at various stages, involve focus on individuals and specific places, but ultimately focuses on the phenomenon. Initial broad investigations began with general internet searches on commentary that included references to one venue, the Szimplakert, which was considered by many commentators to be pioneering. Moreover, it was important to examine what other venues were discussed in relation to or in reference to Szimplakert. Identifying similar venues and descriptions of them helped to understand these venues, not as individual operations, but as points of reference in a constellation of venues. An important aspect of streaming was identifying key streams, which included cultural commentators and virtual outlets where commentary was located. Three key streams were identified, again, though the focused use of search engines. Index.hu is an established media site, which has a number of writers producing regular features on

Investigative research on the internet 849

IJCHM 24,6

850

culture, politics and urban affairs. One commentator, Andra´s Fo¨ldes, wrote several articles for Index.hu on the emergence and growth of the rom/kert phenomenon. He also collaborated on a guide, distinguishing venues according to customers’ “neediness”/“needlessness”, and discussed elsewhere the struggles venues faced in the 2005 summer season. Szimpla also maintained a press centre on their website, which stored scanned copies of articles about the Szimplakert that usually discussed it alongside other, related venues. Index.hu, Fo¨ldes and the Szimpla repository, cross-referenced each other and became the key streams used to trace the evolution of the rom/kert phenomenon. Key sub-streams were subsequently identified. These were used to pursue the emerging themes in the research including the venues’ defining operational characteristics, variations on the rom/kert model, and the collaborations and conflicts between different stakeholders influencing venues’ survival. Importantly, the information gained from Fo¨ldes, Index and Szimpla’s archived material helped to understand not only the venues as individual commercial entities but how they existed in relation to their social, cultural, physical and political environments. Additional important streams were traced through other sites and individual blogs that made reference to venues and to the rom/kert phenomenon. For example, professional reviews and lay commentaries made repeated reference to particular design features and cultural events hosted in the venues. Some websites contained a single article, followed by comments from different users giving their own accounts of venues. The small number of active users, who engaged in dialogue, made it problematic to treat such idiosyncratic commentaries as authentic representations. Therefore, as Hine (2000) suggests, it was necessary to consider individual commentaries and representations in relation to others, which may not have directly acknowledged, engaged with or responded to them. Streaming, using the phenomenon as the focal point from which to extend enquiry was therefore necessary in constructing a rich story of the rom/kert phenomenon. Interpreting different representations in relation to others also became a form of triangulation. Several attempts were made to contact some of the bloggers and writers who provided insightful accounts, but none were successful. Nevertheless, the study was augmented by visits to venues, the collection of photographic data, and recorded interviews with operators. Indicative emerging themes identified through streaming. The emergence of this hospitality genre can be attributed to other venues operating in Budapest in the 1990s, but commentators agree that a pioneering example of a romkert was the Szimplakert, which opened in 2002. The first Szimplakert was housed in the courtyard of a building awaiting demolition. This venue, including the use of courtyards of empty residential buildings, became the template which was reproduced, albeit slightly differently, in subsequent years. In 2003, two similar venues opened in the district; and by 2004 numerous operators had created their own version of the rom/kert. More recently, numerous other venues have opened in surrounding city districts and several have opened on rooftops rather than in courtyards (see Lugosi et al., 2010). Tracing different representations of venues and their histories also helped to understand the hospitality proposition of the venues and how these propositions were shaped by the social, economic, political and physical environment. For example, the decaying nature of the urban fabric, aging population, rising “rent gap” between the potential value of the land and the income actually generated from rents led to the

emptying of numerous buildings. Many of the venues were established because buildings awaiting demolishment could be occupied by operators, and urban decay, therefore, gave these venues distinct identities. Some of the districts were going through a process of social and economic regeneration. Several operators exploited pauses in the regeneration, for example, while developers and local authorities clashed over development plans. Other operators actively contributed to an area’s regeneration by maintaining a “cultural scene” (see Lugosi et al., 2010). Tracing multiple representations of the venues revealed how distressed, mismatched tables and chairs, table football, lampions and artistic decorations were fundamental to the propositions of hospitality in these spaces. Closely linked to this were the various cultural events hosted by the venues, including book launches, talks and debates, fashion shows and art exhibitions. Moreover, tracing consumer representations of the venues, through streaming, helped to understand how consumers experienced the venues. For example: [. . .] last time I was at Szimpla, a huge norwegian contingent was there – love norwegians dont’ [sic] get me wrong, but not the kind who feel the need to paint their flags on their faces and ladies loo was filled with Aussies. Actually, foreigners don’t bother me, as long as they are the right kind (a bit boho, can string an intelligent sentence together etc) but those trendy fake-nailed, hair-straightened, designer clothes wearing idiots that would be better at the White Party or whatever its [sic] called. Leave the romkocsmak to the rest of us who probably wont’ [sic] be spending three hours to get ready to go there (pestcentric.com, 8 May 2007 [emphasis added]).

These types of commentaries provide critical, though fragmented, information about consumer perceptions of the venues and their customer profiles, including how both have changed. Moreover, such reflections on the scene offered insights into commentators’ (sub)cultural values and their sense of identities. Significance, applications and contributions to knowledge. The use of streaming was particularly important in understanding the evolution of this genre and the propositions involved. Because of its fragmented, evolving nature, identifying key and sub-streams helped to create a composite picture of rom/kert hospitality. Beyond the transactions of food and drink, multiple cultural activities were entangled in the genre. Therefore, only by tracing multiple representations, which included those of cultural commentators, critics, journalists, operators and customers, could the complex dynamics of its development be understood. The examination of the rom/kert and similar trends and phenomena has a number of implications. Firstly, understanding this interaction between venues and urban transformation contributes to debates among academics working in areas of geography and social policy, and, potentially, policy makers, concerning the multiple contributions hospitality can make in urban regeneration. Secondly, for commercial operators, investigating the dynamics of this genre of hospitality can inform their business strategies in occupying unorthodox physical spaces and utilising internet-based consumer-to-consumer interaction to understand consumer tastes and create demand. Furthermore, investigating this genre can inform operational policies concerning music, decoration and cultural activities used to define their commercial propositions. Pursuing these lines of enquiry, through streaming, again, provides hospitality researchers opportunities to contribute to wider academic, cultural, commercial and policy agendas.

Investigative research on the internet 851

IJCHM 24,6

852

6. Conclusion This paper has discussed netnography and introduced the notion of streaming to conceptualise a related investigative process on the internet. As we argued, online research might be informed by the principles of netnography, and, more broadly, ethnography, but it may be problematic to call all such investigative internet research “netnography”. In short, all netnography involves streaming, but not all IRI involving streaming is netnography. We introduced three studies to demonstrate different aspects of streaming, including the identification and tracing of key and sub-streams. We suggested that the researcher can begin tracing streams of enquiry from three initial points of focus: (1) a specific virtual space; (2) an individual; and (3) an emerging trend or phenomenon. Table I summarises potential themes and issues to be considered in the different focuses of streaming, which were highlighted in the three cases. Such investigative techniques can be used within exploratory studies, which have individual value (see Stebbins, 2010); but, they may also be deployed within broader data mining exercises. IRI can be used to identify both key “concepts” (e.g. types of behaviour, relationships, venues, (sub)cultural codes etc) and “contexts” (i.e. specific sites or other online sources of data) (Lau et al., 2010). These can then be analysed quantitatively to discover new patterns and test hypotheses (Lau et al., 2010). It is important to acknowledge the challenges and limitations associated with the investigative approaches discussed here. Engaging in unobtrusive research limits the data to that which occurs naturally. It also raises a number of ethical questions, concerning covert research, invasion of privacy and the absence of informed consent (see Lugosi, 2006). Even if greater interaction is sought, engaging with active users has fundamental sampling biases, and although online investigative research can provide access to many sections of the population that may otherwise remain invisible, it still cannot account for how “passive” users receive or react to online material. Finally, as with many other types of qualitative research, online investigations may be criticised for their lack of generalisability, although multiple investigative studies can be “concatenated” (Stebbins, 2010), and, as noted above, IRI can become the first stage in a confirmatory, quantitative exercise.

Initial focus

Table I. Initial focus and potential investigative themes in streaming

Space

Potential themes and Topics of discussion issues in streaming Uses of space Users of space Interactions, including collaborations and conflicts

Individual

Phenomenon

Performances of selves Extending networks Codes and shared meanings in interactions Transactions and reciprocity

Key actors/agencies involved in creation of phenomenon Phenomenon’s characteristics Evolution of phenomenon Conflicts associated with phenomenon development

Regardless of whether it is used within an exploratory study, or as part of a broader data mining exercise, we have argued that engaging in this type of online research has broad applications in examining the experiences of employees, consumers and in developing human resource, marketing and operational strategies. Moreover, as the cases highlight, such research offers hospitality academics opportunities to contribute to broader disciplinary and policy debates. The ubiquity of the internet and the access it provides to data will undoubtedly lead to its increasing use in the future among academics. However, a remaining challenge, certainly for the readership of this journal, is to translate these investigative techniques so commercial operators can also explore and exploit the internet to develop sustainable business practices. References Beaven, Z. and Laws, C. (2007), “‘Never let me down again’: loyal customer attitudes towards ticket distribution channels for live music events: a netnographic exploration of the US leg of the Depeche Mode 2005-2006 world tour”, Managing Leisure, Vol. 12, pp. 120-42. Dholakia, N. and Zhang, D. (2004), “Online qualitative research in the age of e-commerce: data sources and approaches”, Forum: Qualitative Social Research, Vol. 5 No. 2, available at: http://nbn-resolving.de/urn:nbn:de:0114-fqs0402299 (accessed 10 August 2010). Falzon, M.A. (Ed.) (2009), Multi-sited Ethnography, Ashgate, Abingdon. Fielding, N.G., Lee, R.M. and Blank, G. (Eds) (2008), The Handbook of Online Research Methods, Sage, London. Grossnickle, J. and Raskin, O. (2001), Handbook of Online Marketing Research, McGraw-Hill, Blacklick, OH. Guimara˜es, M.J.L. (2005), “Doing anthropology in cyberspace: fieldwork boundaries and social environments”, in Hine, C. (Ed.), Virtual Methods: Issues in Social Research on the Internet, Berg, Oxford, pp. 141-56. Hannam, K., Sheller, M. and Urry, J. (2006), “Editorial: mobilities, immobilities and moorings”, Mobilities, Vol. 1, pp. 1-22. Hewson, C., Yule, P., Laurent, D. and Vogel, C. (2003), Internet Research Methods: A Practical Guide for the Social and Behavioural Sciences, Sage, London. Hine, C. (2000), Virtual Ethnography, Sage, London. Hine, C. (Ed.) (2005), Virtual Methods: Issues in Social Research on the Internet, Berg, Oxford. Janta, H. (2011), “Polish migrant workers in the UK hospitality industry: profiles, work experience and methods for accessing employment”, International Journal of Contemporary Hospitality Management, Vol. 23, pp. 803-19. Janta, H., Ladkin, A., Brown, L. and Lugosi, P. (2011), “Employment experiences of Polish migrant workers in the UK hospitality sector”, Tourism Management, Vol. 32, pp. 1006-19. Janta, H., Lugosi, P., Brown, L. and Ladkin, A. (2012), “Migrant networks, language learning and tourism employment”, Tourism Management, Vol. 33, pp. 431-9. Jones, S.G. (1998), “Information, internet, and community: notes toward an understanding of community in the information age”, in Jones, S.G. (Ed.), Cybersociety 2.0: Revisiting Computer-mediated Communication and Community, Sage, London, pp. 1-35. Kozinets, R.V. (2002), “The field behind the screen: using netnography for marketing research in online communities”, Journal of Marketing Research, Vol. 39, pp. 61-72. Kozinets, R.V. (2010), Netnography: Doing Ethnographic Research Online, Sage, London.

Investigative research on the internet 853

IJCHM 24,6

854

Langer, R. and Beckman, S.C. (2005), “Sensitive research topics: netnography revisited”, Qualitative Market Research: An International Journal, Vol. 8, pp. 189-203. Lau, K.N., Lee, K.H. and Ho, Y. (2010), “Text mining for the hotel industry”, in Enz, C.A. (Ed.), The Cornell School of Hotel Administration Handbook of Applied Hospitality Strategy, Sage, Thousand Oaks, CA, pp. 66-84. Leander, K.M. and McKim, K.K. (2003), “Tracing the everyday ‘sitings’ of adolescents on the internet: a strategic adaptation of ethnography across online and offline spaces”, Education, Communication and Information, Vol. 3, pp. 211-40. Lofland, J. and Lofland, L.H. (1995), Analysing Social Settings: A Guide to Qualitative Observation and Analysis, 3rd ed., Wadsworth, Belmont, CA. Lugosi, P. (2006), “Between overt and covert research: concealment and revelation in an ethnographic study of commercial hospitality”, Qualitative Inquiry, Vol. 12, pp. 541-61. Lugosi, P. (2009), “Ethnography, ethnographers and hospitality research: communities, tensions and affiliations”, Tourism and Hospitality: Planning and Development, Vol. 6, pp. 95-107. Lugosi, P., Bell, D. and Lugosi, K. (2010), “Hospitality, culture and regeneration: urban decay, entrepreneurship and the ‘ruin’ bars of Budapest”, Urban Studies, Vol. 47, pp. 3079-101. Lynch, P., McIntosh, A. and Tucker, H. (Eds) (2009), The Commercial Home: An International Perspective, Routledge, London. Markham, A.N. and Baym, N.K. (Eds) (2009), Internet Inquiry: Conversations about Method, Sage, London. Miller, T.W. (2005), Data and Text Mining: A Business Applications Approach, Pearson, Upper Saddle River, NJ. Patton, M. (2002), Qualitative Research and Evaluation Methods, 3rd ed., Sage, Thousand Oaks, CA. Pink, S. (2007), Doing Visual Ethnography, 2nd ed., Sage, London. Poria, Y. and Oppewal, H. (2003), “A new medium for data collection: online news discussions”, International Journal of Contemporary Hospitality Management, Vol. 15, pp. 232-6. Stebbins, R.A. (1992), Amateurs, Professionals and Serious Leisure, McGill-Queens’s University Press, Montreal. Stebbins, R.A. (2010), “The internet as a scientific tool for studying leisure activities: exploratory internet data collection”, Leisure Studies, Vol. 29, pp. 469-75. Strauss, A. and Corbin, J. (1998), The Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory, 2nd ed., Sage, Thousand Oak, CA. Urry, J. (2007), Mobilities, Polity Press, Cambridge. Watson, P., Morgan, M. and Hemmington, N. (2008), “Online communities and the sharing of extraordinary restaurant experiences”, Journal of Foodservice, Vol. 19, pp. 289-302. Wittel, A. (2000), “Ethnography on the move: from field to net to internet”, Forum: Qualitative Social Research, Vol. 1 No. 1, available at: http://nbn-resolving.de/urn:nbn:de:0114fqs0001213 (accessed 10 August 2010). Corresponding author Peter Lugosi can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0959-6119.htm

Structure and process modeling of seemingly unstructured leisure-travel decisions and behavior Drew Martin College of Business and Economics, University of Hawaii, Hilo, Hawaii, USA, and

Structure of leisure-travel decisions 855 Received January 2010 Revised August 2010 Accepted March 2012

Arch G. Woodside Department of Marketing, Carroll School of Management, Boston College, Boston, Massachusetts, USA Abstract Purpose – This paper aims to introduce a structuring and processing model (SPM) as a framework for tourism decision making research. Design/methodology/approach – The study employs McCracken’s long interview to collect data in field settings. The study introduces advances in Mintzberg et al.’s approach to structuring seemingly unstructured decision making to explain travelers’ decision-making processes. Findings – SPM enables mapping and comparing visitors’ plans, motivations, choices, and consequences. The results demonstrate nuanced decision-behavior dynamics and complexities of visitors’ travel-related unconscious/conscious thinking and behavior. Research limitations/implications – SPM does not attempt to generalize findings to large survey samples. Practical implications – Travel planning and execution dynamics dictate that a decision-making funnel metaphor in consumer research does not capture such trip complexity because additional decisions are made when the traveler arrives at the destination. Originality/value – SPM is dynamic and inclusive explaining simultaneous planning elements as well as considering sub-decisions occurring before and after different phases in the process. This model includes both conscious and unconscious internal retrievals as well as contextual influences relating to current planning affect the decision-making process. Keywords Tourist behaviour, Long interviews, Ethnography, Unstructured decision making, Tourism management, Decision making, Modelling, Journey planning Paper type Research paper

Introduction Unmet needs often trigger a process where consumers collect and sort information before and while implementing decisions. The marketing management literature identifies five phases to explain leisure travelers’ trip planning processes. These phases include: (1) problem recognition; (2) information search; (3) evaluation of alternatives;

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 pp. 855-872 q Emerald Group Publishing Limited 0959-6119 DOI 10.1108/09596111211247209

IJCHM 24,6

856

(4) selection and purchase; and (5) post-purchase reflection (e.g. Kotler and Armstrong, 2001, p. 193). This model is describable metaphorically as a decision funnel suggesting consumers consciously collect information and narrow alternatives down to one final answer or choice. The decision funnel has at least two shortcomings when applied to the leisure travel decision process. First, trip planning is not a single decision; travel planning includes a series of sub-decisions (e.g. Brathwaite, 1992; Woodside and Martin, 2008). Leisure travel decisions are a series of interrelated decisions that lead to destination choice and rejection decisions. Once at the location, more decisions are made as the trip unfolds. Leisure travel decision making is a dynamic process involving a series of seemingly unstructured and unique decisions. Second, recent evidence suggests most consumer decision making occurs only partially with conscious awareness (see Bargh, 2002; Woodside, 2010). Zaltman (2003, p. 50) estimates at least 95 percent of cognition occurs below awareness suggesting many decisions are made without cognitive thought. Because conscious and unconscious memories affect a person’s thought process, people have difficulty explaining their own behavior (see Rapaille, 2006). Researchers must tap into the individual traveler’s unconscious mind to interpret the associative and causal processes resulting in conclusions, choices, and actions. Collecting data by long interviews and guiding participants to use reflexive thinking (see Hall, 2004) generates nuanced and gestalt images that provide insights of the traveler’s behavior and place (e.g. destinations, accommodations, attractions) attitudes. When customers describe their experiences in think-aloud introspections, tourism executives may achieve deep understanding of their brand’s (e.g. destination, hotel, car rental, restaurant, theme park) competitive advantages (see Thompson et al., 1989). Travel decision making’s dynamic properties and the need to understand unconscious thinking inform the need for a new model to understand leisure travel. This paper introduces a structuring and processing model (SPM). SPM’s design incorporates tourist decision making’s dynamic process with qualitative data collection designed to uncover unconscious memories. SPM adapts Mintzberg et al.’s (1976) model of structuring seemingly unstructured decision making to explain travelers’ decision-making processes. SPM shows how travelers break down complex decisions into manageable decision modules. To uncover unconscious memories influencing traveler decision making, SPM researcher employs the long interview method (McCracken, 1988; Woodside, 2010) and interpretations create thick descriptions of traveler behavior. This ethnographic approach examines socio-historical antecedents to show mapping and description of flows of conjunctive thoughts, decisions, events, and outcomes within specific contexts in leisure travel. From the data, processing and behavior streams surface showing relationships among: . antecedent-to-trip conditions; . trip planning strategies; . destination activities-outcomes; and . outcome evaluations.

The SPM assists researchers in developing gestalt understandings of conscious and unconscious thinking and behaviors (Bargh, 2002; Martin and Woodside, 2011a). Field case studies demonstrate application of SPM to travelers’ decision processes. Consumer decision making Information-processing theory is a tenet of consumer behavior. Information-processing theory’s characteristics are found in early decision-making models (e.g. Engel et al., 1968; Nicosia, 1966). Gilbert (1991) summarizes these studies and concludes six commonalities exist among them: (1) consumer behavior is a constant decision-making process; (2) the individual consumer is emphasized; (3) behavior is utilitarian and can be explained; (4) a buyer actively searches, evaluates, and stores information; (5) collected information is narrowed down to choose alternatives; and (6) final purchases affect future purchases. This rational process assumes choices are deliberate, calculated, and seldom unconscious. A decision maker identifies each future choice’s consequences and chooses the best alternative. Consumer information processing steps require significant inputs from socio-environmental and symbolic stimuli. Automaticity allows some steps to be skipped and the process can be disrupted at any point. Finally, completing the steps does not guarantee purchases. Some scholars identify fundamental flaws in rational choice theory (e.g. Wilson, 2002). Rational choice theory assumes all relevant information is available to decision makers, so decisions are based on all possible alternatives. Growing evidence suggests decisions are less deliberate than rational choice theory proposes. Zajonc (1980, p. 154) proposes “[I]t is further possible that we can like something or be afraid of it before we know precisely what it is and perhaps even without knowing what it is.” This proposition regards unconscious thought as a key process in decision making. Bargh (2002) furthers this proposition by concluding unconscious thinking influences most decisions. Constructive choice theorists apply choice heuristics or information processing shortcuts rather than “omniscient rationality” (see Ladhari, 2007). Although unconscious thinking is relatively unexplored, this theory affirms subconscious information processing’s influence. Constructive choice decision makers seek to maximize the choice accuracy while minimizing cognitive effort. Subconscious heuristic processing creates spontaneous choices. Choices are based on ease of justifying the resolution to peers. Ecological systems theory contends an awareness of a person’s environment helps to understand individual choices and behaviors. This approach “incorporates the interactions between the individual, other individuals, and the social structures of society to explain human development” (Raymore, 2002, pp. 41-42). Research insight increases from considering the individuals’ environmental-life constraints in order to fully understand their behavior. Allen’s (2002) FLAG model applies constructive choice theory to ecological systems theory. Detailed descriptions of the informants’ lives include how surroundings and upbringing shape their current implicit and explicit thoughts.

Structure of leisure-travel decisions 857

IJCHM 24,6

Modeling consumer decision making is complex because several antecedents affect the process. Some decisions are cognitive and require considerable planning and some weighing of features of alternatives; however, other decisions just feel right. Leisure travel decision making likely involves both elements (Martin and Woodside, 2011b). Decision making continues throughout the trip – a dynamic process. Modeling this behavior requires a departure from the traditional consumer decision funnel.

858 The dynamics tourist choices The present study recognizes that interpretive researchers carry some implicit, personal views into the field. Considering ying and yang (i.e. the possibilities of opposable beliefs, rationales, and actions that may be representative in a given case study) propositions help interviewers probe in situ with informants – an abductive stance. Abductive inference permits reasoning by considering the possibilities of alternative planned heuristics and behaviors that may be applicable to a specific case (Eco, 1976; Holbrook and Grayson, 1986). A couple traveling together discusses possible travel destination routes for their upcoming trip before booking air travel tickets (and consider alternative heuristics – possible rules for deciding). The couple reports moving from disagreeing to concurrence before finalizing trip plans (the result) – an illustration of pre-trip negotiations. Prior studies about tourist purchase consumption systems inform how unconscious thinking and ecological systems theory help explain consumer decision making (e.g. Woodside and Dubelaar, 2002). Purchase decisions often are sequential mental and observable steps undertaken by consumers. These acquisitions often lead to a purchase sequence involving other products. Qualitative comparative analyses create useful trip decision typologies (Becker, 1998). Multiple dependent variable influences are described in Woodside et al.’s (2004) thick descriptions demonstrate complex destination behaviors and influence travelers’ thoughts and actions. These variables suggest a dynamic tourist decision process beginning prior to the actual trip decision and continuing after the trip ends. Figure 1 displays nine issues relevant to construct flows of travel decisions and behaviors. These issues focus on destination choices such as antecedents and consequences of implementing the decision. Although not displayed in Figure 2, other tourism foci include decisions such as mode/route, accommodations, and participation in specific activities. Prior studies suggest these issues are relevant to traveler decision making (e.g. Martin and Woodside, 2008), so only a brief description of these relationships appear below. Figure 1 provides an abductive template of topics covered during the long interview process. This ethnographic methodology builds minimal structure into the inquiry so the reviewer benefits from surfacing explicitly his or her own tentative system of if-then propositional relationships before entering the field. The arrows represent tentative propositions relevant to following the flow of the general questions. The following descriptions summarize each proposition. Proposition 1 (P1), box 1 to 2 in Figure 1, proposes visitors’ demographics and lifestyles affect how they frame leisure choices (Fodness, 1992; Hsu et al., 2007). This proposition recognizes household related variables with individual-related factors relate to the decision-making process. Proposition 2 (P2) suggests unexpected or unplanned events occur that may affect the framing of leisure choices. A television

Structure of leisure-travel decisions 859

Figure 1. Theoretical map

IJCHM 24,6

860

Figure 2. Structure and process theory of leisure travel decision making

program or movie featuring a destination may trigger thoughts about planning a visit (see Kim et al., 2007). Proposition 3 (P3) proposes external and internal personal influences affect framing leisure choices (e.g. Baggio and Sainaghi, 2011). Tapachai and Waryszak (2000) break these stored images into functional, social, emotional, conditional, and epistemic values creating beneficial images in tourists’ minds and affecting destination selection. Proposition 4 (P4) proposes features and perceived benefits in framing leisure choices influence the destination choice. Long-term memory information is recalled to working memory tippingthe balance in trip decision-making (Woodside and Lysonski, 1989). Proposition 5 (P5) states that information collected for framing and trip planning affects the process of selecting and rejecting destination alternatives (Lockyer and Roberts, 2009). Conscious and unconscious memory retrieval result in either a cognitive decision making process, or a decision based on gut feeling. In the latter case, informant responses express an inability to provide reasons or explanations beyond simply reporting, “It just felt right” (see Allen, 2002). The external stimuli affect leisure choice framing (P2) and influence the final destination choice. Social forces in travel decisions are particularly strong (Moutinho, 1987). Proposition 6 (P6) contends recalling reference groups’ opinions and thoughts influence the selection or rejection of destination alternatives. Proposition 7 (P7) contends key activity drivers solidify destination decisions such as concrete plans and pre-trip actions. Often multiple activity drivers must line up perfectly for an activity to occur (Woodside and Martin, 2008). Proposition 8 (P8) states key activity drivers affect what is planned and done in a destination. Proposition 8 begins the third phase of the visitor’s unstructured decision making process. Box 7 shows proposition 9 (P9). Visitors interpret events and activities while visiting. This behavior extends Weick’s (1995) proposition about often knowing what we think only after listening to what we say. Surprisingly, experience-related research remains under-represented in tourism research (Ritchie et al., 2011). Finally, proposition 10 (P10) concludes that activities done (and not done) affect the attitude and intention consequences resulting from, and associating with, visiting a destination (see Frazer, 1991). Central to consumption decisions is the proposition that prior purchases and experiences trigger later purchases and recommendations to other people. Structure and process model: dynamic modeling theory of leisure travel decision making Leisure travel process’ complexity and opportunities for structural deviation suggest the decision funnel is imprecise. Many decisions are made after the visitor arrives. Jenkins’ (1978) identifies a series of sub-decisions both before and during the trip. Using Porter’s (1985) value chain, Brathwaite (1992) provides evidence that trip perceptions are based on linked experiences. Sharma and Christie (2010) show how weak elements in the value chain affect trip experiences. These studies suggest tourism decision making’s dynamic process is unstructured and full of contingencies. To simplify decision making, travelers split decisions into smaller and more familiar decisions. For example, planning a trip is broken down into subroutines such as destination choice, transportation, lodging, key activities, and contingency activities. This partitioning is similar to the process organizations employ when making

Structure of leisure-travel decisions 861

IJCHM 24,6

862

infrequently encountered decisions. SPM’s primary departure is accounting for the conscious/unconscious dynamic and personal factors influencing travel decisions. Mintzberg et al.’s (1976) unstructured decision making model explains a process organizations employ when encountering unstructured or infrequently encountered decisions. Their model reduces decisions into identification, development, and selection phases. Once a phase is identified, one or more familiar central routines are enacted to guide the decision maker through the process. Interchangeable sets of familiar routines are applied to the phases to reduce the decision’s complexity. Mintzberg et al.’s (1976) model is more complex than SPM because business and other organizations involve more stakeholder groups. Tourist decision making typically is between two or three key decision makers, so fewer subroutines are needed. Figure 2 modifies and extends Mintzberg et al.’s (1976) general structuring of unstructured (mainly) organizational purchasing decision processes to individual and family leisure-travel decisions. For consistency with Figure 1, Figure 2 includes the numbering of the activity boxes. Phase one represents the foundation for destination selection. This subroutine combines demographics, pre-framing based on conscious and unconscious memories, and external influences to create criteria to decide which destination to choose. Once criteria are decided, phase two may begin. Failure to reach a decision in phase two typically results in returning to the first phase. In phase two, alternative destination choices are evaluated based on key activity drivers (e.g. budget constraints). If this subroutine’s key elements line up perfectly, the traveler books the trip. When all key elements do not line up perfectly, the traveler returns to the first phase. For example, Woodside and Martin (2008) describe a husband and wife that required four key elements to line up in order for their trip to pass through phased two. If one tipping points is not met, the trip is cancelled. Once the trip is booked and the traveler departs, the third phase begins. Phase three occurs during the vacation. This phase is very dynamic because additional stimuli add new variables creating new alternatives and eliminating planned activities. A lost piece of luggage in-transit changes a couple’s plans for their first night at a resort. Positive and negative surprises influence an ongoing evaluation of the trip. Next, the fourth phase is a self-reflection process. The trip’s evaluation affects the next travel planning process’ first phase. Surprisingly, memories are dynamic and the interpretations vary over time (Zaltman, 2003). Stories often are fragments pieced together differently depending on the situation. Memories serve as external influences to reference groups also framing their next trips. Phase 4 is important because destination loyalty and traveler influence on reference group behavior are key ingredients to destination growth. Method Previous consumer research studies suggest direct (reason why) questions are not effective for understanding tourist’s unconscious thinking. Schank (1990) concludes that people tend to process and store information as stories rather than using item-by-item categories. Adaval and Wyer’s (1998) findings confirm this view by showing that information stored as a narrative has more impact on consumers’ judgments than when stored as a list. Cox (1967) offers supporting evidence by understanding grocery shopping behavior through weekly conversations with consumers. Long interviews

enable informants a method to become aware and to report on how cultural and socio-historical forces affecting consumers’ thoughts and actions. To demonstrate application of SPM, an ethnographic field study collected data using McCracken’s (1988) long interview method guidelines. This approach allows respondents to access conscious and unconscious memories. The data unfolds using the structure their minds stored the information (Schank, 1990). As respondents tell their story, they self-interpret the experience providing deeper meaning to their actions and feelings (see Weick, 1995). Based on the propositions in Figure 1, loosely structured and probing follow-up questions were developed in order to try to understand the rationale behind the decision making process, outcomes, and respondents’ feelings throughout their trip. Interviewers were trained to ask probing or follow-up questions in the event that unexpected issues or experiences surfaced during the interview process (e.g. Hsu et al., 2007). They practiced by interviewing each other. The field study includes for 60 to 90 minute, in situ interviews. Interviewing tourists while they are experiencing their visits allows interviewers to be reflexive in qualitative analysis (see Hall, 2004). In situ data collection helps interviewers develop etic interpretations of phenomena. In-person interviews also allowed researchers to observe nonverbal communication modalities. An estimated 70 percent of face-to-face conversation is nonverbal suggesting in-person interviews tell most of the story (Hall, 1973). The sample size is more a function of gathering enough evidence to explain the phenomena than a specific number of interviews. While McCracken (1988) recommends at least five interviews, theoretical sampling achieves the desired results. Theoretical sampling leads to an analysis of specific events and continues until representativeness and consistency are achieved (Corbin and Strauss, 1990, p. 9). The key is collecting evidence to represent a phenomena rather than general findings to a broader population. Interviews took place at locations and situations considered to provide the “greatest opportunity to gather the most relevant data about the phenomena under investigation” (Stauss and Corbin, 1998, p. 208). Security concerns limit some relevant locations (e.g. the local airport). A second challenge is finding people willing to spend 60-90 minutes talking about their vacation. People on a one-week vacation are very protective of their leisure time. Despite these limitations, interviews were conducted at a variety of locations frequented by tourists including hotel’s recreational areas, a high-end tourist shopping mall, informal tourists’ social gatherings, and staging areas where tourists adjusted to the high altitude for star gazing tours. The research team deliberately proceeded to sample tourists so that similarities and differences could be maximized. On-site data collection allowed interviewers opportunities for reflexive interpretation to increase their self-discovery of the assessment process (Hollinshead and Jamal, 2007). Informants were tourists visiting Hawaii’s Big Island (BI) between August and October 2006. Both first-time and repeat visitors participated. Care was taken to assure first-time visitors were interviewed at the end of their trips. With the exception of Japanese tourists, all other visitors were interviewed in English. Two bilingual interviewers were employed for Japanese tourist interviews. Each informant received $50 (USD) and a Hawaii-themed t-shirt for their cooperation.

Structure of leisure-travel decisions 863

IJCHM 24,6

864

Prospective informants were approached and pre-screened with general questions about their visit and whether they would be willing to participate in an interview. Informants agreeing to participate signed a consent form and they were told that their compensation was not dependent on answering all the questions. The questionnaire includes questions asking for: . demographic information about members of the traveling party; . pre-trip planning and sources of information; . activities and destinations – planned and unplanned; . issues surrounding flights, accommodations, and ground transportation; and . overall impressions of their travel experience. On average, interviews lasted about 80 minutes and the informants answered most questions. At the end of the interview, the informants signed a receipt for their cash payment and chose from a variety of Hawaii-themed t-shirts. To prevent potential for myopic interpretation of the results, researchers not involved in the specific interviews analyzed and interpreted the results. Written, thick descriptions were completed for each informant. Each case study report was read and revised by the research team. These protocols ensured consistency in the data interpretation. Findings Due to the length of the thick descriptions, the findings are presented in terms of how they relate to the SPM (Figure 2). First, a brief description of the participants informs each case study report. Helmut Helmut is in his early 60s. He and his wife, Helga, are first-time visitors from Bonn, Germany. They visited Hawaii for three weeks and split time between BI, Oahu, and Kauai. Helmut works full-time as a professional scientist (PhD in physics); Helga is in her late 50s and manages their home full-time which includes three children (17, 18, and 21 years old). This trip denotes the first time that the couple felt confident that all three children could manage without adult supervision. While away, the couple kept in daily contact by telephone with their children back home in Bonn. Helmut’s 90-minute interview was conducted poolside at a hotel in Kailua-Kona. Figure 3 shows Helmut’s dynamic trip modeling process. Being a scientist, Helmut was fascinated about volcanoes. Helmut followed stories about Hawaii’s volcanoes in the news for many years and he consciously and unconsciously thought about visiting BI to see lava flows. The couple saved more than 20 years of Delta frequent mileage points to take this trip. Besides concerns about the children, they were very economy minded. A necessary condition was accumulating enough mileage awards for first class airline tickets. Over the 20-year wait, Helmut likely explored other options (e.g. enter phase two); however, everything did not line up perfectly. Once the award miles were sufficient and the children old enough, the trip planning began.

Structure of leisure-travel decisions 865

Figure 3. Helmut’s structure and process theory of leisure travel decision making

IJCHM 24,6

866

Helga insisted they make a stopover on the US mainland. While Helga supported Helmut’s dream, she wanted to see more than the volcano. Helmut booked a stopover in San Diego. Car rentals and hotels were booked over the internet because the couple was very price sensitive. Despite more that 20 years of dreaming about Hawaii, Helmut spent no time gathering information about the destination. His face showed disappointment when describing the rough beaches and coastlines making swimming difficult. Helmut did not realize his dream to witness lava flowing. The lava flowed during his visit; however, the best vantage point required taking a helicopter ride. Helmut decided not to spend $100 (USD) to charter a helicopter. Given the effort saving for the trip, Helmut’s lack of research and planning is surprising. Helmut even described Hawaii as being located in the South Pacific. The couple avoided shopping and tourist package options, preferring a relatively sedentary BI visit. The evidence suggests the couple unlikely will return to BI in the future. Michiko Michiko is a Japanese woman in her mid-30s. She visited Hawaii in October 2006 with her husband and nine-year-old son. The family lives in Hyogo Prefecture in Japan. The husband works full time for an insurance company; Michiko works part-time at a clothing store. BI was their only destination. They stayed for six nights in the Waikoloa area of BI. Although this trip was Michiko’s second visit to BI, neither her husband nor son had visited Hawaii previously. The 70-minute interview was conducted at the Ellison Onizuka Visiting Center on Mauna Kea. This location is a staging place for people waiting to acclimate to the change in atmospheric pressure. The family was returning from a star gazing tour on top of Manua Kea. Figure 4 shows Michiko’s dynamic trip modeling process. When the family started to think about their annual vacation, Hawaii was on the family’s radar. Guidebooks and travel magazines about Hawaii had been collected and read over the previous months. The choice set was limited to splitting time between Oahu and BI, or just BI. Michiko’s first visit to BI apparently left a strong, positive impression in her unconscious memory because she insisted that the entire vacation be spent on BI. She was very impressed with Volcanoes National Park and Michiko had strong feelings that her son must visit that destination. While phase one really did not require too many decisions, several events had to line up perfectly for phase two. First, the husband only had one specific week that he could take the vacation. If affordable flights and accommodations were not available, the planning process would return to the first phase. The family used their Hilton Travel membership to help plan their trip. Both the hotel and car were booked through Hilton. Michiko booked the air ticket through a travel agent because past bookings always had been less expensive than other alternatives. The biggest surprises were during phase three. While the visit to BI fulfilled Michiko’s wishes to show her son the volcano, some unexpected events affected their plans. After reading in the travel magazine about star gazing, the family was disappointed that their rental car was not powerful enough to drive up the mountain. To go star gazing, they were forced to pay for an expensive tour. The star gazing was fabulous; however, Michiko wished they had rented a four-wheel drive vehicle. Also, the family concluded that a three-night stay was sufficient to see everything on BI.

Structure of leisure-travel decisions 867

Figure 4. Michiko’s structure and process theory of leisure travel decision making

IJCHM 24,6

868

How does SPM explain tourism behavior? Quantitative tourism decision making studies have difficulty structuring questions to capture large sets of perceived attributes (see Mazanec and Strasser, 2007). Open-ended qualitative studies allow respondents to tell stories – allowing data to flow naturally rather than fit into positivist categories. Once data are collected, a flexible model is necessary that allows researchers to explain the dynamic process. Brathwaite’s (1992) value-chain approach partially solves these problems by creating a dynamic series of micro-decision funnels. Micro-decision funnels assume decision making is linear and not very holistic. SPM solves these problems by creating a flexible method to collect information and a structure allowing for non-linear and holistic decision making. For Phase 1, memories, demographics, ecological factors, and unconscious memories influenced how both groups framed decision making criteria. Michiko wanted an educational experience for her son. Helmut and Helga chose Honolulu as the last leg of the journey in case they needed to return home early. External influences also were important for framing the destination choice. Helmut’s years of media exposure to Hawaii, unconsciously affected his behavior. Imagine saving 20 years worth of frequent travel awards for a free trip? Phase 2 moves the decision to making actual plans. Most travelers interviewed searched the internet for the booking because of cost savings. Surprisingly, the internet was not used much to collect information about destination activities. Key activity drivers would make or break the trip. Helmut’s wife was not going to agree to the trip unless they stopped on the mainland US for a few days. Michiko’s husband only had a one-week window of opportunity to travel; a possibility existed that the family would need to travel elsewhere. Activity drivers were attached to the specific destination. Both Helmet and Michiko had strong desires to see the volcano. For most travelers interviewed, destination options really were limited to how much time to spend on each Hawaiian island. SPM’s Phase 3 is an important departure from the traditional consumer decision making. At the destination, visitors encounter unexpected occurrences and activities. Leisure activities can be categorized into four quadrants: planned-done; planned-undone; unplanned-done; and unplanned-undone. Planned-done activities typically are key activity drivers. Visitors’ destination choices are influence by planned participation in these activities. Many travelers do not engage in in-depth planning, so unplanned-done activities may represent the largest share of visitors’ leisure time pursuits (Fodness and Murray, 1999). Planned-undone activities may be due to loss of interest, an unexpected situational contingency, or the result of a tradeoff/replacement with a more desirable activity. Finally, unplanned-undone activities are when an activity is a possibility; however, the visitor does not plan nor engage in the pursuit. The case studies provide evidence that exposure to the destination serves as a catalyst for affecting behavior. Visiting the Mauna Kea caldron was interesting; however, Helmut did not see flowing lava. Unexpectedly, Helmut opted to save the money after waiting at least 20 years to realize his dream. Michiko had to literally shift gears when she discovered her rental car was not powerful enough to climb the steep roads for star gazing. She was forced to spend an unbudgeted amount of money on a start gazing tour. Finally, Phase 4 questions whether or not the customer has developed an attachment to BI. Michiko appears to have an emotional attachment to Hawaii. Her passion for exposing her son to something she enjoyed suggests Michiko will return.

She plans to rent a four-wheel drive vehicle next time. Helmut and Helga traveled a great distance to visit BI. Their comments suggest they boarded the airplane knowing this trip was an once-in-a-lifetime adventure. SPM provides a dynamic method of collecting information on tourist decision making. Although McCracken (1988) recommends as few as five long interviews, small samples make generalizing the results to a population problematic. Long interviews are expensive and time consuming. Training interviewers, traveling to interview sites, and locating willing participants is resource intensive. Providing a sufficient incentive is necessary to encourage tourists to participate. Fifty dollars and a $20 T-shirt were not strong enough incentives for most tourists to forfeit 60 to 90 minutes of their precious vacation time. Nine out of ten qualified informants declined to participate due to the time commitment required. Successful recruits typically were resting between activities. Unfortunately, possibly the best interview location, the airport, is no longer available to researchers because of security concerns. Contributions to theory and practice Empirical positivistic-research methods examine pre-specified relationships to test theories. The preceding analysis provides compelling evidence that complex and extended decision making takes place in leisure travel planning. SPM provides a template to examine qualitative data providing important insights for both practitioners and researchers. The examples demonstrate SPM’s ability to explain how tourist decisions are made. SPM extends Mintzberg et al.’s (1976) unstructured decision making model from business strategy to consumer decision making for high involvement service purchases. Planning a major leisure trip involves many steps and variables. SPM allows decision makers to break down the decisions into manageable feedback loops. Once a phase’s feedback loops are completed, the decision process moves to the next phase. When some missing elements occur, either new scenarios are developed, or the decision moves back to the previous phase – sometimes years later. SPM’s data collection method also considers the importance of conscious and unconscious thinking in travel decisions. The results suggest that measuring promotional tools’ effectiveness is difficult with empirical positivistic research methods. Standardized questions cannot capture these long-term planning elements, particularly when most of the information is filed away subconsciously. Long interviews provide a tool to drill down to uncover motivations and to help explain why people act – even when they are not sure themselves. Services typically are produced and consumed simultaneously and SPM accounts for the decisions that take place on-sight. Often these decisions are made based on learning about new activities, or a change in heart because the conditions are not right. Interviewing in situ allows interviewers to observe external influences on tourists’ decisions. Seeing what the tourist is experiencing helps to interpret the comments. Case studies show systems thinking’s value to examine the influences, choices, activities, and consequences of leisure travel decisions and processes (Woodside, 2010). Both tourism and hospitality managers benefit from this research technique because SPM develops the foundation for customer relationship management. Qualitative interpretive methods often offer deeper insights than self-directed survey closed-end self-reports. They may be more useful for describing consumer behavior.

Structure of leisure-travel decisions 869

IJCHM 24,6

Using SPM as an abductive plan for acquiring qualitative data informs a rich and holistic understanding of tourism behavior. Long interviews asking probing follow-up questions allow researchers to ask case-focused relevant questions for achieving deep understanding of behavior – perhaps etic understandings are better than the interviewees’ own (emic) understandings prior to participating in such long interviews.

870

References Adaval, R. and Wyer, R.S. Jr (1998), “The role of narratives in consumer information processing”, Journal of Consumer Psychology, Vol. 7 No. 3, pp. 207-45. Allen, D.E. (2002), “Toward a theory of tourist choice as socio historically shaped practical experience: the fits-like-a-glove (FLAG) framework”, Journal of Consumer Research, Vol. 28 No. 4, pp. 515-32. Baggio, R. and Sainaghi, R. (2011), “Complex and chaotic tourism systems: towards a quantitative approach”, International Journal of Contemporary Hospitality Management, Vol. 23 No. 6, pp. 840-61. Bargh, J.A. (2002), “Losing consciousness: automatic influences on consumer judgment, behavior, and motivation”, Journal of Consumer Research, Vol. 29 No. 2, pp. 280-5. Becker, H.S. (1998), Tricks of the Trade, University of Chicago Press, Chicago, IL. Brathwaite, R. (1992), “Value-chain assessment of the travel experience”, Cornell Hotel and Restaurant Administration Quarterly, Vol. 33 No. 5, pp. 41-9. Corbin, J. and Strauss, A. (1990), “Grounded theory research: procedures, canons, and evaluative criteria”, Qualitative Sociology, Vol. 13 No. 1, pp. 3-21. Cox, D.F. (1967), “Risk-taking and information handling in behavior – an intensive study of two cases”, in Cox, D.F. (Ed.), Risk-taking and Information Handling in Consumer Behavior, Harvard Business School Press, Cambridge, MA, pp. 34-81. Eco, U. (1976), A Theory of Semiotics, Indiana University Press, Bloomington, IN. Engel, J.F., Kollat, D.J. and Blackwell, R.D. (1968), Consumer Behavior, Holt, Rinehart, and Winston, New York, NY. Fodness, D. (1992), “The impact of family life cycle on the vacation decision making process”, Journal of Travel Research, Vol. 16 No. 4, pp. 8-13. Fodness, D. and Murray, B. (1999), “A model of tourism information search behavior”, Journal of Travel Research, Vol. 37 No. 3, pp. 220-30. Frazer, K. (1991), Bad Trips, Doubleday, New York, NY. Gilbert, D.C. (1991), “Consumer behavior in tourism”, in Cooper, C.P. (Ed.), Progress in Tourism, Recreation and Hospitality Management, Vol. 3, Belhaven Press, Lymington, pp. 78-105. Hall, E.T. (1973), The Silent Language, Anchor Press, New York, NY. Hall, M. (2004), “Reflexivity and tourism research”, in Phillimore, J. and Goodson, L. (Eds), Qualitative Research in Tourism: Ontologies, Epistemologies and Methodologies, Routledge, New York, NY, pp. 135-55. Holbrook, M.B. and Grayson, M.W. (1986), “The semiology of cinematic consumption: symbolic consumer behavior in Out of Africa”, Journal of Consumer Research, Vol. 13 No. 3, pp. 374-81. Hollinshead, K. and Jamal, T.B. (2007), “Tourism and ‘the third ear’: further prospects for qualitative inquiry”, Tourism Analysis, Vol. 12 Nos 1/2, pp. 85-129. Hsu, C.H.C., Cai, L.A. and Wong, K.K.F. (2007), “A model of senior tourism and motivations – anecdotes from Beijing and Shanghai”, Tourism Management, Vol. 28 No. 5, pp. 1262-73.

Jenkins, R.L. (1978), “Family vacation decision making”, Journal of Travel Research, Vol. 16 No. 4, pp. 2-7. Kim, S.S., Agrusa, J., Lee, H. and Chon, K. (2007), “Effects of Korean television dramas on the flow of Japanese tourists”, Tourism Management, Vol. 28 No. 5, pp. 1340-53. Kotler, P. and Armstrong, G. (2001), Principles of Marketing, 9th ed., Prentice Hall, Upper Saddle River, NJ. Ladhari, R. (2007), “The effects of consumption emotions on satisfaction of word-of-mouth communications”, Psychology and Marketing, Vol. 24 No. 12, pp. 1085-108. Lockyer, T. and Roberts, L. (2009), “Motel accommodation: trigger points to guest accommodation selection”, International Journal of Contemporary Hospitality Management, Vol. 21 No. 1, pp. 24-37. McCracken, G. (1988), The Long Interview, Sage, Newbury Park, CA. Martin, D. and Woodside, A.G. (2008), “Grounded theory of international tourism behavior”, Journal of Travel and Tourism Marketing, Vol. 24 No. 4, pp. 245-58. Martin, D. and Woodside, A.G. (2011a), “Gestalt modeling of international tourism behavior: advancing grounded theory following etic interpreting of stories Japanese tell about their visits to the United States”, Psychology & Marketing, Vol. 28 No. 10, pp. 998-1026. Martin, D. and Woodside, A.G. (2011b), “Tourists’ dual-processing accounts of reasoning, judgment, and actions”, International Journal of Culture, Tourism and Hospitality Research, Vol. 5 No. 2, pp. 195-212. Mazanec, J.A. and Strasser, H. (2007), “Perception-based analysis of tourism products and service providers”, Journal of Travel Research, Vol. 45 No. 4, pp. 387-401. Mintzberg, H., Raisinghani, D. and The´oreˆt, A. (1976), “The structure of ‘unstructured’ decision processes”, Administrative Science Quarterly, Vol. 21 No. 2, pp. 246-75. Moutinho, L. (1987), “Consumer behavior in tourism”, European Journal of Marketing, Vol. 21 No. 10, pp. 3-44. Nicosia, F.M. (1966), Consumer Decision Process: Marketing and Advertising Implications, Prentice Hall, Englewood Cliffs, NJ. Porter, M. (1985), Competitive Advantage – Creating and Sustaining Superior Performance, The Free Press, New York, NY. Rapaille, C. (2006), The Culture Code: An Ingenious Way to Understand Why People around the World Live and Buy as They Do, Broadway, New York, NY. Raymore, L. (2002), “Facilitators to leisure”, Journal of Leisure Research, Vol. 34 No. 1, pp. 37-51. Ritchie, J.R., Tung, V.W.S. and Ritchie, R.J.B. (2011), “Tourism experience management research: emergence, evolution, and future directions”, International Journal of Contemporary Hospitality Management, Vol. 23 No. 4, pp. 419-38. Schank, R.C. (1990), Tell Me a Story: Narrative and Intelligence, Northwestern University Press, Evanston, IL. Sharma, A. and Christie, I.T. (2010), “Performance assessment using value-chain analysis in Mozambique”, International Journal of Contemporary Hospitality Management, Vol. 22 No. 3, pp. 282-99. Strauss, A. and Corbin, J. (1998), Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory, 2nd ed., Sage, Thousand Oaks, CA. Tapachai, N. and Waryszak, R. (2000), “An examination of the role of beneficial image in tourist destination selection”, Journal of Travel Research, Vol. 39 No. 1, pp. 37-44.

Structure of leisure-travel decisions 871

IJCHM 24,6

872

Thompson, C.R., Locander, W.B. and Pollio, H.R. (1989), “Putting consumer experience back into consumer research: the philosophy and method of existential-phenomenology”, Journal of Consumer Research, Vol. 16 No. 2, pp. 133-46. Weick, K.E. (1995), Sensemaking in Organizations, Sage, Thousand Oaks, CA. Wilson, T.D. (2002), Strangers to Ourselves: Discovering the Adaptive Unconscious, Harvard University Press, Cambridge, MA. Woodside, A.G. (2010), Case Study Research, Emerald Group Publishing, Bingley. Woodside, A.G. and Dubelaar, C. (2002), “A general theory of tourism consumption systems: theory and empirical research”, Journal of Travel Research, Vol. 41 No. 2, pp. 120-32. Woodside, A.G. and Lysonski, S. (1989), “A general model of traveler destination choice”, Journal of Travel Research, Vol. 27 No. 4, pp. 8-14. Woodside, A.G. and Martin, D. (2008), “Applying ecological systems and micro-tipping point theory for understanding tourists’ leisure destination behavior”, Journal of Travel Research, Vol. 41 No. 1, pp. 14-24. Woodside, A.G., MacDonald, R. and Burford, M. (2004), “Grounded theory of leisure travel”, Journal of Travel and Tourism Marketing, Vol. 17 No. 1, pp. 7-39. Zajonc, R.B. (1980), “Feeling and thinking: preferences need no inferences”, American Psychologist, Vol. 35 No. 2, pp. 151-75. Zaltman, G. (2003), How Customers Think, Harvard Business School Press, Boston, MA. Corresponding author Drew Martin can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0959-6119.htm

Using focus groups as a tool to develop a hospitality work-life research study John W. O’Neill School of Hospitality Management, The Pennsylvania State University, University Park, Pennsylvania, USA Abstract

Using focus groups as a tool

873 Received 14 October 2010 Revised 4 February 2011 6 April 2011 Accepted 20 September 2011

Purpose – This paper seeks to describe how the results of the qualitative research method of focus groups may be used as conceptual data at the onset of a research study to inform researchers regarding relevant issues for future more in-depth quantitative study. Design/methodology/approach – Seven focus group sessions with a total of 50 participants were conducted, each focus group with six to eight participants. All focus groups included the homogeneous participants of new entrants to the hospitality industry. Focus group questions were inductive and naturalistic and centered on career expectations and work-life issues. Sessions averaged 1 hour and 15 minutes and were conducted by trained graduate students. Findings – The paper suggests that long, unpredictable hours create both work-related and non-work stress. Further, there is general agreement regarding the stressors and benefits associated with working in the hospitality industry. Research limitations/implications – Limitations include the use of senior, hospitality management majors, all of whom had hospitality industry employment experience, but some of whom had fewer than 1,000 hours of such experience. Originality/value – This work illustrates how focus groups may fit into a larger research study involving the hospitality industry. This work also explores the common issue, but understudied topic of work-life balance in the hotel industry. In so doing, it provides greater understanding of the issue to researchers. Keywords Focus groups, Hospitality, Work-life balance, Qualitative research, Work-related stress, Non-work stress, Stress, Hospitality management Paper type Research paper

Introduction Some of the most significant and robust contributions to the hospitality literature have been made by studies employing qualitative methods (Cohen, 1988), and more qualitative methods are needed in hospitality research (Mehmetoglu and Altinay, 2006). At the commencement of a large-scale research study regarding work and life issues in the hospitality industry, a qualitative, conceptual, exploratory understanding of the work and life issues of entrants to the hospitality industry can be obtained through focus groups. The focus group method provides data in a social context, and it is an appropriate method for use in the initial development of a larger research This research project was supported by the Alfred P. Sloan Foundation. The contents of this publication are solely the responsibility of the author and do not necessarily represent the official views of this foundation. Special acknowledgement goes to Kathy Christensen, Nan Crouter, Jan Cleveland, Anna Mattila, and Amit Sharma.

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 pp. 873-885 q Emerald Group Publishing Limited 0959-6119 DOI 10.1108/09596111211247218

IJCHM 24,6

874

program (Bojlen and Lunde, 1995). The present work portrays outcomes of the initial conceptual step in a program of research regarding work and life issues in the hotel industry. This work aimed to obtain first-hand information from senior hospitality management majors by conducting focus groups. This work further aimed to analyze all of the data and identify areas of similarities and differences across the focus groups. According to Bojlen and Lunde (1995), focus group research is a qualitative method with interesting properties. It is performed by planned discussion and interview with a small group of people conducted by a moderator. The participants are sampled from the study population. The aim is to obtain knowledge of the participant’s considerations and ideas regarding a relatively specific topic. The method is feasible in illuminating the variation of viewpoints held in a population. It is used as a single source of data or in combination with other methods. As the method provides data in a social context, it is used as an alternative to individual interviews, when appropriate, or in the initial development of a research program (Bojlen and Lunde, 1995). The method typically provides the researcher with with more surprises than other research methods (Grudens-Schuck et al., 2004). Focus groups have emerged as a popular technique for gathering qualitative data, both among sociologists and across a wide range of academic and applied research areas (Morgan, 1996). Focus groups are currently used as both a self-contained method and in combination with surveys and other research methods, most notably individual, in-depth interviews. Comparisons between focus groups and both surveys and individual interviews help to show the specific advantages of focus groups in producing interaction among participants (Morgan, 1996). Focus groups have challenges. Measuring strength of opinion from focus group data is problematic (Sim, 1998). The indicators used to measure attitudes in orthodox survey research are largely inapplicable to the context of focus groups. When comparing data from different focus groups, inferences may be drawn as to the presence of absence of certain views or issues across groups, but not in terms of their relative strength. Also, both methodological and epistemological objections can be raised against attempts to generalize from focus group data. Theoretical generalization is likely to be more feasible than empirical generalization, and if the latter is considered fruitful, it is likely to be of a provisional nature (Sim, 1998). The subject work accounts for these relevant considerations. Krueger and Casey (2000) summarize that focus groups tend to be an effective research methodology when researchers are seeking to identify a range of ideas or feelings that people have about something, to uncover factors influencing multifaceted concepts, to have ideas emerge from the group, and to later design a larger quantitative study. On the other hand, they indicate that focus groups tend to be ineffective when researchers want participants to arrive at a consensus, want participants to be educated, are seeking sensitive information that should not be shared in a group, or need statistical projections (Krueger and Casey, 2000). Since in the subject work, the researchers sought to identify a range of ideas, to identify factors influencing multifaceted concepts, to have ideas emerge from groups, and to design a larger quantitative study, focus groups are deemed an appropriate method for this work. Further, since the researchers did not want participants to arrive at a consensus, to be educated, to share sensitive information, or need statistical projections, focus groups are deemed an appropriate method for this work. It should be noted that in the subject study, focus groups were one of two qualitative methods initially implemented to inform the researchers regarding relevant issues of

work and life in the hotel industry. The results from this conceptual stage were employed for future, more in-depth quantitative study, including for the development of survey instruments. The other method used for this purpose was in-person interviews with hotel managers taking place in the managers’ homes. This paper, however, concentrates on the focus groups.

Using focus groups as a tool

Work and life issues The hotel industry is a large and important segment of the economy, providing continuous service 24 hours per day, seven days per week (O’Neill et al., 2009). A review of the academic literature shows that work and family issues have received limited attention in hospitality and tourism journals, and there is no evidence that focus groups have been employed to study this multifaceted issue. While other literature in hospitality has focused on new entrants’ perceptions of issues related to work-related job characteristics (e.g. Weaver, 2009), there is a dearth of literature regarding work-family-related issues. Many of the characteristics of jobs in the hospitality field (e.g. long hours scheduled at nonstandard times, on-call hours, emphasis on face time, and geographic mobility as a prerequisite for career advancement), however, have long been associated with work-family concerns in other literatures. There are also undoubtedly important ways in which people’s lives off the job shape their job choices and their experiences on the job in the hospitality industry. Turnover is one of the most prevalent concerns in the hotel industry, which has been characterized as having a “turnover culture” (Deery and Shaw, 1997, p. 377; O’Neill et al., 2009). Wasmuth and Davis (1983a,b,c) were two of the first researchers to investigate turnover in the hotel industry. They found that annual turnover among European and North American hotel employees averaged 60 percent, and was disproportionately high in the food and beverage, front office, and housekeeping departments. Annual turnover rates have been estimated to be 50 percent for hourly and 21 percent for management positions (Smith Travel Research, 2003). During the first decade of the 2000s, annual employee turnover rates were estimated to range between 17 and 28 percent on average throughout the US, with the hospitality industry recording the highest annual turnover rates, ranging from 49 to 60 percent (Hauknecht and Trevor, 2011). Marriott estimated that with each one percent increase in its employee turnover rate, the company loses between $5 and $15 million in revenues (Schlesinger and Heskett, 1991). Low pay, difficult work, and irregular (often unpredictable) work hours have been identified as the three primary obstacles in the recruitment and retention of managers in the hotel industry (Dermody and Holloway, 1998). Work schedules seem to be a particularly critical issue in the hotel industry. In particular, participants in Stalcup’s and Pearson’s (2001) study indicated long hours and the necessity of relocating regularly were key reasons for turnover. In almost every case, when long hours were mentioned as a reason for turnover, the complaint was not regarding working too much, but about not having enough time to spend with families. Recent declines in business travel and tourism forced hotel organizations to critically examine operations and performance and to reassess strategies for gaining competitive advantage in this highly competitive sector of the economy. Given the challenges and difficulties the hospitality industry historically has faced with work-family conflict (e.g. Dermody and Holloway, 1998; Stalcup and Pearson, 2001), hotel organizations may be able to gain strategic advantage over their competitors by effectively addressing work-life issues that affect their employees’ productivity.

875

IJCHM 24,6

876

Specifically, attention to the work and family interface, an increasingly popular issue in contemporary organizational research, may reveal new ways in which hotel companies can position themselves to attract and retain valued employees, improve performance and guest satisfaction, and reduce industry problems such as employee turnover. The subject research tapped the views of entrants to the hospitality industry using focus groups to obtain untainted information regarding work and life expectations. The importance of an individual’s expectations regarding the work experience is noted in research regarding employee turnover. According to the met-expectation hypothesis, an individual is more likely to be dissatisfied or absent from work and leave an organization or an industry when their work experience violates their expectations (Wanuous et al., 1992). Gaining the perspectives of work and life issues of individuals months before their entering the workforce, provides an interesting perspective into factors that predict negative outcomes. Realistic job previews and expectation-lowering procedures have been found to reduce turnover by providing information about both rewards and challenges associated with the job (Buckley et al., 1998). Work and family issues need to be examined to better understand the turnover problem. Cohen (1997) noted that non-work factors play a significant role in predicting employee turnover. Additionally, conflict between work and non-work roles explained significant variance in predicting job dissatisfaction and turnover cognitions of work factors alone (Hom and Kinicki, 2001). The unfolding model of turnover (Lee et al., 1996) suggests that individuals experience a “shock to the system,” often due to non-work factors in which they evaluate their job and make a decision to leave. Method Krueger and Casey (2000) indicate that focus groups should be comprised of a relatively homogeneous group of people. An earlier study in the International Journal of Contemporary Hospitality Management that polled 21 participants concluded that future research of new entrants to the hospitality industry would benefit from a larger and more homogeneous sample (Weaver, 2009). This work involved conducting focus groups with a total of 50 students majoring in hospitality management at a large, US university, a relatively homogeneous group. A total of 93 percent of participants who were contacted agreed to participate in the study. There were a total of seven focus groups, ranging in size between six and eight participants each. Krueger and Casey (2000) suggest that focus groups should contain between six and eight participants. All of the participants had experience inside the hospitality industry either professionally and/or through internships. Most of the participants planned to work in hotels, and most planned to become managers in either the rooms or food and beverage areas. Focus group interview questions should be both inductive and naturalistic (Krueger and Casey, 2000). Thus, the subject work did not employ the rigid, a priori approach tyical of a quantitative study. The focus group question instrument for the subject work included a series of relevant questions, many of which were intended to compel participants to consider both the positives and negatives of work in the hotel industry, and some of which asked the participants to consider the issues from different vantage points. Since effective focus group research is intended to illuminate viewpoints regarding a relatively specific topic (Bojlen and Lunde, 1995), the majority of questions focused on work and life in the hotel industry. The questions in the subject work were categorized into four related topic areas:

(1) Immediate career expectations and plans, including expectations about the types of positions they would like to hold, skills needed to succeed, and career mobility. Similarly, participants were asked to identify what positive features of the industry or hotels attracted them to the field, and the features they believe present challenges. (2) Family planning and expectations of partner/spouse, number and timing of children, delaying or sequencing of work and family commitments and expectations regarding the division of labor in the home. (3) Work and family life concerns. (4) Suggestions regarding corporate human resource innovations within the hospitality industry. The instructions clearly stated that responses would be anonymous, that respondents did not need to answer any particular question, and that they were free to leave at any time, consistent with the university’s institutional review board requirements. These protections were designed to increase the participants’ comfort level when discussing the issues. In addition, to prevent less vocal respondents from being excluded from the discussion, each participant was asked to introduce him or herself at the beginning, and all participants responded to at least some of the questions. Focus group administration Each facilitator for the focus groups was trained for one hour regarding the protocols. The facilitators were graduate students who were not affiliated with the hospitality school to enhance confidentiality and to help create the comfortable, permissive environment that tends to make focus groups effective (Krueger and Casey, 2000). Facilitators commenced each focus group by making the same statements, requested that subjects read a statement regarding the purpose and intent of the study, and to sign their name to that statement if they agreed to participate. Consistent with recommendations by Morgan (1996), moderators encouraged interaction among participants. The focus groups were conducted in a private meeting room located on the university campus. Participants were recruited from the capstone course for the hospitality major, ensuring that most were seniors and hence were in reasonably close proximity to career if not family issues. Participation in the sessions was voluntary. Each focus group session lasted approximately one hour and 15 minutes. This method was approved by the university’s Institutional Review Board (IRB). Focus group data analyses All focus group discussions were audio-taped and then transcribed by the trained graduate students. The transcribed data were archived in the QSR NVivo 7.0 software program for subsequent analyses. Triangulation of the data revealed significant similarity of the responses among the different focus groups as described in depth in the Results section of this paper. Results Sample characteristics Of the 50 focus group participants whose responses were taped and used in the qualitative analyses, there were 31 females and 18 males (plus one of unidentified gender) and 45 White, four non-White, and one of unknown ethnic background. Given

Using focus groups as a tool

877

IJCHM 24,6

878

the small number of non-white respondents, information is reported regarding the gender of those responsible for specific quotations, but ethnic information is not reported. In general, results from the focus groups paint a consistent picture regarding how life and work operate, blend, and often conflict for hospitality employees. However, the picture is seen from different angles by different participants, so unique contributions are provided by each group of participants, as well. This article continues with the common themes, and then focuses on specific and unique themes provided by each group. Common themes Focus group analysis tends to be based on pattern identification (Krueger and Casey, 2000). The subject work found a pattern that there is something about careers in the hotel industry that attracts certain types of individuals. The focus groups all suggest that working in the hospitality industry is very social, service-oriented, fast-paced, and varied. Many of the participants had weighed alternative career options, and decided on hospitality careers precisely because of these factors. Many participants liked the task variety, while others felt the rapid pace made the work shift pass by quickly. Effective reporting of focus group results includes quotes (Krueger and Casey, 2000). Multiple quotes follow. For example, a female participant claimed, “I like the fast paced world . . . You’re always dealing with different things” while another female participant mentioned, “I need a job where it’s going to be something completely different every day, and it’s going to challenge me to walk into the door. [That’s] the only thing that’s going to get me out of bed”. Some participants liked the idea of service and helping others, and almost all were very socially-oriented. For example, a female participant stated, “[For] any position in this industry, you have to be a people person. You have to like people and you have to not be bothered by them.” One male participant cited: What would normally be a slower period for the hotel over a quarter and you’d have two really crazy busy weeks and the rest of the time really slow. So it really depends. I mean, that’s one of the things that I like about the industry, it’s not just flat line, steady the whole time. Every day is different, every week is different . . . They’re all open 24/7.

Related to the nature of hotels themselves was the notion that, for employees, hotels are often family like. Many participants mentioned hotel events that were held for the employees and their families, that they had friends both on the job and in the industry, and that expressing family concerns and bringing family members to the workplace were welcome, particularly in well-run hotels. For example, a female participant stated: Another good thing about working at a hotel is that there’s a support structure . . . So whether it’s myself or another employee in the hotel, there’s always somebody who can take up or finish or do whatever you do to get through that day. So there’s always that support that anybody can leave and take care of a family.

Similarly, the participants mentioned hotels as having a quality almost like family, with friendliness and support as positive aspects of the jobs. Working time arrangements in the industry often involve hours that are long, unpredictable, and non-standard. For entry-level managers, the jobs may effectively demand almost all of one’s time, and participants were aware of this. For instance, one female participant stated, “[E]very single person I tell them about my major [responds], ‘Do you know the hours?’ ‘You don’t wanna get involved with that . . . ’ And I say ‘Yeah, I know’” Similarly, participants understood that, as managers, they often will be

on call 24/7 and may need to drop everything else at a moment’s notice to solve a crisis at the hotel. The participants also were aware that managerial positions in the industry commonly require work at night, on weekends, and during holidays due to the 24/7 nature of the industry. These working time arrangements may improve as one advances in a career. For example, a male participant noted, “If you work hard in the beginning . . . and you build yourself up to that management position, then you’re the one making the decisions when your vacation time is, what day you want off . . . ” The participants also generally reported that the total number of hours worked generally reduced with advancement. This point should not be overstated. There is no reason to believe that the participants thought that senior managers were working 40-hour weeks. Something like a reduction from a 60 to a 50 (or 55) hour workweek is probably more accurate. Geographic mobility and willingness to relocate are important in the early stages of hotel industry careers. Akin to high levels of time commitment, successful hotel managers must be willing to relocate to advance in the industry. The participants all were aware of this issue. Generally, all focus groups tended to view job-related relocations positively prior to taking on family commitments, but viewed them as more negatively once family responsibilities emerged. Time arrangements and related high levels of time commitment and geographic mobility early in the career are linked to a strong belief in the meritocratic nature of the industry, and existence of fast-track careers. The participants indicated that “paying dues” early in the career, through long hours and willingness to do anything for the job including relocation, would facilitate quick movement up the career ladder to where hours would become more manageable and relocation less frequent. One female participant stated, “If you’re willing to move around, you can advance a lot more quickly than if someone did not . . . ”. Control over hours and reduced relocation as one advances are linked to the job structure of the industry. The participants felt that the perks of advancement came with and in large part were due to differences between the types of jobs available in the industry. For those who moved straight up the career ladder to a corporate office position, hours were expected to become more manageable because corporate offices do not typically operate at night, on weekends, or on holidays. And once in a corporate office job, the need to relocate is less likely. Most participants believed that jobs could be taken later in human resources, marketing, or accounting to escape the 24/7 character of hotel operations. Across all focus groups, concerns were expressed regarding negative outcomes of positions in the hotel industry, including stress and burnout. The high levels of time commitment, energy required, emotion control and task variety, while seen as positive in some sense, were also viewed as problematic in the long-term. Participants worried about stress and burnout, and felt that the demands of a hotel career would eventually decrease over time. One female participant expressed this concern during the focus groups and said, “I worry a lot just about being burned out after eight to ten years . . . You just really invest a lot of yourself and . . . I think it’s very easy to just burn out and just be so tired of working . . . ”. All focus groups felt that delayed childrearing is associated with hotel careers. Whether expressed positively as a way to “have fun” when one is young, or negatively as a cost required by a hotel career; participants felt that delayed childrearing is common among hotel managers. One female participant cited an example of a couple she knew, “They’re in their early 30s and they both were working in the hospitality

Using focus groups as a tool

879

IJCHM 24,6

880

industry and then whenever they were 28, they decided that they wanted to have kids and they knew they couldn’t do it.” Although there are no hard data regarding the phenomenon, its existence seems very likely given the intense demands of hotel careers on employees in their 20s and even beyond. The intense demands of careers in hotel management were also linked to a belief that marital conflict and dissolution are common in the industry. The participants believed that the demands of the careers would eventually lessen, but also wondered how senior managers held their family lives together in the face of these tough careers. The participants also were aware of the frequency of divorce among hotel managers. One female participant recalled: My general manager was the most amazing person you’ll ever meet, the sweetest guy in the world. He wasn’t married. I thought, “Hmmm . . . that’s weird because he’s such a nice guy.” You just totally think he’d be the best husband ever. Come to find out he was divorced because his job just totally consumed him. I mean, he’s (a general manager) amazing but he sacrifice[d] a wife and two kids for it. They said, “You’re not ever here.”

Finally and relatedly, all of the focus groups expressed a belief in the prevalence and advantages of assortive mating. The participants believed that the practice of hoteliers meeting and marrying other hoteliers is common. They did not express the view that assortive mating occurred because it was easy to meet others in the industry. Instead, they believed that only other employees in the industry could understand the often relentless demands of hotel careers. A female participant reflected such a view by saying, “. . . I think that helps a lot sometimes . . . when your spouse is somebody who also works in the industry because they understand. They understand like when you get called in . . . ”. In part, assortive mating may serve to inhibit marital conflict and dissolution. Different themes Also, themes emerged regarding the hospitality industry that either appeared in only a minority of focus groups, or that cast a different light on a theme reported from other groups. Among these differences was the way that the social and service aspects of the industry were viewed. On the one hand, some participants viewed service as almost noble. For example, one female participant stated, “I’m always concerned with other people’s feelings . . . making sure everyone’s happy all the time” Others were more likely, however, to view “caring” as part of the job, and one which, because it involved high levels of emotion control, was difficult to maintain over the course of a career. Another difference between participants surrounded issues of childcare. Many participants frequently expressed a fear of strangers caring for their children, and presumed that either they or their partner would perform most childcare. For example, one male participant relayed his experience with a summer camp and said, “I don’t like childcare. I’ve worked in camps, and the parents drop them off as early as they could and pick them up as late as they can. And that just made me mad. I would never do that to my kids.” For other participants, it was not fear of strangers that concerned them, but rather the high cost of quality child care (particularly relative to the relatively low incomes associated with hotel employment, at least during years of employment), and not being available for their children, particularly over holidays or during important events in a child’s life. The time demands of the

industry were also viewed differently across different groups of participants. Although most of the participants devoted much of their discussions to time demands, a few discussed the issue of the ebbs and flows of time demands. The up and down nature of the jobs, with some days of the week being busier than others, and some weeks and months of the year being more demanding, was explicitly mentioned by a few participants. For example, a male participant characterized such demands as, “Seasonally driven certainly, conventions especially on the catering side there is truly an ebb and flow.” There is no doubt that such ebbs and flows exist, and perhaps most of the participants simply took their existence for granted, and not worth mentioning. Absent further data or sheer speculation, this difference is a puzzle for exploration in future research. Finally, some of the participants discussed childhood experiences through family employment in the industry as a pathway into the industry, but this topic did not arise in most groups. In part, these differences may be attributed to the divergent perspectives of participants in this work. Some participants might view emotion control as noble because it enhances the experiences of and often helps guests. Other participants may find emotion control more of a burden. Similarly, some participants may believe that childcare should be avoided, but could end up using formal or informal care once a child is born. Especially in light of the fact that the subject work included seven different focus groups, however, it is remarkable how many themes were commonly expressed by most groups of participants. Such agreement on similar themes provides reason to believe that this qualitative data have shed light on relevant issues for future quantitative research regarding these issues. Discussion and conclusions The present research provided an opportunity to obtain a targeted, magnified look at the work and family issues associated with future managers in the hotel industry based on the results of focus groups. Focus groups appear to have been effective as a methodology to employ in this work because ideas emerged from the focus groups, and the study identified a range of multifaceted ideas and feelings that participants have regarding work and life in the hotel industry, each of which tend to be the types of outcomes of successful focus groups (Krueger and Casey, 2000). Because the problems of interest span both work and non-work domains, it was useful to have several different focus groups. From a macro-industry view, a unique characteristic of the hospitality industry may be the almost “theater-like” quality of the workplace. Many participants also cited the beautiful settings, the good food and accommodations as benefits, and that interactions with guests were often in a party-like atmosphere. Further, each hotel employee has his or her own role to perform and arrive at work dressed formally (e.g. suits, jackets, ties, etc) or in a uniform (e.g. hourly employees) when interacting with the public. Good performance and success at work are defined as displaying positive affect, being responsive to guest requests (or demands), generally creating a pleasant environment and delivering excellent service. On the other hand, like many occupations today, long and unpredictable hours within hospitality are cited as a major source of stress. One thing that sets the industry apart from other occupations with long and unpredictable hours is that hotel managers must model emotional restraint and regulation and positive affect throughout the day with both external paying guests and internal personnel. Interestingly, it is one of the

Using focus groups as a tool

881

IJCHM 24,6

most positive features of the hotel industry (e.g. the party atmosphere) coupled with one of the industry challenges, the long hours, that provides the basis for this unique characteristic. From the results of the focus groups, it appears that this unique characteristic may be a predictor of hotel industry turnover. As previously discussed, turnover is known to be a significant problem in the hotel industry.

882

Implications According to most participants, the long, unpredictable hours, the norm to be infinitely accessible to employees and guests, and the requirement to consistently show positive affect in social interactions pose significant potential physical, mental and emotional burdens for hotel managers and their families. These burdens may underlie problems such as substance use, infidelity, and divorce. There is a need for researchers in future studies to more fully explore these emotionally arousing interactions from a number of perspectives and research levels. Specifically, there is a need to more closely examine the daily work experiences of hotel managers and understand the relationship between such occupational conditions and their daily experiences on and off the job. For example, a daily diary study that examines managers’ emotional well-being, quality of interpersonal interaction, and work performance in more depth may be informative. One significant contribution of this project is that it tapped the work and life perceptions and expectations of entrants into the hospitality industry. The work psychology and management research consistently demonstrates that employees’ violated expectations (and possibly spousal expectations) are critical predictors of intentions to turnover or voluntarily leave an organization. Therefore, it was important to determine the initial expectations of newly educated hotel industry entrants. The knowledge of work-related expectations for college seniors in focus groups was similar to the realities of work provided by quantitative surveys of hospitality managers in other research (e.g. O’Neill et al., 2009). However, hotel management seniors did not have specific ideas regarding how they would handle family issues and how work and family demands may influence each other. Although they wanted to have children (most after approximately 10 years of employment), they generally did not know how they would manage childcare and maintain a marital relationship while working in the hospitality industry. An implication for the practice of hospitality management education may be for faculty to provide hospitality management seniors with a more complete and realistic preview regarding not only the benefits of working in the industry, but also challenges regarding the management and balance of both work in the hospitality industry and simultaneous family demands from both spouses and children. An implication for the practice of hospitality management is that such realistic job previews could become a component of new employee orientations programs. It is plausible that during the first ten years in the hospitality industry, greater formation and development of family expectations may occur among hotel managers. During this time period, individuals may examine their personal life priorities, and reassess and revise their work and life expectations. Also during this time, work and family issues (and possible violated expectations or perceived mismatches) may significantly contribute to managerial voluntary turnover. It is necessary for future research projects to more carefully study hotel managers to assess the development of family formation expectations, the integration of work and family, and the perceptions of possible work-family mismatch or violations of expectations, work strain and voluntary turnover.

Finally, like most industries, the labor force of the hotel industry has a pyramidal structure. At the top of the pyramid are corporate executives who oversee hotel operations, plan and develop new hotels, and build strategies to competitively position themselves among their industry competitors. At the bottom of the pyramid are hourly employees: the housekeepers, front desk clerks, and food service workers who are primarily response for the daily operations of hotels. In the middle of the pyramid are hotel-level managers who act as liaisons between the top and bottom of the organization as well as outside the organization. That is, managers are responsible for connecting with hourly employees, guests, vendors, and corporate headquarters. The long and unpredictable hours that seem to unbalance hotel managers’ lives may result from juggling the strategic goals of top-level corporate strategists and the needs of hourly employees. Future research should involve more detailed studies to determine whether and how corporate executives in the hotel industry conceptualize the quality of hotel managers’ and hourly employees’ work lives as a strategic issue, one that, if handled innovatively and well, could help the organization to recruit and retain strong, talented employees, save recruitment, training, and health care costs, and help the organization compete. In other words, hospitality organizations proactively addressing work and life balance issues could in doing so obtain competitive advantage. In addition, there is a need to interview incumbents of hourly jobs in the hotel industry. Often these jobs require little education or training, offer relatively low compensation and minimal benefits or flexibility, and are frequently characterized by high levels of turnover, which creates costs and other challenges for hotel managers, organizations, and the industry. Limitations As with all research, strengths and weaknesses are present in the sampling and methodology of the present study. The inclusion of hotel management students provided a unique understanding of issues that novices to the industry face, which was enriched by interactive group discussion in a focus group setting (Morgan, 1997). This approach also may have systematically contributed to biased responses (Maxwell, 1992). Further, the results may not be entirely applicable to practicing managers. Future research, such as conducting exit interviews, is necessary to fully explore this possible sampling bias and better understand the role family or life issues may play in turnover. Also, as previously discussed, while the study reported in this paper concentrates on focus groups, a simultaneous project involving in-person, in-home interviews with hotel managers was conducted to improve the generalizability of the findings. It should be noted that there was general agreement between these two groups of participants regarding issues and concerns pertaining to working in the hospitality industry and managing/balancing family and home demands. Again, though the subject of this paper is on focus groups, both the focus groups and the managerial interviews assisted the researchers with forming a larger, quantitative study regarding work and life in the hotel industry, including the development of survey instruments, a typical use of focus groups. A large percentage of potential participants who were contacted agreed to participate in the study. However, a small percentage did not participate, indicating the possibility that those truly experiencing serious time burdens and conflict may not have believed they had the time or resources to commit to a 1.25-hour focus group. Additionally, such individuals may not have desired to discuss work-life issues. Of those who agreed to participate, social desirability (Podsakoff et al., 2003) may have

Using focus groups as a tool

883

IJCHM 24,6

884

biased their responses. Although various attempts were made to create a safe, secure, comfortable, and permissive environment as recommended by previous research involving focus groups (Krueger and Casey, 2000), and guarantee confidentiality of responses, some individuals may have been unwilling to disclose sensitive negative information that they may have believed would reflect poorly on their character or abilities. Although qualitative methods such as focus groups with open-ended, semi-structured interviewing allow researchers to gain a deep and rich sense of how the work and life domains interact in the hospitality industry (Lee, 1999), drawbacks cannot be overlooked. The quality of responses are still subject to participant recollection of events and what they deemed to be worthy of mentioning (Silverman, 2000). While interview techniques may not be the optimal method of gathering objective information surrounding an event, indicating decreased descriptive validity, more relevant to this work is what specific events actually mean to those engaged in them, indicating high interpretative validity (Maxwell, 1992). Additionally, gathering several focus groups provided an opportunity to generate multiple perspectives regarding the same events or issues. In the subject work, the focus group results can be used to later design a larger quantitative study regarding work and life in the hotel industry, which is a typical use of focus group results (Krueger and Casey, 2000). References Bojlen, N.S. and Lunde, I.M. (1995), “Focus group interview as a qualitative research method”, Ugeskr Laeger, Vol. 157 No. 23, pp. 3315-8. Buckley, R.M., Fedor, D.B., Veres, J.G., Wiese, D.S. and Carraher, S.M. (1998), “Investigating newcomer expectations and job-related outcomes”, Journal of Applied Psychology, Vol. 83, pp. 452-61. Cohen, A. (1997), “Nonwork influences on withdrawal cognitions: an empirical examination of an overlooked issue”, Human Relations, Vol. 50, pp. 1511-36. Cohen, E. (1988), “Traditions in the qualitative sociology of tourism”, Annals of Tourism Research, Vol. 15 No. 1, pp. 29-46. Deery, M. and Shaw, R. (1997), “An exploratory analysis of turnover culture in the hotel industry in Australia”, International Journal of Hospitality Management, Vol. 16 No. 4, pp. 375-92. Dermody, M.B. and Holloway, R.W. (1998), “Recruitment and retention of managers”, The Cornell Hotel and Restaurant Administration Quarterly, Vol. 39 No. 6, December, pp. 20-5. Grudens-Schuck, N., Allen, B.L. and Larson, K. (2004), Focus Group Fundamentals, Iowa State University Extension, Ames, IA. Hauknecht, J.P. and Trevor, C.O. (2011), “Collective turnover at the group, unit, and organizational levels: evidence, issues, and implications”, Journal of Management, Vol. 37 No. 1, pp. 352-88. Hom, P.W. and Kinicki, A.J. (2001), “Toward a greater understanding of how dissatisfaction drives employee turnover”, Academy of Management Journal, Vol. 44, pp. 975-87. Krueger, R.A. and Casey, M.A. (2000), Focus Groups, 3rd ed., Sage Publications, Thousand Oaks, CA. Lee, T.W. (1999), Using Qualitative Methods in Organizational Research, Sage Publications, Thousand Oaks, CA. Lee, T., Mitchell, T., Wise, L. and Fireman, S. (1996), “An unfolding model of voluntary employee turnover”, Academy of Management Journal, Vol. 39 No. 1, pp. 5-36.

Maxwell, J.A. (1992), “Understanding validity in qualitative research”, Harvard Educational Review, Vol. 62, pp. 279-300. Mehmetoglu, M. and Altinay, L. (2006), “Examination of grounded theory analysis with an application to hospitality research”, International Journal of Hospitality Management, Vol. 25, pp. 12-33. Morgan, D.L. (1996), “Focus groups”, Annual Review of Sociology, Vol. 22, pp. 129-52. Morgan, D.L. (1997), Focus Groups as Qualitative Research, Sage Publications, Thousand Oaks, CA. O’Neill, J.W., Harrison, M.M., Cleveland, J., Almeida, D., Stawski, R. and Crouter, A.C. (2009), “Work-family climate, organizational commitment, and turnover: multilevel contagion effects of leaders”, Journal of Vocational Behavior, Vol. 74, pp. 18-29. Podsakoff, P.M., MacKenzie, S.B., Lee, J.-Y. and Podsakoff, N.P. (2003), “Common method biases in behavioral research: a critical review of the literature and recommended remedies”, Journal of Applied Psychology, Vol. 88 No. 5, pp. 879-903. Schlesinger, L.A. and Heskett, J.L. (1991), “The service-driven company”, Harvard Business Review, Vol. 69, pp. 71-81. Silverman, D. (2000), Doing Qualitative Research, Sage Publications, London. Sim, J. (1998), “Collecting and analysing qualitative data: issues raised by the focus group”, Journal of Advanced Nursing, Vol. 28 No. 2, pp. 345-52. Smith Travel Research (2003), Proprietary Report, Hendersonville, TN. Stalcup, L.D. and Pearson, T.A. (2001), “A model of the causes of management turnover in hotels”, Journal of Hospitality & Tourism Research, Vol. 25 No. 1, pp. 17-30. Wanuous, J.P., Poland, T.D., Premack, S.L. and Davis, K.S. (1992), “The effects of met expectations on newcomer attitudes and behaviors: a review and meta-analysis”, Journal of Applied Psychology, Vol. 77, pp. 288-97. Wasmuth, W.J. and Davis, S.W. (1983a), “Managing employee turnover”, The Cornell Hotel and Restaurant Administration Quarterly, Vol. 24 No. 1, February, pp. 15-22. Wasmuth, W.J. and Davis, S.W. (1983b), “Managing employee turnover: why employees leave”, The Cornell Hotel and Restaurant Administration Quarterly, Vol. 24 No. 2, May, pp. 11-18. Wasmuth, W.J. and Davis, S.W. (1983c), “Strategies for managing employee turnover”, The Cornell Hotel and Restaurant Administration Quarterly, Vol. 24 No. 3, August, pp. 65-75. Weaver, A. (2009), “Perceptions of job quality in the tourism industry”, International Journal of Contemporary Hospitality Management, Vol. 21 No. 5, pp. 579-93. Further reading Bret, J.M. and Stroh, L.K. (1995), “Willingness to relocate internationally”, Human Resources Management, Vol. 34, pp. 405-24. Crouter, A.C. (1984), “Spillover from family to work: the neglected side of the work-family interface”, Human Relations, Vol. 37 No. 6, pp. 425-42. Corresponding author John W. O’Neill can be contacted at: [email protected] To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Using focus groups as a tool

885

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0959-6119.htm

IJCHM 24,6

Consulting the oracle? Applications of modified Delphi technique to qualitative research in the hospitality industry

886 Received 17 August 2010 Revised 1 December 2010 7 March 2011 31 July 2011 20 September 2011 Accepted 20 September 2011

Abu Elnasr E. Sobaih Faculty of Tourism and Hotel Management, Helwan University, Cairo, Egypt, and

Caroline Ritchie and Eleri Jones Cardiff School of Management, Cardiff Metropolitan University, Cardiff, UK Abstract Purpose – The Delphi technique is used to achieve consensus among experts and/or gain judgment on complex matters. This paper aims to discuss the classical Delphi and its advantages and disadvantages in qualitative research, particularly in hospitality. Design/methodology/approach – The classical Delphi is characterized by the involvement of experts and its iterative nature. In an industry with high turnover and limited pools of specialist expertise this can lead to problems of attrition and management of the process. The paper presents two qualitative hospitality research case studies in which the classical Delphi is successfully modified to overcome its limitations. Findings – Identifying potential problems early in the research process enables critical design decisions to be made. Case one used a parallel expert group with similar experience to develop a research instrument for a limited number of prestigious experts well-acquainted with one another who might have reached specious consensus through channels not accessible to the researcher. Case two enabled the addition of new experts to an expert panel to overcome attrition in successive Delphi rounds. Practical implications – Despite its growing popularity in social science, Delphi has rarely been used in qualitative hospitality research. The modifications suggested in this paper can enhance the robustness of the classical Delphi technique for qualitative hospitality research. Originality/value – The paper shows how the classical Delphi technique can be successfully modified for use in qualitative hospitality research. Keywords Delphi method, Modified Delphi technique, Experts, Research design, Qualitative research, Hospitality management, Research methods Paper type Research paper

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 pp. 886-906 q Emerald Group Publishing Limited 0959-6119 DOI 10.1108/09596111211247227

Introduction The Delphi technique recognises expert knowledge as legitimate and useful in generating forecasts and achieving consensus (Yousuf, 2007; Hsu and Sandford, 2007; Green et al., 2007). It was introduced by the RAND corporation in the 1950s to prepare forecasts of future technological capabilities of potential interest to the military (Dalkey and Hekmer, 1951; Dalkey, 1967; 1968a,b; Brown, 1968; Brown et al., 1969; Helmer, 1972; Hsu and Sandford, 2007). Linstone and Turoff (1975) suggested that although many people categorise the Delphi technique as a forecasting procedure, it has many other applications, e.g. in empirical studies making judgements or achieving consensus on complex matters where precise information is unavailable in information systems

(Brancheau and Wetherbe, 1996; Schmidt et al., 2001); corporate management (Linstone and Turroff, 2002); environmental management (Needham and de Loe, 1990; Gkhale, 2001); marketing (Hayes, 2007); leisure planning and education (Riley, 1990; Briedenhann and Wickens, 2007); government planning (Linstone and Turroff, 2002); health research (de Meyrick, 2003); tourism research (Kaynak et al., 1994; Kaynak and Marandu, 2006) and occasionally hospitality research (Singh, 2000; Birdir and Pearson, 2000; Yao-Fen and Chen-Tsang, 2012). Although popular and widely-used, the Delphi technique has received its share of criticism. Sackman (1975), for example, criticised it as being unscientific because it is very sensitive to the mode of application and panel composition (Donohoe and Needham, 2009). However, along with other phenomenological research methods (Stierand and Do¨rfler, 2012), it can be argued that the epistemological strength of the Delphi technique is its ability to recognise its inherent subjectivity. This is particularly significant in the definition and selection of expert panellists, the exposure of pluralities and paradoxes inherent in their multiple perceptions and their rationalisation; building consensus through an interpretivist research paradigm. As Goldschmidt (1975) argued, there have been many poorly-conducted Delphi studies but the problem lies in its application rather than the method itself. Donohoe and Needham (2009) suggest that some mystification about the technique and its use leaves it open to criticism arguing that critical design decisions early in the research process enable an appropriate research design to be developed. This paper focuses upon the use of the Delphi technique in qualitative, particularly hospitality, research and modifications to both Normative and Policy Delphi techniques (Paraskevas and Saunders, 2012). It analyses the classical Delphi technique critically reviewing its characteristics and advantages/disadvantages in the context of qualitative hospitality research, identifying issues and practical implications in relation to its application. Finally, it discusses and contextualises modifications to the technique, which have been successfully used in two qualitative hospitality-based research studies, both of which were part of larger PhD projects. The cases illustrate how modification of the classical Delphi technique can overcome some of the disadvantages of the technique in qualitative hospitality research, e.g. the limited availability of experts and attrition as a result of the iterative nature of the technique. Although a less commonly-adopted approach (Paraskevas and Saunders, 2012), case one used a modified Policy Delphi technique at the beginning of the research process. Instead of a pilot study it is used to develop consensus on a research instrument for use in identifying UK wine consumer behaviour within hospitality and other public consumption contexts to address the issue of a limited number of experts. Case two involved the use of a modified Normative Delphi technique towards the end of the study to develop consensus on a practitioner toolkit for managing part-time restaurant staff. This case study directly addressed the issue of attrition. The classical Delphi technique The originators of the Delphi technique believed that it can identify best knowledge from diverse realities and is “a set of procedures for predicting and refining the opinions of a group of people . . . a group of experts or especially knowledgeable individuals” (Dalkey, 1967, p. 1). Linstone and Turoff (1975) and Skutsch and Hall (1973) described the Delphi technique as a method for gaining judgement on complex matters where precise

Consulting the oracle?

887

IJCHM 24,6

888

information is unavailable to overcome the disadvantages of single and group meetings. Adler and Ziglio (1996) pointed out that single experts may be biased and group meetings may suffer from a hegemonic “follow the leader(s)” syndrome. The Delphi technique allows experts to deal systematically, anonymously and in their own time with complex issues for which their level of knowledge is necessary (Okoli and Pawlowski, 2004; Yousuf, 2007). It can produce results which are as reliable and valid as other more scientific research methods, such as traditional surveys (Okoli and Pawlowski, 2004). Thus the objectives of most Delphi studies can be summarised into four points: (1) collecting and distilling knowledge from a group of experts; (2) achieving consensus and/or gaining judgement on complex matters where precise information is unavailable to underpin a prediction of the future; (3) the reliable and creative exploration of ideas; and (4) the production of suitable information for critical decision making. The Delphi technique has been used in both quantitative and qualitative research (Rowe and Wright, 1999; Skulmoski et al., 2007). Creswell (2007) and Skulmoski et al. (2007) suggest that it is the knowledge, experience and habitus of the panellists that makes them expert in a Delphi study and their rich explanations are particularly beneficial in qualitative research studies and enhanced by the iterative and consensual nature of the three stages of a Delphic study: (1) initial; (2) core; and (3) final (Linstone and Turoff, 1975; Hsu and Sandford, 2007; Green et al., 2007). Initial stage The initial stage of the Delphic process should achieve three key outcomes: (1) critical identification of the research question(s) and theme(s) expressed as a problem statement; (2) development of criteria to identify experts; and (3) criteria to identify the size and types of the expert sample. Each outcome will be discussed in turn. Critical identification of the research question(s) and theme(s). After clarifying the research question(s) it is important to start the Delphi process with carefully-selected themes which are normally grounded in theory. Traditionally research themes have been identified in two ways – critical review of literature or by a pilot study – especially where there is little or no existing knowledge. Custer et al. (1999, p. 51) note that themes “may be drawn from various sources including . . . synthesized reviews of the literature, and interviews with selected content experts” providing a solid foundation upon which new theory can be constructed. The research questions and the emergent themes should be piloted in the initial stage to test them for validity and reliability before presentation to experts in the core stage. Development of criteria to identify and select experts. Okoli and Pawlowski (2004) and Donohoe and Needham (2009) confirmed that a crucial requirement of any Delphi

study is the selection of appropriately-qualified experts if the results are to have critical authority. Whilst an “expert” can be defined as someone with a high level of knowledge or skill in a particular area, the literature on the Delphi technique does not provide a clear definition. Simon (1965) identified expertise as synonymous with authority and a level of experience/knowledge that differentiates an expert from a novice. Adler and Ziglio (1996) stated that Delphi participants should meet four expertise requirements: (1) knowledge/experience of the issue under investigation; (2) capacity and willingness to participate; (3) time to participate in a Delphic research process; and (4) effective communication skills. Needham and de Loe (1990) and Donohoe and Needham (2009) suggested another essential criterion; extended, continuous closeness to the issue/problem, which ensures experts bring direct in-depth knowledge to the discussion. For example, research profile and achievements may be the criteria for being considered an academic expert. In other contexts employment history, the judgement of peers and/or a particular job role (Paraskevas and Saunders, 2012) may be the basis for the criteria. The criteria for “expert” always depends on the aim and objectives of the research and is therefore subjective to some extent. An objective in any Delphi study therefore is to reduce subjectivity and increase critical authority by clearly defining “expert”. However the consensual, iterative nature of Delphi studies means that paradigm shifts may occur amongst expert panellists during the process and impact the final consensus. Therefore researchers using the Delphi technique need to be aware of its ontological subjectivity. Criteria to identify the size and type of the expert sample. There is no hard and fast rule to determine how many participants should be in a Delphi study (Adler and Ziglio, 1996) since it does not require “a statistical sample that attempts to be representative of any populations [because] it is a group decision mechanism” (Okoli and Pawlowski, 2004, p. 20). There are various opinions on expert panel size ranging from seven to 15 (Archer, 1980). Skulmoski et al. (2007) argue that participant numbers depend upon the type of sample and whether it is heterogeneous (i.e. comprising disparate groups of experts) or homogenous. Briedenhann and Butts (2006) and Skulmoski et al. (2007) suggested that smaller panels are best suited to homogenous samples and larger panels to heterogeneous samples. Briedenhann and Butts’ (2006) study of rural tourism projects identified and invited 115 experts to participate in their study and finally involved 60 experts in three sub-groups: (1) academia; (2) the public sector; and (3) consultants. Inevitably the iterative nature of a Delphic study leads to attrition (see Table I) and it is important to monitor attrition to ensure a valid consensus is achieved. Donohoe and Needham (2009) suggested beginning with a larger initial panel was necessary to overcome attrition so the numbers involved in the final iteration did not fall below what was judged to be critical mass. However, expert numbers may be limited (see case one) or there may be particular reasons for high attrition (see case two). Thus the most

Consulting the oracle?

889

IJCHM 24,6

890

Table I. Disadvantages of the Delphi technique in qualitative hospitality research

Disadvantages of the Delphi technique

Specific issues for qualitative research in hospitality

1. The Delphi technique is very sensitive to design characteristics including: level of participants’ expertise and clarity of the questions

The hospitality industry is notorious for high staff turnover; job titles are not synonymous across the industry or indicative of a particular level of expertise

2. Questionnaire management, and reporting

The technique is very sensitive to the management of questionnaires, understanding of written documents, symbolic or idiomatic meaning and cultural barriers

3. Experts’ willingness to allow their judgements to be re-formed by the opinions of others

Where the experts can identify other participants because of their small community, influence of personality may be difficult to remove

4. Experts are vulnerable to high rates of attrition

Issues important to an academic researcher may seem dull or obvious to an industry-based practitioner, particularly in small businesses where time is at a premium

5. The potential use of incentives or moral persuasion to identify experts and/or encourage sustained participation

Power issues and respondent bias resulting from the use of financial incentives or moral persuasion to reduce attrition may reduce the validity and reliability of the outcome

6. Risk of specious consensus

Participants acquiescing and conforming to dominant voices or comfortable familiar arguments may increase the risk of specious consensus (“group think”) in the Delphi forum

7. Problems in determining what actually constitutes consensus

Researchers need to be acutely aware of the nuances of the hospitality industry so as to be able to define what constitutes consensus in terms of their research

8. Multiple iterations take time to complete and can be costly in terms of researchers’ time

Achieving consensus may require several iterations each taking considerable time, which can be costly for the researcher and for participants

important criterion for a Delphi study may not be the number of participants at each stage but the knowledge and ongoing willingness to participate that the experts bring to the discussion. Core stage In the core stage documentation outlining the problem statement and the research questions and themes developed in the initial stage is circulated to the experts for comment and response to the researcher administering the Delphi study. The responses are analysed to reflect on extant theory (if any) and a summary report with a revised problem statement based on the comments is circulated to the experts if consensus has not been achieved. If some responses are extreme, in a Normative Delphi study the experts would be asked to reassess their responses in the light of the group’s responses and provide justification for their comments (Gordon, 1994; Paraskevas and Saunders, 2012). The articulation, interpretation and testing of the experts’ belief systems is the basis of the interpretivist research paradigm underpinning the Delphi technique. This process is iterated as many times as necessary to achieve consensus

because opinions are always subjective (see Figure 1). Ideas in a Delphic study are judged upon their quality and their closeness/contribution to the issue/theme under discussion. Number of iterations. Some Delphic studies have achieved consensus in one round (Hartman and Baldwin, 1995; Kuo and Yu, 1999), others have needed two (Duncan, 1995; Robinson et al., 2005), three (Brungs and Jamieson, 2005) and some have taken up to five iterations (Thomson, 1985; Kincaid, 2003). Delbecq et al. (1975) and Skulmoski et al. (2007) suggest that two or three iterations should be sufficient for most qualitative research projects. The hospitality studies reviewed in Table II generally used two or three rounds. Modes and management of interaction. Researchers and experts can interact in different ways. The traditional Delphi technique used paper-based methods with documents sent by post – a time consuming, laborious approach achieving low response rates. Other methods can be used, for example face-to-face meetings similar to semi-structured interviews with discussion of what others have said and feedback of the comments from each round for further comment. However, there may be power bias if this is not appropriately managed (see Tables I and III). As Lugosi et al. (2012)

Consulting the oracle?

891

Figure 1. The typical Delphi process

IJCHM 24,6

Delphi application

Publication

Forecasting studies Lloyd et al. (2000) Singh (2000)

892 Singh and Schmidgall (2000)

Paradigmatic

Table II. Applications of Delphi technique in hospitality research

Purpose of research Forecast changes in the Hong Kong hotel industry resulting from sovereignty changes in 1997 Forecast changes in the US lodging industry structure, performance and capital resources between 2000 and 2005 Forecast which lodging products are most likely to gain financing in the future and which financial institutions are likely to supply that capital

Yao-Fen and Chen-Tsang (2012) Identify career competencies for food and beverage managers Ko and Chiu (2006) Identify factors involved in planning new coffee shops in Taiwan Garrigos-Simon et al. (2007) Identify the importance of various capabilities and psychological factors of hotel managers to improve their perception of the competitive environment emerging in the post-internet era Murphy and Murrmann (2009) Develop a high performance management system construct for US restaurant managers Hu (2010) Develop a core competency model of innovative culinary development

point out, via modern technology (email, blogs, etc) the researcher may be able to reach a much larger pool of experts or participants (otherwise separated by geography and time zones) but still maintain control of the discussion. Table III highlights the particular importance of this in an industry as fragmented as hospitality. Final stage When consensus is achieved, that is nothing more is being added to the discussion, the Delphi process is terminated. Following this a final summary report and consensus statement should be prepared for distribution to all participants for comment. The new knowledge/theory which results can then be applied to the initial research question and the results publicly disseminated. The Delphi technique in qualitative research Both review of literature on the Delphi technique per se (Dalkey and Hekmer, 1951; Dalkey, 1967; 1968a;, 1968b; Brown, 1968; Brown et al., 1969; Rowe and Wright, 1999; Garrod and Fyall, 2005; Yousuf, 2007) and its specific application in hospitality research (Birdir and Pearson, 2000; Murphy and Murrmann, 2009; Yao-Fen and Chen-Tsang, 2012)) show that it is suitable for qualitative research in the social sciences and offers several advantages relating to its systematic nature and flexibility (see Table III). Advantages of the Delphi technique in qualitative research The Delphi technique is flexible and systematic and, in developing forward, issues relating to time zones and geography may be eliminated without compromising its

Advantages of the Delphi technique

Compatibility with qualitative research

1. It is systematic and flexible

It can be applied in diverse situations and to a wide range of complex problems. It allows experts to deal systematically with complex problems where precise information is unavailable

2. It utilises expert judgment

The process enables participants to think through ideas in their own time (important in an industry where shift work is common) before communicating them, leading to a more reflective response

3. It is an iterative process

The iterative approach allows experts to reconsider their judgements in the light of feedback from peers without the need to be in the same physical space for discussion

4. It is anonymous

The anonymity enables experts to express their opinions freely, without institutional loyalties or peer group pressures (particularly important where many of the group may know each other)

Consulting the oracle?

893

5. It eliminates the influence of personality The potential influence of dominant individuals, a problem in face-to-face group interviewing, is removed 6. It combines advantages of both single and group interviews

The process generates a record of individual and group thoughts, which can be reviewed as required

7. It produces reliable and valid results

It can produce a more valid, reliable and critical outcome than some other methods of data collection

8. It lends itself to the use of modern communications methods

The use of email, blogs, wikis and other communications methods can accelerate progress to a valid consensus and overcome temporal and geographic constraints

9. Feedback is controlled

Issues outside the debate can be controlled by the researcher

10. It is well-established and insightful

It facilitates progress via an iterative approach, which lends itself to modern communications methods

validity; for example through the use of technology to enable virtual meetings. This is important in an industry where shift working prevails. Additionally this flexibility means that experts can reflect on their ideas in their own time whilst remaining fully integrated into iterative discussions with the researcher. This fosters anonymity and encourages experts to express their opinions freely away from pressure re institutional loyalties, peer groups or dominant or “significant” personalities (such as Michelin rated chefs). This is important in an industry where the experts may know each other through working for the same organisation (e.g. the Hilton group), by being members of the same association (e.g. British Institute of Inn Keeping or Institute of Hospitality) or by being ranked in the same listing (Michelin Guide Rouge) but may be at different career points. The Delphi technique is well-established and insightful and facilitates the collection and evaluation of a depth and breadth of information essential for critical qualitative

Table III. The Delphi method and its compatibility with qualitative research

IJCHM 24,6

894

research. It combines the advantages of single and group interviews and the recording of individual and group comments enables exploration of the perspectives that inform them. A further advantage is that since many hospitality industry-related issues are widely discussed in the media (e.g. alcohol consumption and related behaviours) the researcher can control the discussion should it stray into areas peripheral to a particular debate. These advantages are summarised in Table III. Despite its applicability in wider qualitative research, the Delphi technique has had limited usage in qualitative hospitality research. Taking a pragmatic approach much hospitality industry research uses one-to-one interviews. However in this high turnover, fragmented industry it can be difficult to conduct follow-up interviews with the same/similar people in the same/similar jobs since interviewees may have moved on to new employment and no longer be appropriate or available. As Table II shows the Delphi technique has been used in a number of hospitality studies, usually for forecasting and model development, but its use is relatively new (the first studies occurring in 2000) and underutilised in comparison to tourism research where the technique has been used since the 1970s. There are, however, a number of disadvantages when moving beyond forecasting and paradigmatic studies which are discussed in the next section. Disadvantages of the Delphi technique in qualitative hospitality research The subjectivity of the Delphi technique makes it very sensitive to design characteristics, particularly the level of participants’ expertise and the clarity of the research question. Within the hospitality industry it can be difficult to identify experts for particular studies because job titles differ across the industry and may not reflect differing levels of expertise. Additionally experts identified early in the research process may well change employment and stop meeting the expert criteria for that study or access to them may cease. Furthermore in an industry with an international workforce, which values personal skills highly, good oral skills may not reflect good written skills, particularly where complex nuances are involved. This may be compounded by diverse cultural perspectives and linguistic issues. A final factor is that although hospitality employment is very fluid, knowledge of others in specialist fields often develops, for example chefs in Michelin-starred restaurants may have trained or worked together. As Paraskevas and Saunders (2012) acknowledge, although technically participants are anonymous they may be able to identify co-participants from their micro-community. Therefore removing the influence of personality is difficult particularly as the experts may use informal channels of communication, which the researcher is unaware of or has no access to (see case one); Lugosi et al. (2012) had to gain permission to enter a personal blog. Delphic studies also suffer high rates of attrition. Issues of interest to an academic researcher may seem dull or obvious to industry-based practitioners, especially if the research question is not well constructed. For hospitality researchers there are two additional factors: the industry is dominated by small businesses so time is at a premium and there is high staff turnover. This can mean that experts do not engage fully with a study even before they withdraw (see case two). To reduce attrition it is not uncommon to offer financial incentives. The UK hospitality industry is notorious for low pay, especially for operatives, making the use of financial incentives attractive to both researcher and expert. In a fragmented industry where potential participants are

known to each other, particularly if snowballing techniques are used, the potential for bias in the identification and recommendation of experts is significant. As Stierand and Do¨rfler (2012) confirm researchers need to be acutely and subjectively aware of the full range of the hospitality experience so that they can correctly identify what defines “expert” in a particular study. The social, hospitable and small community nature of the hospitality industry may also give rise to another issue – arrival at unofficial, specious, consensus through informal communication channels. Participants may believe they have identified, and so acquiesce to, dominant voices if they are not fully engaged in the process. Given all of these issues it may be difficult to determine what actually constitutes consensus and researchers need to be alert to the nuances of the hospitality industry in defining consensus in a particular study. Finally, the technique usually requires several iterations which can be time consuming and therefore costly (see case two). The disadvantages of the Delphi technique are summarised in Table I. These problems can impact on the use of the Delphi technique and the quality of outcome if not appropriately managed. Critical design decisions can address these disadvantages by consciously designing them out of a particular study. This inevitably results in modifications to the traditional Delphi technique. Examples of modifications of Delphi technique in qualitative hospitality research This section discusses modifications of the Delphi technique, which have been successfully applied in two recent qualitative hospitality PhD research projects. In case one a modified Delphi technique was used to generate initial questions for a research project in which the pool of experts was very limited. In case two a modified Delphi was used at the end of the research process to test the application of new theory to industry. Case one: developing a semi-structured interview schedule for trade wine buyers to identify UK wine consumers and their behaviour in the on-trade This case focused on identifying UK wine consumers and how they interacted with wine in a variety of hospitality/public consumption situations. Despite being the largest import market by value in the world, very little is written about UK wine consumers (Ritchie, 2006). Additionally, what little data was available was based upon dated research (Ritchie et al. 2010). The initial stage of a Delphi study must ensure that there is a real research question(s) with valid theme(s) and identify experts for the core round. The research focus was identified when the researcher was studying for the Wine and Spirit Education Trust[1] (WSET) Diploma. She was the only non-trade candidate on the course. Descriptions of UK wine consumer behaviour by early career trade wine buyers (TWBs) – those who source wine for retail outlets – did not correlate with her experience either as a wine consumer or educator. Thus a potential research question was identified but its validity needed to be tested and a research instrument developed. TWBs were identified as the most significant stakeholder group influencing wine consumption, since if wine is unavailable in bars, restaurants or other retail outlets it cannot be bought or consumed. Collecting data from TWBs became the first phase in the larger PhD study. At the time of the study in the UK just ten supermarket chains

Consulting the oracle?

895

IJCHM 24,6

896

sold approximately 70-75 per cent of the wine in the off-trade using wine-buying teams with one to four members. Many TWBs had similar career paths and qualifications – most following the WSET Diploma and/or Master of Wine (MW) route – so most of this elite community knew each other. Similarly, in the wine wholesale trade just four companies dominated selling to both the on and off-trade. Thus just fourteen significant organisations provided wine to UK consumers and it was anticipated that many potential experts would be known to each other and would interact informally. Initial stage Critical identification of the issue(s) and theme(s). In order to establish a solid basis for the core stage and test the validity of the overall research question, six focus groups were run across the UK to establish whether the author’s experience, the (very) limited literature and/or the wine trades’ perceptions of UK wine consumer behaviour were leading to misconceptions about actual behaviours. The criteria for participation in the focus groups was that there was a gender balance and participants should be over 18 years of age and consume wine (see Ritchie (2006) for full details). The results showed that, whilst there were some areas where similar behaviours were identified, there were many areas where perceptions of behaviour and attitudes differed markedly both from extant research and industry beliefs. In addition, available research about wine consumers in other counties did not appear to apply to UK wine consumers who demonstrated much more sophisticated behaviours than expected. Thus it was established that a real research question existed: Who are UK wine consumers and how do they behave? Usually the identification of experts for a Delphi study is complex and subjective. In this case the TWBs were identified as experts. The challenge was how to develop appropriate themes for semi-structured interviews with senior TWBs without contacting them. A poorly-formulated pilot interview with one or more of the major buyers could have damages the researcher’s credibility and, via informal communication channels, resulted in others being reluctant to participate thus denying access to this essential community. In addition it would have reduced the very limited number of senior TWBs available. Therefore a modified Policy Delphi technique was used to develop the semi-structured interview themes rather than developing them through a more conventional piloting process. It was considered to be Policy Delphi rather than Normative because the author was “seeking to explore possible alternatives [being] unaware of the full spectrum of issues” (Paraskevas and Saunders, 2012: p. 15). Development of criteria to identify and select the Delphic experts. The consumer focus groups had identified various themes for investigation but the researcher was aware that if the themes were not appropriately presented to the TWBs they were unlikely to participate. Finding experts with similar levels of knowledge, sharing the same linguistic and cultural traditions but not part of the TWB community was essential. An MW employed as the UK representative on the European Union Wine Board of Trade and the Education Officer for the Institute of Masters of Wine were identified. Although both were highly qualified and involved in the wine trade neither was employed as a TWB. Additionally both were interested in expanding understanding of UK wine consumer behaviour. However the perspective of only two experts was not considered sufficient to provide the critical mass needed to identify and validate the themes therefore other experts were sought.

Despite the limited pool of senior TWBs in the UK, a large number of small business owner-operators sourced their own wine. Approaching this group would not impact on the expert pool and had the added advantage of being much closer to wine consumers – their direct customers – than those who bought for the national and international chains. However, the researcher was aware that there would be temporal and physical issues associated with interacting with the owner-operators to be overcome. Opportunistic access was made by joining a week-long wine study tour to Italy with 16 members of the wine trade. The research project was explained prior to the trip and potential participants invited to comment on the themes during the tour. Twelve small business TWBs agreed to participate. The final panel of experts for the Delphi study was therefore a heterogenous group comprising small business TWBs and MWs. The MWs contributed a depth of professional knowledge and language that the small business TWBs could not; the small business TWBs, contributed a depth of knowledge about UK wine consumer behaviour from observations of their customers, which the MWs could not. Using both expert types ensured the multi-layered perspective needed for a valid consensus. Core stage Having completed the initial stage the study moved into the core stage – the consultation rounds. Number of iterations. There were three iterations. In the first iteration the MW and the Education Officer were sent a list of the themes identified by the consumer focus groups, asked to comment on them and add in any points that they felt were missing. Analysis of their comments resulted in modification of original themes by the addition of two new themes. In the second iteration the revised list of themes was issued to the twelve small business TWB participants. A variety of single and group discussions took place during the tour about the research. By the end of the week all the lists were returned with comments. No new themes were identified but these comments added depth to several themes. In the final iteration the modified themes were returned to the MW and Education Officer for further comment. They had nothing more to add and consensus was reached. Final stage Therefore the final stage of a Delphic study was reached and the themes were developed into a consensus statement – the semi-structured interview schedule for 14 senior TWBs. The modifications to the Delphi technique used to develop the research instrument are summarised in Figure 2. The process was judged successful in replacing a conventional piloting process on two counts: first, it enabled data collection from the TWBs whose companies supplied over 70 per cent of wine sold in the UK at that time; second, although asked, none of the senior TWBs questioned the validity of the themes discussed in their interviews or identified any missing ones. As discussion of this case shows there was one other modification of the conventional Delphi technique relating to maintaining anonymity. The researcher knew that there was only one professional qualification route (WSET perhaps followed by MW), only one major wine trade fair in the UK annually and that most of the small business TWBs would either have worked for, and or bought from, the major wine

Consulting the oracle?

897

IJCHM 24,6

898

Figure 2. A summary of modified Delphi technique for case one

wholesalers. Therefore it was accepted from the start that the study (and its participants) would become known via many informal communication channels. Whilst this had the potential for dominant personalities to influence the final consensus accepting that anonymity was not possible also enabled the researcher to take advantage of the small business TWB expertise on the study tour to Italy. The potential for power bias was addressed by intentionally separating the more prestigious MW experts from the small business TWBs in the consultation rounds. However during the study tour it was discovered that one of the participants was an MW student. This gave more weight to his comments at the beginning of the week but familiarity had tempered this by the end. The identification of the MW student on the study tour confirmed the researchers’ original view that anonymity was not possible. Had it been attempted it is likely that discussions unknown to the researcher would have taken place resulting in less valid and reliable, potentially specious, themes being identified. Case two: testing a toolkit to assist with the management of part-time employees in UK restaurants Early research had shown that part-time restaurant employees were often managed inappropriately due to a lack of understanding by managers/employers of their part-time employees’ needs (Sobaih et al., 2008; 2011; Sobaih, 2011). The specific issue was that part-time staff, although not a homogeneous group, were often treated as such by their managers/employers (Sobaih et al., 2011) and there was a need for industry-friendly best practice advice to enable restaurant owners/managers to better interact with their part-time employees to enhance the customer experience. To develop an effective toolkit it was necessary to identify a method whereby consensus could be achieved. As highlighted in Table I there are many linguistic, cultural and organisational differences within the UK restaurant industry. Any toolkit would need to address these issues and achieve symbolic and cultural consensus if it was to have resonance across the industry therefore a Delphic study was considered the most appropriate. Since, unlike case one, this part of the PhD study was to develop findings for industry practitioners use not to generate new knowledge, it was decided to use and modify the Normative Delphi process most commonly used in qualitative hospitality research. The initial stage Critical identification of research question(s) or issue(s). The traditional Delphi technique starts by identifying the research issue/question(s). The first modification made to the technique was to begin the process with themes identified through research into the management of part-time employees in the UK restaurant sector rather than developing new themes. During the initial stage a draft of the toolkit was piloted with two practitioner experts to ensure the critical authority of the study. The experts to be used in the core stage had limited time and responded only to requests they saw as having professional credibility and or were of personal benefit (see Table I). Agreeing with Custer et al. (1999), the piloting had two advantages: it confirmed the validity of the research question – how to turn academic theory into an industry-friendly tool and produce a best practice toolkit for managing part-time employees in the UK restaurant sector? It also undoubtedly improved the initial response rate in the core stage.

Consulting the oracle?

899

IJCHM 24,6

900

Development of criteria to identify and select experts. The next step was to recruit the expert panel (see Figure 1). Unlike case one where the pool of experts was very limited the critical issue for this study was identifying a sufficiently wide range of participants to make the toolkit useful in a range of restaurant environments. To reap the full benefit of the Delphi technique whilst overcoming its shortcomings, clear criteria for selecting the expert panel were developed. The experts were identified based upon their knowledge, experience, closeness, capacity and commitment to the research. All the experts had at least 15 years experience of, and were employed in, managing the human element within the restaurant sector. Having validated the questions and identified the experts the study moved into the core stage. The core stage Number of iterations. In the first round of the core stage the draft toolkit and a list of questions were circulated to six experts. During the second round two experts withdrew from the study because they shifted employment – a reflection of the high turnover that characterizes the hospitality industry. Since participation had fallen, a second modification was to add more experts to each successive round reflecting a variant on Donohoe and Needham’s (2009) funnel approach. As shown in Table IV, the new participants ensured that critical mass was maintained and, also reflecting the diversity of the industry, brought new perspectives to each round. However this modification meant that more time and effort needed to be invested in the newly-recruited experts in each round. Mode of interaction. An appointment was made with the selected experts and the research aim was discussed. A copy of the toolkit was left with the experts to read and, wherever possible, to use with staff in their restaurants for a period of one month (sometimes it took two months). After this time the experts were asked to give their feedback on the toolkit through in-depth, face-to-face, semi-structured interviews. As Table IV shows three rounds were used to achieve consensus amongst the experts about the toolkit with no new comments being added. The final stage The final stage of the Delphic process therefore began. A summary report and a print-ready copy of the toolkit were produced and circulated to participants for final comment. A summary of case two is shown in Figure 3. Participants

Table IV. Summary of the rounds and participants in the modified Delphi study for case two

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Expert1 Expert2 Expert3 Expert4 Expert5 Expert6 Expert7 Expert8 Expert9 Expert10 Expert11 Expert12

First round

Second round

Third round

U U U U U U

U U U

U

U U U U U

U U U

U U U

Consulting the oracle?

901

Figure 3. A summary of modified Delphi technique in case two

Conclusion Like Paraskevas and Saunders (2012) this paper started from the premise that the Delphi technique is much underused in hospitality research because it is often seen as too complex for, or irrelevant to, qualitative research within the hospitality industry. The paper began by summarising the advantages of the Delphi technique to qualitative research generically (Table III) and applications of Delphi technique in hospitality research specifically (Table II). They show that this premise is incorrect although Table I acknowledges the disadvantages of the technique which must always be taken into account.

IJCHM 24,6

902

The paper then focused on modifications to the Delphi technique, which address the potential shortcomings identified. Case one involved a modified Policy Delphi process being used to develop a research instrument (the semi-structured interview schedule) away from the proposed interview sample. The modifications helped to ensure the credibility of the researcher in the interviews to minimise the likelihood of access to an entire expert community being withdrawn and, since the pool of potential experts was extremely limited, avoiding conventional piloting which would have reduced the expert pool below a critical volume. Case two used a modified Normative Delphi technique to turn academic research into a user-friendly tool. Thus it bridged the gap between academia and industry to develop a community of practice and addressed a criticism often levelled at academic research – that is that it is of little relevance to operational practice. As these cases highlight, it is the multi-paradigmatic but consensual nature of the Delphi technique and particularly the concept of ‘expert’ which is both its strength and weakness. Definitions of expert are inevitably subjective and great care must be taken in constructing the expert criteria in any particular research context. Stierand and Do¨rfler (2012) suggest that taking a phenomenological approach as these cases did, which utilises lived experiences or qualia, can assist significantly in this process and so enhance the richness of the data generated. For example, a hotel manager may manage a 600-room or six-room hotel – in one instance perhaps as manager, in the other as owner – one is potentially on the lookout for better employment, the other is likely to stay in place. Which then is the expert and in relation to what? Since the authors of the cases had worked for many years in the hospitality industry they understood both the nuances of job titles of their experts and how to classify the expertise, which lay behind them. Additionally they understood the fluid and international nature of hospitality employment with its small, intimate communities of people with the same specialism (chefs, sommeliers, risk managers, etc). They knew that experts may be aware of, or have worked with, other experts in their field even though they currently work in different environments and/or live in different countries. Potentially these issues can have major implications for a Delphic study, however by pragmatically modifying the technique as in case one, via a face-to-face forum, they can be addressed without invalidating the process. These results suggest that it is unlikely that a classical Delphic study could be carried out within the hospitality industry because of the difficulty of defining an expert, ensuring anonymity, controlling the discussion and retaining experts’ interest through successive iterations. However, as the two case studies discussed in this paper have shown, this does not necessarily invalidate the Delphi technique as a research method for the hospitality industry. Instead it reinforces the point that a hospitality researcher must fully understand both the layered multi-perspective nature of the Delphi technique and the fluid interactive nature of the hospitality industry before they are able to use it effectively. Furthermore researchers proposing to use the Delphi technique - Policy, Normative or modified – must be able to construct a valid research question and a critical definition of what constitutes an expert in terms of their study before data collection begins. Lugosi et al.’s (2012) work suggests that other modifications of the Delphi technique to exploit digital communication channels could enable participation by a wider pool of experts. This could be of great significance to researchers in an industry predicated upon shift, weekend and night-time working, internationally across time zones.

By including these particular cases – one where the modification was used early in the study and another where it was used at the end – this paper demonstrates that not only can the Delphi technique be used effectively in qualitative hospitality research but it can be applied as an element in the research process rather than the complete methodology. Modifying the Delphi technique does not reduce its effectiveness providing sufficient consideration is given to the design of each stage. Thus a modified Delphi technique can be a useful and effective research tool and should be included in the hospitality researchers’ portfolio. Note 1. WSET is the professional and lead educational body for the wine trade. Obtaining their highest qualification, the WSET Diploma, enables graduates to apply to the Institute of Masters of Wine to study to become a Master of Wine, internationally recognised as the highest professional qualification within the wine trade. References Adler, M. and Ziglio, E. (1996), Gazing into the Oracle: The Delphi Method and its Application to Social Policy and Public Health, Jessica Kingsley Publishers, London. Archer, B.H. (1980), “Forecasting demand: quantitative and intuitive techniques”, International Journal of Tourism Management, Vol. 1 No. 1, pp. 5-12. Birdir, K. and Pearson, T.E. (2000), “Research chefs’ competencies: a Delphi approach”, International Journal of Contemporary Hospitality Management, Vol. 12 No. 3, pp. 205-9. Brancheau, J.C. and Wetherbe, J.C. (1996), “Key issues in information system management”, MIS Quarterly, Vol. 20 No. 2, pp. 225-42. Briedenhann, J. and Butts, S. (2006), “Challenges and shortcomings: use of the Delphi technique in rural tourism project evaluation”, Current Issues in Tourism, Vol. 9 No. 2, pp. 171-90. Briedenhann, J. and Wickens, E. (2007), “Developing cultural tourism in South Africa: potential and pitfalls”, in Richards, G. (Ed.), Cultural Tourism: Global and Local Perspective, The Haworth Hospitality Press, Binghamton, NY, pp. 69-90. Brown, B. (1968), Delphi Process: A Methodology Used for the Elicitation of Opinions of Experts, The RAND Corporation, Santa Monica, CA. Brown, B., Cochran, S. and Dalkey, N. (1969), Delphi Method II: Structure of Experiments, The RAND Corporation, Santa Monica, CA. Brungs, A. and Jamieson, R. (2005), “Identification of legal issues for computer forensics”, Information Systems Management, Vol. 22 No. 2, pp. 57-66. Creswell, J. (2007), Qualitative Inquiry and Research Design: Choosing among Five Approaches, Sage Publications, Thousand Oaks, CA. Custer, R.L., Scarcella, J.A. and Stewart, B.R. (1999), “The modified Delphi technique: a rotational modification”, Journal of Vocational and Technical Education, Vol. 15 No. 2, pp. 1-10. Dalkey, N.C. (1967), Delphi, The RAND Corporation, Santa Monica, CA. Dalkey, N.C. (1968a), Predicting the Future, The RAND Corporation, Santa Monica, CA. Dalkey, N.C. (1968b), Experiments in Group Prediction, The RAND Corporation, Santa Monica, CA. Dalkey, N.C. and Hekmer, O. (1951), The Use of Experts for the Estimation of Bombing Requirements, RM-727-PR, November.

Consulting the oracle?

903

IJCHM 24,6

904

de Meyrick, J. (2003), “The Delphi method and health research”, Health Education, Vol. 103 No. 1, pp. 7-16. Delbecq, A.L., Van de Ven, A.H. and Gustafson, D.H. (1975), Group Techniques for Program Planning, Scott-Foresman and Co., Glenview, IL. Donohoe, H.M. and Needham, R.D. (2009), “Moving best practice forward: Delphi characteristics, advantages, potential problems and solutions”, International Journal of Tourism Research, Vol. 11 No. 5, pp. 415-38. Duncan, N.B. (1995), “Capturing flexibility of information technology infrastructure: a study of resource characteristics and their measure”, Journal of Management Information Systems, Vol. 12 No. 2, pp. 37-58. Garrigos-Simon, F.J., Palacios-Marque, P. and Narangajavana, Y. (2007), “Improving the perception of hotel managers”, Annals of Tourism Research, Vol. 35 No. 2, pp. 359-80. Garrod, B. and Fyall, A. (2005), “Revisiting Delphi: the Delphi technique in tourism research”, in Ritchie, B.W., Burns, P. and Palmer, P. (Eds), Tourism Research Methods: Integrating Theory and Practice, CAB International, Wallingford, pp. 85-98. Gkhale, A.A. (2001), “Environmental initiative prioritization with a Delphi approach: a case study”, Environmental Management, Vol. 28 No. 2, pp. 187-93. Goldschmidt, P.G. (1975), “Scientific inquiry or political critique? Remarks on Delphi assessment, expert opinion, forecasting, and group process by H. Sackman”, Technological Forecasting and Social Change, Vol. 7 No. 2, pp. 195-213. Gordon, T.J. (1994), “The Delphi methods”, AC/UNU Millennium Project, Future Research Methodology, available at: www.gerenciamento.ufba.br/Downloads/delphi%(1).pdf (accessed 11 August 2010). Green, K.C., Armstrong, J.S. and Graefe, A. (2007), “Methods to elicit forecasts from groups: Delphi and prediction markets compared”, Foresight: The International Journal of Applied Forecasting, Vol. 8, pp. 17-20. Hartman, F. and Baldwin, A. (1995), “Using technology to improve the Delphi method”, Journal of Computing in Civil Engineering, Vol. 9 No. 4, pp. 244-9. Hayes, T. (2007), “Delphi study of the future of marketing of higher education”, Journal of Business Research, Vol. 60 No. 9, pp. 927-31. Helmer, O. (1972), On the Future State of the Union, Institute for the Future, Menlo Park, CA. Hsu, C. and Sandford, B. (2007), “The Delphi technique: making sense of consensus”, Practical Assessment Research & Evaluation, Vol. 12 No. 10, pp. 1-8. Hu, M. (2010), “Developing a core competency model of innovative culinary development”, International Journal of Hospitality Management, Vol. 29 No. 4, pp. 582-90. Kaynak, E. and Marandu, E.E. (2006), “Tourism market potential analysis in Botswana: a Delphi study”, Journal of Travel Research, Vol. 45 No. 2, pp. 227-37. Kaynak, E., Bloom, J. and Leibold, M. (1994), “Using the Delphi technique to predict future tourism potential”, Marketing Intelligence & Planning, Vol. 12 No. 7, pp. 18-29. Kincaid, S.O. (2003), “Web-based courses in human services: a comparison of student and faculty perceptions of factors that facilitate or hinder learning”, Digital Abstracts International, Vol. 64 Nos 07, 2403 (UMI No. 3098041). Ko, W. and Chiu, C.P. (2006), “A new coffee shop location planning for customer satisfaction in Taiwan”, International Journal of the Information Systems for Logistics and Management, Vol. 2 No. 1, pp. 55-62.

Kuo, N.-W. and Yu, Y.-H. (1999), “An evaluation system for national park selection in Taiwan”, Journal of Environmental Planning and Management, Vol. 42 No. 5, pp. 735-43. Linstone, H.A. and Turoff, M. (1975), The Delphi Method: Techniques and Applications, Addison-Wesley Publishing Company, Reading, MA. Linstone, H.A. and Turroff, M. (2002), “General applications: introduction”, in Linstone, H.A. and Turroff, M. (Eds), The Delphi Approach: Technique and Applications, Addison-Wesley Publishing, Reading, MA. Lloyd, J., La Lopa, J.M. and Braunlich, C.G. (2000), “Predicting change in Hong Kong’s hotel industry given the change in sovereignty from Britain to China in 1997”, Journal of Travel Research, Vol. 38 No. 4, pp. 405-10. Lugosi, P., Janta, H. and Watson, P. (2012), “Investigative management and consumer research on the internet”, International Journal of Contemporary Hospitality Management, Vol. 24 No. 6, pp. 838-54. Murphy, K.S. and Murrmann, S. (2009), “The research design used to develop a high performance management system construct for US restaurant managers”, International Journal of Hospitality Management, Vol. 28 No. 4, pp. 547-55. Needham, R.D. and de Loe, R. (1990), “The policy Delphi: purpose, structure, and application”, The Canadian Geographer, Vol. 34 No. 2, pp. 133-42. Okoli, C. and Pawlowski, S.D. (2004), “The Delphi method as a research tool: an example, design considerations and applications”, Information & Management, Vol. 42 No. 1, pp. 15-29. Paraskevas, A. and Saunders, M. (2012), “Beyond consensus: an alternative use of Delphi enquiry in hospitality research”, International Journal of Contemporary Hospitality Management, Vol. 24 No. 6, pp. 907-24. Riley, R.G. (1990), An Exploratory Study of Quality Assurance Methodology in Therapeutic Recreation Using the Delphi Technique, Microform Publications, College of Human Development and Performance, University of Oregon, Eugene, OR. Ritchie, C. (2006), “Identifying the UK wine consumer”, PhD thesis, University of Wales Institute, Cardiff. Ritchie, C., Elliott, G. and Flynn, M. (2010), “Buying wine on promotion is trading-up in UK supermarkets: a case study in Wales and Northern Ireland”, International Journal of Wine Business Research, Vol. 22 No. 2, pp. 102-10. Robinson, Q.M., Collins, C.J. and Oreg, S. (2005), “The effects of recruitment message specificity on applicant attraction to organizations”, Journal of Business & Psychology, Vol. 19 No. 3, pp. 319-40. Rowe, G. and Wright, G. (1999), “The Delphi technique as a forecasting tool: issues and analysis”, International Journal of Forecasting, Vol. 15 No. 4, pp. 353-75. Sackman, H. (1975), Delphi Assessment: Expert Opinion, Forecasting, and Group Process, R-1283-PR, The Rand Corporation, Santa Monica, CA, April. Schmidt, R.C., Lyytinen, K., Keil, M. and Cule, P. (2001), “Identifying software project risks: an international Delphi study”, Journal of Management Information System, Vol. 17 No. 4, pp. 5-36. Simon, H.A. (1965), The Shape of Automation, Harper and Row, New York, NY. Singh, A.J. (2000), “The US lodging industry in the new millennium – a Delphi study to predict changes in the lodging industry structure, performance and capital sources in years 2000 and 2005”, Journal of Hospitality Financial Management, Vol. 8 No. 1, pp. 1-22. Singh, A.J. and Schmidgall, R.S. (2000), “Financing lodging properties”, Cornell Hotel and Restaurant Administration Quarterly, Vol. 41 No. 4, pp. 39-47.

Consulting the oracle?

905

IJCHM 24,6

906

Skulmoski, G.J., Hartman, F.T. and Krahn, J. (2007), “The Delphi method for graduate research”, Journal of Information Technology Education, Vol. 6, pp. 1-21. Skutsch, M. and Hall, D. (1973), Delphi [Microform]: Potential Uses in Educational Planning, Project Simu-School, Chicago, IL. Sobaih, A.E. (2011), The Management of Part-time Employees in the Restaurant Industry: A Case Study of the UK Restaurant Sector, Lambert Academic Publishing GmbH & Co., London. Sobaih, A.E., Coleman, P., Ritchie, C. and Jones, E. (2008), “Part-time employees and service quality in the UK restaurant sector”, The Hospitality Review, Vol. 10 No. 4, pp. 63-9. Sobaih, A.E., Coleman, P., Ritchie, C. and Jones, E. (2011), “Part-time restaurant employee perceptions of management practices: an empirical investigation”, The Service Industries Journal, Vol. 31 No. 11, pp. 1749-68. Stierand, M.B. and Do¨rfler, V. (2012), “Reflecting on a phenomenological study of creativity and innovation in haute cuisine”, International Journal of Contemporary Hospitality Management, Vol. 24 No. 6, pp. 946-57. Thomson, B.R. (1985), “Appropriate and inappropriate uses of humor in psychotherapy as perceived by certified reality therapists: a Delphi study (Delphi method), (UMI No. 8606095)”, Digital Abstracts International, Vol. 47 No. 1, p. 90. Yao-Fen, W. and Chen-Tsang, (S.)T. (2012), “Analysis of career competency of food and beverage managers in international tourist hotels in Taiwan”, International Journal of Hospitality Management, Vol. 31 No. 2, pp. 612-6. Yousuf, M.I. (2007), “Using experts’ opinions through Delphi technique”, Practical Assessment Research & Evaluation, Vol. 12 No. 4, pp. 1-8. Corresponding author Abu Elnasr E. Sobaih can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0959-6119.htm

Beyond consensus: an alternative use of Delphi enquiry in hospitality research Alexandros Paraskevas Department of Hospitality, Leisure & Tourism Management, Oxford Brookes University, Oxford, UK, and

Mark N.K. Saunders

Beyond consensus

907 Received 18 January 2011 Revised 8 August 2011 28 November 2011 Accepted 21 December 2011

School of Management, University of Surrey, Guildford, UK Abstract Purpose – This paper’s aim is to critically review the use of Delphi techniques in qualitative research for utilising “expert” opinions and to explore through a detailed example how Policy Delphi can be used by hospitality researchers as an alternative to the more widely used Normative Delphi. Design/methodology/approach – The paper reflects on the research methodology of a project that explored organisational crisis signals detection using Policy Delphi with a criterion sample comprising 16 senior hotel executives involved in crisis management. Findings – The main methodological concerns regarding Delphi are the definition of consensus, the expertise of the panel, its lack of scientific rigour, and – due to its lack of uniformity – reliability and validity of findings. Policy Delphi by default addresses the first since it does not seek consensus and can, through its design and execution, address the remaining concerns. Research limitations/implications – Carefully designed Policy Delphi can offer a powerful research tool for exploratory research in hospitality, particularly for development of policies and strategies within an organisation. Unlike Normative Delphi, it is not intended as a decision making tool, but rather as a tool to generate options and suggest alternative courses of action for consideration. Originality/value – The paper presents a valuable research tool that has evaded the attention of many hospitality researchers offering an illustrative example of its use in exploratory research to deliver credible, transferable and confirmable findings. Keywords Consensus building, Crisis management, Delphi method, Exploratory research, Expert panel, Critical incident technique, Methods of enquiry, Hospitality management Paper type Research paper

1. Introduction A large part of phenomenological research in hospitality and tourism has relied upon elicitation and subsequent analysis of expert-participant opinion from individuals who were part of the phenomenon under study and had their own unique experiences and deep understandings of the issues of concern. In such cases, researchers construct knowledge through the collection of multiple sets of such interpretations, the combination of individual experts’ experiences and understandings enabling common themes to be generated and insight gained (Becker, 1992). Central to this is the inductive representation of participants’ “lived experiences” (see Stierand and Do¨rfler, 2012 in this Special Issue) from which the researcher can build up a set of factual propositions, explanations and theories (Saunders et al., 2012). Invariably, emphasis is placed in the “process of interpretation” by which these outcomes are developed

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 pp. 907-924 q Emerald Group Publishing Limited 0959-6119 DOI 10.1108/09596111211247236

IJCHM 24,6

908

(Bogdan and Taylor, 1975, p. 14) as well as in the associated “audit trail” (Gill and Johnson, 2010, p. 172) to ensure possible concerns of subjectivity and bias are addressed. In interpreting such data, researchers therefore need to be aware of and make clear their own impact that of the research process on data interpretation. This raises challenges for qualitative researchers utilizing expert opinion. Alongside the need to build up factual propositions, explanations and theories, they must demonstrate the credibility, transferability and confirmability of their findings (Lincoln and Guba, 1985), providing sufficient information to enable the impact of method on findings to be assessed. Such challenges can, for some research projects, be addressed through the use of research designs incorporating Normative or Policy Delphi techniques through which expert participants are involved in the data co-creation and interpretation. Delphi techniques are argued to offer two key benefits: first, the data collected are the opinions or judgments of expert-participants, however defined, providing a rich base for analysis and discussion (Delbecq et al., 1986; Hsu and Sandford, 2007; Sackman, 1975); and second, the validity of findings can be evaluated in terms of either sufficient or widespread agreement among these participants or, in participants recognising that there is a significant spread of opinions (Dalkey and Helmer, 1963). Of these, Normative Delphi (also referred to as Conventional or Classical Delphi) is the most widely used and employed normally for building consensus on a specific real world issue (Hsu and Sandford, 2007). Defined as a means for structuring group communication processes to address complex problems (Linstone and Turoff, 2002), Normative Delphi provides a way of integrating anonymous experts’ knowledge and by making use of their diverse and subjective judgements, opinions and experiences to achieve consensus. In contrast, Policy Delphi uses the experts’ judgements, opinions and experiences to identify spread of opinion. However, despite its widespread real world applicability in areas such as strategy determination, resource utilisation and programme planning, the use of Policy Delphi within business management and, in particular hospitality and tourism research, has been minimal. This is surprising since Normative Delphi has received considerable methodological criticism (e.g. Gupta and Clarke, 1996; Sackman, 1975, etc) and presents a number of challenges (see Sobaih et al., 2012, in this Special Issue) that lead researchers to modify their research designs. In contrast whereas Policy Delphi appears to overcome several of these challenges, offering more flexibility. The aim of this paper is to provide an insight into Policy Delphi and illustration of how it can be used by qualitative researchers as an alternative technique for researching complex subjects. It commences with an overview of the Normative and Policy Delphi techniques. This is followed by an example in which the latter is applied in a hospitality context. It concludes with a discussion of the benefits that Policy Delphi can offer to hospitality and tourism researchers. 2. Normative Delphi Normative Delphi is a method for eliciting, aggregating and obtaining a consensus of expert opinion (Linstone and Turoff, 2002). Designed as a group decision mechanism, it utilizes a panel of qualified experts who have deep understanding of the issues of concern. These experts are either top management decision makers, professional staff members together with their support team or other respondents whose judgments are being sought (Delbecq et al., 1986). There is no recommended number of expert

participants for the panel, this varying according to the scope of the problem and the available resources with studies recording panels ranging from ten to 1,685 participants (e.g. Donohoe and Needham, 2008; Reid, 1988). Rather an essential characteristic of Normative Delphi is that expert-participants are anonymous throughout the study thereby eliminating possible influences by more prominent members of the group and providing each participant a non-threatening format (Moore, 1987). Normative Delphi is conducted typically in three to five rounds (a detailed discussion on Normative Delphi’s design is provided by Sobaih et al., 2012, in this Special Issue). In the first round, a questionnaire with open questions is sent to each participating expert. Their responses are analysed to generate a series of statements, which are incorporated into another questionnaire and, subsequently, sent to each expert asking them to rank their agreement with each. Rankings from this second round are summarised and included in a repeat version of the questionnaire which is sent back to the same experts who can now see how other experts (who remain anonymous throughout the process) ranked statements and may decide to revise their own ranking. The process is repeated several times to develop a degree of consensus through iterative and controlled feedback (Delbecq et al., 1986). Although theoretically the Normative Delphi process can continue through as many subsequent rounds as needed to achieve consensus, three rounds are usually sufficient (Custer et al., 1999). Within hospitality and tourism, Normative Delphi has, to date, been used predominantly to forecast tourism demand (e.g. Cunliffe, 2002; Kaynak and Macaulay, 1984; Singh, 2000; Yong et al., 1988) and, to a lesser extent, generate ideas, theories and practice related to hospitality and tourism management issues. For example, Garrod and Fyall (2000) explored constraints and imperatives relating to the long-term management of built heritage attractions; Miller (2001) used a two-round Delphi to develop indicators measuring movements in the tourism product’s sustainability at a company/resort level; Spenceley (2005) identified environmental factors that were essential to, or incompatible with, sustainable nature-based tourism in South Africa; Garrigos-Simon et al. (2007) identified the importance of various capabilities and psycho-sociological factors of hotel managers, in order to improve their perception of the competitive environment emerging in the post-internet era; and Murphy and Olsen (2009) developed set of work practice dimensions in a high performance management system in the US casual restaurant sector. In not requiring selected participants to meet face-to-face, Normative Delphi overcomes difficulties and costs associated with bringing participants together at a single time and place whilst helping to maintain anonymity. Apart from such practical advantages, Linstone and Turoff (2002) maintain that Normative Delphi also provides participants with greater freedom of individuality and expression. Face-to-face interactions can present difficulties if dominant individuals influence the discussions’ outcome, focussing the exchange of ideas on individual and group interests rather than on problem solving (Dalkey and Helmer, 1963). In such situations group pressure for conformity can potentially discourage minority views to be aired resulting in the phenomenon of “groupthink”. Normative Delphi has received considerable methodological criticism (e.g. Gupta and Clarke, 1996). Sackman (1975, p. 74) dismissed Normative Delphi for lack of scientific rigour and psychometric validity concluding that “the massive liabilities of Delphi, in principle and in practice, outweigh its highly doubtful assets”. However, Normative Delphi advocates argue it should not be used as a substitute for other

Beyond consensus

909

IJCHM 24,6

910

scientific testing but as an alternative for complex subjects that other methods cannot resolve (Hsu and Sandford, 2007). According to Linstone (2002, p. 559) the technique “is of value not in the search for public knowledge but in the search for public wisdom; not in the search for individual data but in the search for deliberative judgment”. One central question in the use of Normative Delphi is the percentage of agreement a researcher should accept as synonymous with consensus. Some researchers set out very explicit percentage cut-off points for consensus at the beginning of their study (e.g. Loughlin and Moore, 1979 suggest that consensus should be equated to 51 per cent, whereas Green et al., 1999 use an 80 per cent consensus level). Others use an arbitrary cut off simply to limit the number of items considered. Rowe and Wright (1999) propose an alternative measure, the “post-group consensus”, which refers to the extent panellist experts individually agree with the final group aggregate; their own final round rating and those of other participants. A further question is whether the reduction in variance of opinions over rounds reflects true consensus (that is a reasoned acceptance of a position by the panellist experts) or if this is “false” consensus resulting from a drop out by discouraged dissenters whose disagreements were ignored (Linstone and Turoff, 2002; Sackman, 1975). With the concept of consensus being debated, Delphi technique advocates developed an alternative approach which did not necessarily seek participants’ consensus. 3. Policy Delphi Policy Delphi was introduced by Turoff (1970) as a means of establishing “all the differing positions advocated and the principal pro and con arguments for those positions” (p. 153) rather than the consensus of unbiased experts, and as a means of analysis instead of a tool for decision-making or prognostication (Turoff, 1970). Strauss and Zeigler (1975) do not overrule the possibility of consensus stating that the purpose of the Policy Delphi is to identify a variety of alternatives to a policy issue and provide a constructive forum in which consensus might occur. However, Turoff (1970) suggests that this version is more appropriate for those not interested in the panel of experts generating a solution but in having an informed group present all the options available with supporting evidence for their consideration. Policy Delphi is a multistage process, the number of stages ranging between two and five (Critcher and Gladstone, 1998). Turoff (1970) suggests that the process normally requires at least “five rounds as opposed to the two or three that are usually sufficient for the technological type [Normative] Delphi” (p. 161), recognising that, as with the Normative Delphi, multiple iterations can tire participants and lead to attrition (Table I). In practice, Policy Delphi can usually be carried out effectively in three rounds by allowing the participants to add their own options and positions on the topic and their underlying assumptions or supporting arguments from the first round (Turoff, 2002). In the first round of Policy Delphi the researcher spends a significant amount of time identifying and recording issues and options derived from these issues according to the participants. McKenna (1994) maintains that, in this stage, face-to-face interviews greatly increase participation and investment in the project, especially with participants in leadership positions. This round is followed by data analysis, distribution of findings to all participants and design of a new research instrument (normally a questionnaire) based on the responses to the previous questions for a

Policy Delphi stage 1 Formulation of the issues. What is the issue that really should be under consideration? How should it be stated?

Delphi round

Beyond consensus

1

2 Exposing the options. Given the issue, what are the policy options available?

911

3 Determining initial positions on the issues. Which are the ones everyone already agrees upon and which are the unimportant ones to be discarded? Which are the ones exhibiting disagreement among the respondents? 4 Exploring and obtaining the reasons for disagreements. What underlying assumptions, views, or facts are being used by the individuals to support their respective positions?

2

5 Evaluating the underlying reasons. How does the group view the separate arguments used to defend various positions and how do they compare to one another on a relative basis? 6 Re-evaluating the options. Re-evaluation is based upon the views of the underlying “evidence” and the assessment of its relevance to each position taken

3þ

Source: Adapted from Turoff (2002, p. 84)

second measurement of opinions. Participants are also encouraged to add their own options and are asked for their positions on every option along with their reasons or supporting arguments. This process allows participants to reconsider their opinions in light of the views of other stakeholders and can be repeated until either consensus is reached or saturation of opinion occurs (Critcher and Gladstone, 1998). Policy Delphi has been used in public administration and business management research to examine, among other issues, consumer problems in a marketing system (Wikstrom, 1978), city planning and development in the Middle East (Smit and Mason, 1990), public health issues (Rainhorn et al., 1994; O’Loughlin and Kelly, 2004), public administration policies (Buck et al., 1993; Critcher and Gladstone, 1998), taxation issues (Evans and Walpole, 1999) and organisational strategic planning (Akkermans et al., 1999; Mara, 2000). Such studies showed the technique allows researchers to overcome most of the methodological criticisms (Sobaih et al., 2012) of Normative Delphi. However, it appears that this method has not been adequately adopted by hospitality and tourism qualitative researchers since the authors have been unable to find examples of its use in hospitality research other than one presented in this paper. 4. Application of Policy Delphi in hospitality research: an example Research background and aim Organisational crises are seen more as a norm rather than an exception in the contemporary business environment (Fink, 1986). Although several organisational crises are considered “sudden”, expert studies have demonstrated that over 60 per cent of them normally emit warning signals early enough to be prevented (Institute of Crisis Management, 2006). These signals are either ignored or misinterpreted with often tragic results (Bradford, 2003; Choo, 2005; Thomas, 2006). The mainstream literature of crisis management acknowledges the importance of early crisis signal detection (Clair, 1993; Kash and Darling, 1998, Mitroff, 2005), yet, this aspect of crisis management has not

Table I. The six stages of a Policy Delphi

IJCHM 24,6

912

been adequately researched. Similarly most of the studies undertaken in hospitality and tourism field in the last decade, focus upon causes of crises that affected tourism and industry responses (Henderson, 2003; McKercher and Chon, 2004), their impact on the industry (Blake and Sinclair, 2003; Raab and Schwer, 2003; Vassilikopoulou et al., 2009) and recovery strategies (Armstrong and Ritchie, 2008; Johnson Tew et al., 2008; Lo et al., 2006). Initial attempts to develop a more strategic framework for managing crises in hospitality and tourism were made by Cassedy (1991) and then by Santana (1999), Faulkner (2001), Ritchie (2004), Paraskevas and Arendell (2007) and Hystad and Keller (2008). Yet, although these scholars acknowledge in their models the different stages of crisis management, and refer to signal detection, they do not explore it at any depth. The study, presented here aimed to narrow this gap by exploring the concept of crisis signal detection in an organisational context and, using a Policy Delphi, to analyse the conditions that influence the processes and practices of detection, transmission and interpretation of crisis signals by the management of a hospitality organisation. Further, it proposed a framework of enabling conditions that can be established in a hospitality organisation to make these processes and practices more effective. Research approach A review of the crisis management literature showed that the knowledge about how crisis signal detection can be developed in an organisational context is relatively limited, dispersed and could not serve as an external referent point for this study. It was therefore deemed appropriate for the study to take a phenomenological, inductive perspective with the help of individuals with expertise in organisational crisis management. From such a perspective, any knowledge on the subject of crisis signal detection would be constructed from a multiple set of interpretations that was part of the context in which it occurred. The concerns over Normative Delphi’s consensus highlighted the question of whether, consensus was needed for this study or if just “broad agreement” among participants about the signal detection framework would be sufficient. It was considered more important for the purpose of this study to adopt a technique that highlighted possible opposing views rather than aiming for consensus on conditions that might potentially influence the effectiveness of crisis signal detection (Dalkey and Helmer, 1963). Consequently, it was decided to use Policy rather than Normative Delphi. Sample selection Dunn (1994) suggested that the participants in a Policy Delphi should be selected to represent a wide range of opinions. Studies employing Delphi techniques normally recruit participants who have advanced knowledge of the subject under investigation, hence the term “panel of experts” (Linstone and Turoff, 2002). One of the major areas of methodological debate with regards to Normative Delphi is the selection and participation of “experts”. Sackman (1975, p. 703) questioned the definition of an expert in a field arguing that “it is almost impossible to find current psychometric or social science literature on ‘experts’”. Rowe and Wright (2001, p. 368) were also concerned with the participants’ “absolute and relative” level of expertise and the way that this affects the results of Normative Delphi. Policy Delphi overcomes this methodological hurdle by employing “informed advocates” who have both knowledge and experience of the subject under investigation (Eggers and Jones, 1998). However, although an “informed advocate”

may be knowledgeable about one aspect of a complex issue, they may not necessarily be able to formulate an opinion about the totality of the issue. Consequently, there is a need for participants’ heterogeneity to be preserved to assure the validity of the results (Linstone and Turoff, 2002). Linstone (1978, p. 294) argued that in the formulation of national policy “[. . .] it is important to include in the Delphi representatives of a large or wide spectrum of vested interests, ranging from bureaucrats to minority groups”. Policy Delphi studies therefore, usually employ purposive or criterion sampling. Both these sampling techniques assume researchers will have sufficient knowledge of the relevant population to select individuals that meet the purpose of their study. Purposive sampling allows researchers to use their judgement to select the participants that will best enable them answer their research questions (Saunders et al., 2012), whereas criterion samples are selected on the basis they meet a predetermined criteria of importance (Patton, 2002). In this crisis signal detection study, the sample selected met three predetermined criteria: (1) corporate level hospitality professionals in charge or directly involved with decisions related to risk/crisis management; (2) working in a multi-unit national or international hotel groups; and (3) having experienced at least one crisis incident in their organisations. .

The first criterion was necessary according to Patton (2002, p. 238) to distinguish between potential participants able to “[. . .] understand cases that are likely to be information-rich because they may reveal major system weaknesses that become targets of opportunity for program or system improvement”. The second criterion provided the specific context in the study; whilst the third, Patton’s (2002) “criterion of importance”, ensured relevant first-hand experience. In order to benefit fully from participants’ crisis management expertise, the study explored critical incidents in organisations, this focus on specific events enabling participants to provide a full, detailed description of each experience as it was lived (Stierand and Do¨rfler, 2012). This allowed the subjective nature of the “lived experience” of a crisis and its signal detection to be understood from the perspective of those experiencing it. The study could then develop inductively a conceptualization of crisis signal detection in the context of a hotel group. Participants were recruited from the Global Council on Safety, Security and Crisis Management of the International Hotel and Restaurant Association (IH&RA). A total of 22 Council members matched the selection criteria. However, for reasons ranging from increased workload and travelling commitments to high sensitivity of information to be shared, only 16 agreed to participate. This group conformed to guidelines given by Linstone (1978, p. 296) who suggest “a suitable minimum panel size is seven” and Turoff (2002, p. 82) who maintain “a Policy Delphi can be given to anywhere from ten to 50 people”. Within the prescribed context, the group was reasonably heterogeneous, participants varying in work background and years of experience (6-36 years), representing national and international hotel chains based in the Americas, Europe, Middle East and Africa (EMEA) and Asia Pacific (five companies being Anglo-American). Although the group was male-dominated (two female participants) this reflected industry norms.

Beyond consensus

913

IJCHM 24,6

914

Round one of Policy Delphi: issue identification and policy options available One major criticism of Normative Delphi based studies is what Rowe and Wright (2001) and Linstone (2002), among others, call “sloppy execution” or “simplification urge” in questionnaire design. They argue that the questionnaire used in the first round often contains far too simplistic statements or designs, associated tasks and measures being of uncertain relevance to the research purpose (Rowe and Wright, 2001). Linstone (2002) notes that some questionnaires are too long or allow contradictions in responses. Rowe and Wright (2001) also highlight the danger of structured first rounds in which statements are imposed without giving experts the opportunity to indicate the issues they believe to be more important for the subject. In addition, poorly designed questionnaires in the first round have been found to have an adverse effect on participants who, as a result, often opt not to continue in the subsequent rounds (Webler et al., 1991). In Policy Delphi, the first round may employ any appropriate method for data collection and, following McKenna’s (1994) advice, this study used face-to-face interviews with the participants. This round was designed to incorporate phases 1-3 of the design suggested by Turoff (2002) (Table I) namely: . formulation of issues; . exposing the options; and . determining initial positions on the issues. Within each interview critical incident technique (CIT) (Flanagan, 1954; Chell, 1998) was used to maximise the chance of capturing data about the most important conditions and issues. CIT is usually employed in the investigation of significant occurrences (events, incidents, or issues), which are identified by the interviewee, whom describes the way they are managed, and the outcomes in terms of perceived effects. “The objective is to gain understanding of the incident from the perspective of the individual, taking into account cognitive, affective, and behavioural elements” (Chell, 1998, p. 56). This allows participants to express their personal views of the described incident (Stauss and Weinlich, 1997), the inductive nature – especially when the topic has not been widely researched (Grove and Fisk, 1997), enabling powerful insights to be offered (Zeithaml and Bitner, 2003). Moreover, its “cultural neutrality” (De Ruyter et al., 1995), allows participants to offer their own perceptions on an issue, rather than indicate perceptions in response to researcher-initiated questions. CIT interviews in this study focussed on two discrete areas. The first asked participants to “re-live” a crisis they experienced; whereas the second explored their insights about what was learnt from the crisis, focussing on early detection of possible warning signals. Through these interviews participants were encouraged to offer as many personal insights as possible about the factors and conditions that influence crisis signal detection. To form a fuller picture of the context, details and effects of the critical incident on the organisation, content analysis was undertaken of documentary and other evidence provided by participants from other sources. These included company documentation such as internal reports, consultancy reports, meeting minutes, memos and e-mails, policy statements, standard operating procedure manuals, training manuals, as well as video and audio recordings of Crisis Team meetings and training materials. The analysis of this material corroborated and augmented the evidence from the participants’ accounts. Finally, articles and press coverage of the critical incidents were

utilized whenever possible. In certain cases, analysis of secondary evidence indicated the need of a follow-up interview with a participant to clarify aspects of the critical incident. These follow-up interviews were granted whenever needed. At the end of the interview cycle, responses were transcribed verbatim and participants asked to review them for content and interpretation accuracy. This “member checking”, helped establish credibility through establishing both internal (authenticity check) and external validity (transferability) in the overall process (Lincoln and Guba, 1985). Any amendments made in the transcripts were considered primary data, the data they replaced being discarded. Verified transcripts were coded and analysed using N-Vivo 7 and a summary report sent to participants. The report presented all responses anonymously, the themes that emerged being articulated as condition statements that would enable the timelier and more effective detection of crisis signals by an organisation. Each participant could distinguish her or his own response as it was highlighted in a different colour thereby addressing the criticisms of scarce feedback (Sackman, 1975) and “poor techniques of summarising and presenting the group response” (Linstone and Turoff, 2002, p. 6). Round two of Policy Delphi: measuring the spread of opinions The second round of Policy Delphi aimed to reveal participants’ opinions about the condition statements that emerged from Round One and explore the underlying assumptions, views, or facts used to support their respective positions (phases 4 and 5, Table I). Turoff (2002) argues that since Policy Delphi deals predominantly with statements and comments, the most appropriate way for evaluating participants’ ideas and opinions is to measure each against relative dimensions of desirability, feasibility, importance, and validity or confidence using four point rating scales. Desirability and feasibility ratings provoke discussions or debates among participants, which may generate new options, whilst importance and validity ratings are used to evaluate underlying assumptions or supporting arguments. Turoff (2002, p. 86) suggests that neutral answers should not be offered, arguing that “the lack of a neutral point promotes a debate which is in line with developing pros and cons as one primary objective”. Participants may however use the option of “no judgement” where they consider that they have insufficient expertise to evaluate the statement or are not able to express any opinion about it. A questionnaire was designed and distributed to participants requesting they evaluate, comment upon and rate each of condition statements developed in Round One using Turoff’s (2002) four dimensions. Participants were also encouraged to add new conditions where they felt appropriate and suggest ways that these could be created in their organisation. Their ratings for each condition statement were summarised to determine the levels of agreement (or disagreement) for this round. Linstone and Turoff (2002) argue that group agreement can be based on differing, or even opposing, assumptions and notes failure to pursue these assumptions may lead to false conclusions. Individual comments were therefore consolidated and analysed in order to identify the underlying themes and assumptions. A report was produced at the end of Round Two listing modified condition statements and highlighting those where broad agreement was reflected, as well as points and rationale of disagreement. This modified and expanded list of condition statements, along with their associated comments, allowed participants to also have “a feel” of the underlying reasons for responses.

Beyond consensus

915

IJCHM 24,6

916

Round three of Policy Delphi: final evaluation and conclusion In the final round of Policy Delphi participants were asked to “operationalise” their thoughts and views about how an effective crisis signal detection system should be planned in their own organisation. In doing this, they were instructed to take into consideration the points and rationale for disagreements expressed in the previous round, assess their relevance to their own organisations and propose actions for addressing them (phase 6 of Turoff’s design, Table I). Participants acted as representatives of their organisations rather than as individual anonymous participants (although anonymity was still maintained). This addressed criticisms that individuality and anonymity are often reasons for the lack of participant commitment who, behind their veil of anonymity, often rush “snap judgments” in the form of short or thoughtless written arguments instead of careful ponder and analysis of the issue (Sackman, 1975; Webler et al., 1991). Having achieved a broad understanding about the conditions needed to be in place for effective crisis signal detection, each participant proposed an action plan for the development and refinement of their own organisation’s signal detection process and practice (taking into consideration the disagreements and concerns highlighted in the report of Round Two). Subsequently, participants’ responses were analysed and the proposed actions consolidated to produce a general plan for the development of effective organisational crisis signal detection. Findings The first round of this Policy Delphi was used to identify the conditions that affect the effectiveness of organisational crisis signal detection and determine the participants’ views about them. Here the purpose was to elicit as many as possible differing views in order to create a “rich picture” of crisis signal detection and ways that it could be improved. The spread of participants’ opinions was, as expected, wide and even when they were talking about ostensibly the same topic they would often view it differently. For example, when participants referred to “culture” in aggregate, they meant between one and four different types of culture, namely, management culture, occupational culture, crisis culture and information culture. The first round yielded a number of conditions, which were refined by the researchers and articulated in ten condition statements to be explored and further refined or enriched in the subsequent rounds. It allowed a first identification of the detectors that can be part of the signal detection system as well as other building blocks of the detection system. In this round, the first elements of consensus appeared with regards to the different networks to be used. Alongside there were differences in opinions regarding potential detectors, reflecting differences in organisational culture, company size and budgets and participant’s expertise. Conditions identified in Round One were enriched and re-articulated, this time in eleven condition statements at the end of Round Two and were reconfirmed by the participants in Round Three as follows: Primarily the organisation needs to develop the right crisis culture which means that the organisation has to realise that it is vulnerable to a number of internal and external threats. Once this realisation is made, the organisation will understand that it needs to develop and actively support (with funding, tools and training) a quite complex network of detectors both internally and externally. This network will continuously scan the organisation’s physical,

information and cognitive domains. The signal detection process and practices as well the diversity of the detectors in the network need to continually evolve as crises themselves evolve and become more complex. One condition that will facilitate the more effective engagement of detectors in these networks is a decentralised, empowering management culture that favours continuous learning from own and other organisations’ experiences. Such culture will enable the creation of genuine crisis awareness and vigilance as well as better use of the organisation’s accumulated crisis knowledge for timely and effective response to the crisis signals. Decisions about action can be thus collectively made (in network hubs) at local regional or senior management level. Technology plays an important role in the connectivity of the network through multiple horizontal and vertical communication platforms as well as in the decision making about signals though decision support systems and crisis knowledge databases that provide a repertoire of responses to different types of crises. In the case of unknown threats, the latter, can be used as a reference point that will facilitate the recognition of patterns of signals which will help the decision makers to make better sense of the emergent crisis and take appropriate action. Although consensus was not sought in Policy Delphi and justification for differing opinions was shared among all participants, in these rounds the researchers encouraged a prioritisation of conditions and actions. In this prioritisation, a broad consensus emerged regarding how the detection system should be constructed and under which conditions it becomes more effective. Differences in opinions still existed; for example, some participants found that signal fusion hubs or that decision support systems were too expensive for their companies to afford. However, these elements were not considered primary for detection system design. Although some companies had them or were planning to incorporate them, they agreed that the system could work well even without them. It is noteworthy, however, that “quasi anonymity” worked in favour rather than against the purpose of the study. The possibility that responses might be attributed to persons encouraged – even forced – the participants to be open and truthful, and in many cases to maintain their views until the final round, providing insightful information. It can be therefore argued that, in Policy Delphi, “quasi anonymity” may reduce the threat of “groupthink” or false consensus. The outcome of this round was a commonly agreed Crisis Signal Detection Framework offering a “bird’s eye” view of an organisation’s crisis detection system. Further analysis of participants’ responses in these rounds resulted in the classification of human detectors in an Organisational Crisis Detector Typology. This was considerably differed from those presented in the extant crisis management literature offering two more dimensions: crisis specialism and the strength of the detector’s ties with the organisation. Differing opinions of participants allowed the classification of detectors into core, ad hoc and expert networks which should aid crisis managers in prioritising the detectors to be included in the system according to the organisation’s needs. Finally, Round Three showed development of a signal detection system can be “operationalised” through a gradual process which “matures” in five stages as proposed by the Signal Detection Maturity Model resulting from this round. Again, the difference of opinions about what is more important, feasible and desirable as the organisation’s signal detection strategy matures allowed the creation of this model.

Beyond consensus

917

IJCHM 24,6

918

Validity and reliability of Policy Delphi In their discussion of the validity of constructionist research designs Easterby-Smith et al. (2008, p. 97) emphasise that results should be believable and reached through transparent methods. Although precise replicability of Policy Delphi cannot be guaranteed, every effort was made to ensure reliability and validity of the findings. The criteria used here were based on Lincoln and Guba (1985) who propose: . credibility (in place of internal validity); . transferability (in place of external validity); . dependability (in place of reliability); and . confirmability (in place of objectivity). Credibility of Policy Delphi was ensured first through the selection and use of an expert participant panel as in the study’s sample. The criteria for selection were clear and relevant to the study, the possibility of bias not being considered significant as the panel was reasonably heterogeneous. Other steps to ensure credibility included the use of multiple sources of evidence in order to understand and interpret the context, causes and consequences of the critical incidents reported by the experts and the member checking employed during the first round of Policy Delphi. Credibility of the overall Policy Delphi findings was also enhanced by the confirmation of the “conditions” through a process of decision, strengthened by reasoned argument in which assumptions were challenged and tested by participants in successive rounds. Transferability of findings was ensured by what Lincoln and Guba (1985, p. 362) term a “thick description”. This comprised “thorough description of the context or setting within which the inquiry took place” and of “the transactions or processes observed in that context”. Participants were able to assess the relevance of conditions to their own organisations and, subsequently, the consolidation of their responses into a general plan. Since the study was designed to solicit experiences and opinions of experts in the field, the likelihood that the resulting findings could be transferred to other contexts and settings was increased. Dependability was ensured primarily by iterative feedback with participants. This is at the core of the Policy Delphi technique and also helps minimise possible researcher bias through the use of participants as independent “judges” in establishing reliable condition statements in the end of the first round. Finally, confirmability was ensured first by the findings and proposed model being grounded in the data collected; second, by all inferences, categorisations and analyses being confirmed by the participants and, third, by transparency of the method used and how sense was made of the data. 5. Conclusion The focus of this paper is purposefully on Policy Delphi since this technique has, until now, received no interest from hospitality researchers. The paper provides a number of insights into Policy Delphi and offers the first illustration of how it can be used as an alternative technique for researching complex subjects. Unlike the more widely known Normative Delphi that is often used as forecasting or a decision making tool based on the consensus of expert-participants, Policy Delphi is a research technique which seeks more to explore possible alternatives, identify and seek

the range of opinions on the issues that were germane to the research topic, and to potentially emerge with a consensus on key issues. Situations that appear particularly suited to Policy Delphi include those where new strategies or policies are sought and there is a need to explore or expose underlying assumptions, opinions or gather information to support new strategic or policy directions. For example, in the development of a marketing communications strategy for a tourism destination with multiple stakeholders (hoteliers, tour operators, airlines, local tourism authorities, etc) or in the development of a delegation of authority policy within an organisation, it is recognised that the idea of unbiased technical expertise does not apply as both these issues are defined by conflicting special interests. Moreover, where it is unlikely that consensus can be created, Policy Delphi can be used to identify the breadth of differing opinions. Through the iterative processes of prioritising elements of the strategy/policy in terms of confidence, importance, feasibility and desirability, Policy Delphi can either facilitate a trend to consensus or allow informed decisions to be made for the benefit of the majority of the involved parties. This default also removes the need to determine explicit percentage cut-off points for consensus. Policy Delphi can also be used to construct a feedback system that channels views and opinions of the various stakeholders with the goal of developing a more realistic and implementable policy. It is also likely to be of use in situations where management needs to find out the extent of diversity in opinions, or where the researcher is unaware of the full spectrum of issues and literature offers insufficient information. From a methodological perspective Policy Delphi offers researchers more flexibility than Normative Delphi. Rather than having to recruit proven “experts” on the research topic, a wide range of “informed” participants are required to ensure the maximum possible heterogeneity of the sample of informants. It also does not confine the first round of data collection to a prescribed method such as questionnaires. Anonymous recording of opinions and their use throughout the process without any urge to reach consensus appears to have had a positive effect on participant retention at all the stages of the study. The feedback structure of the technique allowed a focus upon issues that are often disregarded through the influence of powerful or vocal special interests and can set the stage for constructive debate. Such characteristics can make the Policy Delphi particularly attractive not only to researchers but also to government officials and policy makers. Although it is still the researcher’s responsibility to provide sufficient information to enable the impact of method to be assessed, unlike interviews, observation and focus groups, participants both remain anonymous and are actively involved in data generation and interpretation. This offers a significant advantage when undertaking inductive research by reducing the influence of dominant individuals on outcome. At the same time participants’ active involvement in the iterative interpretation processes further support the credibility, transferability, dependability and confirmability of findings. Because consensus is not the objective, Policy Delphi allows researchers to observe unforced shifts of opinion about different aspects of the topic in the course of a study through considering how participants’ desirability, feasibility, importance, and confidence ratings or their general stance towards an issue alter. Finally, future exploratory studies on policy and strategy development in hospitality and tourism destination contexts would do well to consider the use of Policy

Beyond consensus

919

IJCHM 24,6

920

Delphi as a tool to generate options and suggest alternative courses of action. Sample selection allows incorporation of the multiplicity of interests involved, recognising the value of different kinds of front-line expertise. Subsequent exploration of the distribution of opinions and reasons for these and the development and testing of condition statements enables both researchers and policy makers to expand their knowledge and understanding of the issue under investigation, isolate underlying reasons for non-constructive opinions and derive participant confirmed transferable findings. References Akkermans, H., Bogerd, P. and Vos, B. (1999), “Virtuous and vicious cycles on the road towards international supply chain management”, International Journal of Operations and Production Management, Vol. 19 Nos 5/6, pp. 565-81. Armstrong, K.E. and Ritchie, B.W. (2008), “The heart recovery marketing campaign: destination recovery after a major bushfire in Australia’s national capital”, Journal of Travel and Tourism Marketing, Vol. 23 Nos 2,3,4, pp. 175-89. Becker, C.S. (1992), Living and Relating: An Introduction to Phenomenology, Sage Publications, Newbury Park, CA. Blake, A. and Sinclair, T.M. (2003), “Tourism crisis management: US response to September 11”, Annals of Tourism Research, Vol. 30 No. 4, pp. 813-32. Bogdan, R. and Taylor, S.J. (1975), Introduction to Qualitative Research Methods: A Phenomenological Approach to the Social Sciences, Wiley, New York, NY. Bradford, L. (2003), “Fat foods: back in court: novel legal theories revive the case against McDonald’s”, Time On Line Edition, 3 August, available at: www.time.com/time/ insidebiz/article/0,9171,1101030811-472858,00.html (accessed 10 October 2008). Buck, A.J., Gross, M., Hakim, S. and Weinblatt, J. (1993), “Using the Delphi process to analyze social policy implementation: a post hoc case from vocational rehabilitation”, Policy Sciences, Vol. 26 No. 4, pp. 271-88. Cassedy, K. (1991), Crisis Management Planning in the Travel and Tourism Industry: A Study of Three Destinations and a Crisis Management Planning Manual, PATA, San Francisco, CA. Chell, E. (1998), “Critical incident technique”, in Symon, G. and Cassell, C. (Eds), Qualitative Methods and Analysis in Organizational Research: A Practical Guide, Sage, London, pp. 51-72. Choo, C.W. (2005), “Information failures and organizational disasters”, Sloan Management Review, Vol. 46 No. 3, pp. 8-10. Clair, J.A. (1993), “Turning poison into medicine: a grounded theoretical analysis of processes, pathologies, and designs in the detection of potential organizational crises”, unpublished doctoral dissertation, University of Southern California, Los Angeles, CA. Critcher, C. and Gladstone, B. (1998), “Utilizing the Delphi technique in policy discussion: a case study of a privatized utility in Britain”, Public Administration, Vol. 76 No. 3, pp. 431-49. Cunliffe, S. (2002), “Forecasting risks in the tourism industry using the Delphi technique”, Tourism, Vol. 50 No. 1, pp. 31-41. Custer, R.L., Scarcella, J.A. and Stewart, B.R. (1999), “The modified Delphi technique: a rotational modification”, Journal of Vocational and Technical Education, Vol. 15 No. 2, pp. 1-10. Dalkey, N. and Helmer, O. (1963), “An experimental application of the Delphi method to the use of experts”, Management Science, Vol. 9 No. 3, pp. 458-67.

De Ruyter, K., Perkins, D.S. and Wetzels, M. (1995), “Consumer-defined service expectations and post purchase dissatisfaction in moderately-priced restaurants: a cross-national study”, Journal of Consumer Satisfaction, Dissatisfaction, and Complaining Behavior, Vol. 8 No. 1, pp. 177-87. Delbecq, A.L., Van de Ven, A.H. and Gustafson, D.H. (1986), Group Techniques for Program Planning: A Guide to Nominal Group and Delphi Processes, Green Briar Press, Middleton, WI. Donohoe, H.M. and Needham, R.D. (2008), “Moving best practice forward: Delphi characteristics, advantages, potential problems, and solutions”, International Journal of Tourism Research, Vol. 11 No. 5, pp. 415-37. Dunn, W.N. (1994), Public Policy Analysis: An Introduction, 2nd ed., Prentice Hall, Englewood Cliffs, NJ. Easterby-Smith, M., Thorpe, R. and Jackson, P.R. (2008), Management Research, 3rd ed., Sage Publications, London. Eggers, R.M. and Jones, C.M. (1998), “Practical considerations for conducting Delphi studies: the oracle enters a new age”, Educational Research Quarterly, Vol. 21 No. 3, pp. 53-66. Evans, C. and Walpole, M. (1999), Compliance Cost Control: A Review of Tax Impact Statements in the OECD, (Research Study No. 27), Australian Tax Research Foundation, Sydney. Faulkner, B. (2001), “Towards a framework for tourism disaster management”, Tourism Management, Vol. 22 No. 2, pp. 135-47. Fink, S. (1986), Crisis Management: Planning for the Inevitable, American Management Association, New York, NY. Flanagan, J.C. (1954), “The critical incident technique”, Psychological Bulletin, Vol. 51 No. 4, pp. 327-58. Garrigos-Simon, F.J., Palacios-Marque, P. and Narangajavana, Y. (2007), “Improving the perception of hotel managers”, Annals of Tourism Research, Vol. 35 No. 2, pp. 359-80. Garrod, B. and Fyall, A. (2000), “Managing heritage tourism”, Annals of Tourism Research, Vol. 27 No. 3, pp. 682-708. Gill, J. and Johnson, P. (2010), Research Methods for Managers, 4th ed., Sage Publishing, London. Green, B., Jones, M., Hughes, D. and Williams, A. (1999), “Applying the Delphi technique in a study of GPs’ information requirements”, Health and Social Care in the Community, Vol. 7 No. 3, pp. 198-205. Grove, S.J. and Fisk, R.P. (1997), “The impact of other customers on service experiences: a critical incident examination of ‘getting along’”, Journal of Retailing, Vol. 73 No. 1, pp. 63-85. Gupta, U.G. and Clarke, R.E. (1996), “Theory and applications of the Delphi technique: a bibliography (1975-1994)”, Technological Forecasting and Social Change, Vol. 53 No. 2, pp. 185-211. Henderson, J.C. (2003), “Terrorism and tourism: managing the consequences of the Bali bombings”, Journal of Travel & Tourism Marketing, Vol. 15 No. 1, pp. 41-58. Hsu, C.-C. and Sandford, B.A. (2007), “The Delphi technique: making sense of consensus, practical assessment”, Research and Evaluation, Vol. 12 No. 10, pp. 1-8, available at: http:// pareonline.net/pdf/v12n10.pdf (accessed 28 September 2010). Hystad, P.W. and Keller, P.C. (2008), “Towards a destination tourism disaster management framework: long-term lessons from a forest fire disaster”, Tourism Management, Vol. 29 No. 1, pp. 151-62.

Beyond consensus

921

IJCHM 24,6

922

Institute of Crisis Management (2006), Annual ICM Crisis Report: News Coverage of Business Crises During 2005, Vol. 15 No. 1, June, available at: www.crisisexperts.com/05report.htm (accessed 8 June 2011). Johnson Tew, P., Lu, Z., Tolomiczenko, G. and Gellatly, J. (2008), “SARS: lessons in strategic planning for hoteliers and destination marketers”, International Journal of Contemporary Hospitality Management, Vol. 20 No. 3, pp. 332-46. Kash, T.J. and Darling, J.R. (1998), “Crisis management: prevention, diagnosis and intervention”, Leadership & Organization Development Journal, Vol. 19 No. 4, pp. 179-86. Kaynak, E. and Macaulay, J.A. (1984), “The Delphi technique in the measurement of tourism market potential”, Tourism Management, Vol. 5 No. 2, pp. 87-101. Lincoln, Y. and Guba, E. (1985), Naturalistic Inquiry, Sage Publications, Beverly Hills, CA. Linstone, H.A. (1978), “The Delphi technique”, in Fowles, R.B. (Ed.), Handbook of Futures Research, Greenwood, Westport, CT, pp. 271-300. Linstone, H.A. (2002), “Eight basic pitfalls: a checklist”, in Linstone, H.A. and Turoff, M. (Eds), The Delphi Method: Techniques and Applications, Addison-Wesley, Reading, MA, pp. 559-71. Linstone, H.A. and Turoff, M. (Eds) (2002), The Delphi Method: Techniques and Applications, Addison-Wesley, Reading, MA. Lo, A., Chung, C. and Law, R. (2006), “The survival of hotels during disaster: a case study of Hong Kong in 2003”, Asia Pacific Journal of Tourism Research, Vol. 11 No. 1, pp. 65-79. Loughlin, K.G. and Moore, L.F. (1979), “Using Delphi to achieve congruent objectives and activities in a paediatrics department”, Journal of Medical Education, Vol. 54 No. 2, pp. 101-6. McKenna, H.P. (1994), “The Delphi technique: a worthwhile research approach for nursing?”, Journal of Advanced Nursing, Vol. 19 No. 6, pp. 1221-5. McKercher, B. and Chon, K. (2004), “The over-reaction to SARS and the collapse of Asian tourism”, Annals of Tourism Research, Vol. 31 No. 3, pp. 716-9. Mara, C.M. (2000), “A strategic planning process for a small non-profit organization”, Nonprofit Management and Leadership, Vol. 11 No. 2, pp. 211-23. Miller, G. (2001), “The development of indicators for sustainable tourism: results of a Delphi survey of tourism researchers”, Tourism Management, Vol. 22 No. 3, pp. 351-62. Mitroff, I.I. (2005), Why Some Companies Emerge Stronger and Better from a Crisis: 7 Essential Lessons for Surviving Disaster, AMACOM, New York, NY. Moore, C.M. (1987), Group Techniques for Idea Building, Sage Publishing, Newbury Park, CA. Murphy, K. and Olsen, M. (2009), “Dimensions of a high performance management system: an exploratory study of the US casual restaurant segment”, International Journal of Contemporary Hospitality Management, Vol. 21 No. 7, pp. 836-53. O’Loughlin, R. and Kelly, A. (2004), “Equity in resource allocation in the Irish health service: a policy Delphi study”, Health Policy, Vol. 67 No. 3, pp. 271-9. Paraskevas, A. and Arendell, B. (2007), “A strategic framework for terrorism prevention and mitigation in tourism destinations”, Tourism Management, Vol. 28 No. 6, pp. 1560-73. Patton, M. (2002), Qualitative Research and Evaluation Methods, 3rd ed., Sage Publications, Thousand Oaks, CA. Raab, C. and Schwer, R.K. (2003), “The short- and long-term impact of the Asian financial crisis on Las Vegas strip baccarat revenues”, International Journal of Hospitality Management, Vol. 22 No. 1, pp. 37-45.

Rainhorn, J., Brudon-Jakobowicz, P. and Reich, M. (1994), “Priorities for pharmaceutical policies in developing countries: results of a Delphi survey”, Bulletin of World Health Organisation, Vol. 72 No. 5, pp. 257-64. Reid, N. (1988), “The Delphi technique: its contribution to the evaluation of professional practice”, in Ellis, R. (Ed.), Professional Competence and Quality Assurance in the Caring Professions, Chapman & Hall, London, pp. 230-354. Ritchie, B.W. (2004), “Chaos, crises and disasters: a strategic approach to crisis management in the tourism industry”, Tourism Management, Vol. 25 No. 6, pp. 669-83. Rowe, G. and Wright, G. (1999), “The Delphi technique as a forecasting tool: issues and analysis”, International Journal of Forecasting, Vol. 15 No. 4, pp. 353-75. Rowe, G. and Wright, G. (2001), “Expert opinions in forecasting: role of the Delphi technique”, in Armstrong, J.S. (Ed.), Principles of Forecasting, Kluwer Academic Press, Norwell, MA, pp. 125-44. Sackman, H. (1975), Delphi Critique, Lexington Books, Lexington, MA. Santana, G. (1999), “Tourism: towards a model for crisis management”, Tourizam, Vol. 47 No. 1, pp. 4-12. Saunders, M.N.K., Lewis, P. and Thornhill, A. (2012), Research Methods for Business Students, 6th ed., FT Prentice Hall, London. Singh, A.J. (2000), “The US lodging industry in the new millennium – a Delphi study to predict changes in the lodging industry structure, performance and capital sources in years 2000 and 2005”, Journal of Hospitality Financial Management, Vol. 8 No. 1, pp. 1-22. Smit, J. and Mason, A. (1990), “A policy Delphi study in the socialist Middle East”, Public Administration and Development, Vol. 10 No. 4, pp. 453-65. Sobaih, A.E.E., Ritchie, C. and Jones, E. (2012), “Consulting the oracle? Applications of modified Delphi technique to qualitative research in the hospitality industry”, International Journal of Contemporary Hospitality Management, Vol. 24 No. 6, pp. 886-906. Spenceley, A. (2005), “Nature-based tourism and environmental sustainability in South Africa”, Journal of Sustainable Tourism, Vol. 13 No. 2, pp. 136-70. Stauss, B. and Weinlich, B. (1997), “Process-oriented measurement of service quality: applying the sequential incident technique”, European Journal of Marketing, Vol. 31 No. 1, pp. 33-55. Stierand, M.B. and Do¨rfler, V. (2012), “Reflecting on a phenomenological study of creativity and innovation in haute cuisine”, International Journal of Contemporary Hospitality Management, Vol. 24 No. 6, pp. 946-57. Strauss, H.J. and Zeigler, L.H. (1975), “The Delphi technique and its uses in social science research”, Journal of Creative Behavior, Vol. 9 No. 4, pp. 253-9. Thomas, W.J. (2006), “The Vioxx story: would it have ended differently in the European Union?”, American Journal of Law and Medicine, Vol. 32 Nos 2-3, pp. 365-80. Turoff, M. (1970), “The design of a policy Delphi”, Technological Forecasting and Social Change, Vol. 2 No. 2, pp. 149-72. Turoff, M. (2002), “The Policy Delphi”, in Linstone, H.A. and Turoff, M. (Eds), The Delphi Method: Techniques and Applications, Addison-Wesley, Reading, MA, pp. 80-96. Vassilikopoulou, A., Siomkos, G., Chatzipanagiotou, K. and Triantafillidou, A. (2009), “Hotels on fire: investigating consumers’ responses and perceptions”, International Journal of Contemporary Hospitality Management, Vol. 21 No. 7, pp. 791-815. Webler, T., Levine, D., Rakel, H. and Renn, O. (1991), “A novel approach to reducing uncertainty: the group Delphi”, Technological Forecasting and Social Change, Vol. 39 No. 3, pp. 253-63.

Beyond consensus

923

IJCHM 24,6

924

Wikstrom, S. (1978), “Information about consumer problems by the policy-Delphi procedure: a new method for evaluating the marketing system”, Advances in Consumer Research, Vol. 5 No. 1, pp. 596-602. Yong, Y.W., Keng, K.A. and Leng, T.L. (1988), “A Delphi forecast for the Singapore tourism industry: future scenario and marketing implications”, International Marketing Review, Vol. 6 No. 3, pp. 35-46. Zeithaml, V.A. and Bitner, M.J. (2003), Services Marketing: Integrating Customer Focus across the Firm, 3rd ed., McGraw-Hill, New York, NY. Further reading Anderson, J.C., Rungtusanatham, M. and Schroeder, R.G. (1994), “A theory of quality management underlying the Deming management method”, Academy of Management Review, Vol. 19 No. 3, pp. 472-509. Mills, S.F. and Hudson, R. (2000), “Foodservice manager 2000”, Journal of Hospitality and Tourism Research, Vol. 17 No. 1, pp. 147-59. About the authors Alexandros Paraskevas is Senior Lecturer in Strategic Risk Management at Oxford Brookes University (UK). His hospitality industry background includes auditing and operations management positions for over a decade with Marriott and Starwood. His research interests focus on the management of risks/crises and disasters both in the organisational and tourism destination context. He has served as Advisor of the International Hotel and Restaurant Association’s (IH&RA) Global Council on Security, Safety and Crisis Management and is member of ASIS (the American Society for Industrial Security professionals) and HEAT (ASIS Council for Hospitality and Tourism). He is a member of the Complexity Research Group at the London School of Economics and the author of Planning Research in Hospitality and Tourism (2007, Butterworth-Heinemann) together with Levent Altinay. Alexandros Paraskevas is the corresponding author and can be contacted at: [email protected] Mark N.K. Saunders is Professor in Business Research Methods at the University of Surrey, School of Management (UK). His research interests focus on two themes. The first, research methods, includes the development of tools to learn about, understand and improve organisational relationships within a process consultation framework, online research methods and methods for researching trust. The second, human resource aspects of the management of change, is concerned particularly with trust and justice. Recent books include: Research Methods for Business Students (2009, FT Prentice Hall), now in its fifth edition, co-authored with Phil Lewis and Adrian Thornhill; and Organizational Trust: A Cultural Perspective (2010, Cambridge University Press) co-edited with Denise Skinner, Nicole Gillespie, Graham Dietz and Roy Lewicki.

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0959-6119.htm

How to promote cooperation in the hospitality industry Generating practitioner-relevant knowledge using the GABEK qualitative research strategy Harald Pechlaner Catholic University of Eichsta¨tt-Ingolstadt, Eichsta¨tt, Germany, and European Academy of Bolzano-Bozen, Bolzano-Bozen, Italy, and

Michael Volgger European Academy of Bolzano-Bozen, Bolzano-Bozen, Italy

How to promote cooperation

925 Received 26 December 2010 Revised 28 April 2011 8 September 2011 26 November 2011 27 January 2012 Accepted 22 February 2012

Abstract Purpose – While it is possible to classify the previously suggested conditions to the promotion of interorganizational cooperation as either referring to strategic interdependence or to structural and procedural conditions, it is unclear which approach is more critical to the promotion of local and regional cooperation in the hospitality industry at the network level. Therefore, the purpose of this paper is twofold: to inductively develop propositions regarding the promotion of such cooperation in order to evaluate the relative importance of the two conflicting positions, and to demonstrate the suitability of GABEK to the development of these propositions. Design/methodology/approach – Following a qualitative case study design, data were gathered by conducting 15 open interviews in a South Tyrolean destination and analyzed with the aid of the GABEK technique. Findings – The results suggest that the structural and procedural conditions are relatively more critical to the promotion of interorganizational cooperation in the hospitality industry: first, the balancing of the efficiency and legitimacy dimensions via the organizational and procedural design of the cooperation seems crucial; second, intermediary organizations, if characterized by good corporate governance, may have a positive influence on cooperation; and third, the application of GABEK-like tools may facilitate the finding of commonly shared solutions in cooperative configurations. Practical implications – The research reveals actionable insights regarding the improvement of interfirm cooperation by offering managers tangible measures and a tool for implementing the theoretical concepts. Originality/value – The paper is original because it contributes to evaluating the two main approaches to the promotion of cooperation and because it proposes an innovative methodology (GABEK) for practice-oriented qualitative research. Keywords Hospitality industry, Destination governance, Cooperation, Qualitative research, Hospitality services, Research methods Paper type Research paper

1. Introduction The hospitality industry is facing increasing international competition and ever-more sophisticated and rapidly changing customer demands (Hu et al., 2009). To meet these challenges, hospitality and tourism scholars recommend overcoming the traditional fragmentation in the industry by amplifying cooperative agency (Pechlaner and Tschurtschenthaler, 2003). The specific economic rationale behind the suggested

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 pp. 925-945 q Emerald Group Publishing Limited 0959-6119 DOI 10.1108/09596111211247245

IJCHM 24,6

926

interorganizational cooperation – understood in a very general sense to mean “work[ing] jointly towards the same end” (Pearsall, 2001) by “recurring exchange relationships among a limited number of organizations that retain residual control of their individual resources” (Ebers, 1997) – is the bundle structure of the tourism product (Candela and Figini, 2010). Despite the general recommendation for interorganizational cooperation, hospitality research has provided little insight into how it could actually be put into practice, especially as far as local and regional cooperation is concerned. Interorganizational cooperation encompasses phenomena such as establishing linkages between a group of actors (private firms, as well as public and private-public organizations) and sharing knowledge, resources, risks, and profits within this group. However, because interorganizational cooperation also carries significant risks (due to opportunism, conflicts of interest, and asymmetric information), the willingness of single actors to cooperate does not necessarily exist a priori. Drawing on Galaskiewicz (1985), Oliver (1990) proposes a list of six conditions under which interorganizational relationships generally tend to form: necessity, asymmetry, reciprocity, efficiency, stability, and legitimacy. The first three conditions – necessity, asymmetry and reciprocity – are rooted in research traditions that ascribe the willingness to cooperate to some kind of strategic interdependence (Pfeffer and Salancik, 1978; Casciaro and Piskorski, 2005; Wang and Zajac, 2007). The latter three – efficiency, stability, and legitimacy – are backed by approaches in the interorganizational relationship literature that consider institutional, structural and procedural aspects (such as laws, contracts, routines, organizations and social embeddedness) to be more important (Williamson, 1979; Burt, 1992; Gulati, 1998). However, the evaluation of the relative importance of the different conditions for interorganizational cooperation, as well as the application of these insights in the field of hospitality, remains incomplete (see Babiak, 2007). In particular, it is still unclear whether some of these conditions are more influential and critical for the promotion of interorganizational relationships in hospitality than others. For instance, necessity, asymmetry and reciprocity appear to be ubiquitous in regional and local networks in hospitality, but despite this fact, few hospitality scholars have stressed the role of strategic interdependence (Wang and Krakover, 2008). On the other hand, theoretical contributions to destination governance insistently argued for the importance of the structural and procedural conditions to the promotion of local and regional cooperation in the hospitality industry (Beritelli et al., 2007). Nevertheless, such research on destination governance lacks specified recommendations and methods for translating these general considerations into practical measures. Thus, on an empirical and practical level, we do not yet know which dimensions proposed by Oliver (1990) fit to what extent the “management of food, beverages and/or accommodation in a service context” (Higher Education Funding Council for England HEFCE, 1998). Thus, the problem narrows to the following research questions: How can local and regional interfirm cooperation in the hospitality industry be successfully promoted in a tangible way? In particular, are structural, procedural and institutional conditions really relatively more critical for the establishment of such cooperation in hospitality, as suggested by the research on destination governance, or is strategic interdependence the more important facilitator?

Our study addresses these questions by inductively developing a set of propositions concerning the promotion of local and regional interorganizational cooperation in the hospitality industry. We present and apply GABEK as a qualitative research strategy to develop such empirically rooted and practically relevant propositions (Zelger, 2000). GABEK is an abbreviation of a German phrase that describes a holistic handling of complexity. It places special emphasis on the data analysis phase, and provides a sophisticated software tool (WinRelan) for its execution. The GABEK method may also function as part of the solution to the “promoting-cooperation-in-practice problem”: while it is a useful research strategy, it can also be employed as a practical decision support device to foster decisions that promote cooperation on a regular basis. Hence, this study shows that GABEK may be useful for tourism researchers and practitioners who need to collect, order and interpret existing collective knowledge. Thus, the specific purposes of this paper are threefold: first, to inductively generate propositions regarding the tangible organization of cooperation in tourist destinations in practice in order to judge the applicability of the list developed by Oliver (1990) to the context of tourist destinations; second, to demonstrate a GABEK qualitative research method application to the development of these propositions; and third, to show how GABEK can be applied as a decision support tool in practice. 2. Literature review The literature review is organized as follows: First, we discuss the gaps in the treatment of interorganizational cooperation in the specific hospitality literature; second, we examine in detail the theoretical contribution of the research on destination governance and the related empirical studies; and third, we outline GABEK as a methodological basis for understanding the promoters of local and regional cooperation in hospitality in practice. Cooperation in hospitality research A review of four major journals of hospitality management (International Journal of Hospitality Management, International Journal of Contemporary Hospitality Management, Journal of Hospitality and Tourism Research, and Cornell Hospitality Quarterly – see Hu and Racherla, 2008) shows that the treatment of the cooperation theme in this specific research area is limited in at least two respects. First, regarding the functional and geographic extensiveness of cooperation, the hospitality literature limits analyses either to the intra-organizational level (“teamwork”) (Hu et al., 2009; Yang, 2010; Chathoth et al., 2011) or to the inter-organizational level on an explicit international scale (Altinay, 2006; Lee, 2008; Xiao et al., 2008). Thus, astonishingly little treatment of local and regional forms of cooperation (e.g. destinations) is evident in the more specific hospitality literature, in spite of their special importance to tourism. The study of local and regional cooperation is important because, in contrast to international alliances, within tourist destinations actors are much less free to choose their alliance partners. Moreover, contrary to international alliances, which mostly involve large corporations, cooperation within tourist destinations must take into account different types of actors and SMEs (Fyall and Garrod, 2004; Wang, 2008; Novelli et al., 2006). A second limitation of the hospitality literature on cooperation is that it largely ignores non-business actors, even though they often play crucial mediating roles in

How to promote cooperation

927

IJCHM 24,6

928

local and regional networks. In fact, most papers focus exclusively on the various forms of corporate alliances at the expense of collective actors (Crotts et al., 2001; Chathoth and Olsen, 2003). Yet it is imperative to consider also public actors, because they differ significantly from private actors with respect to their values, ideologies and interests (d’Angella and Go, 2009). Overall, the consequent assessment of the organization of cooperation within destinations is missing from the specific hospitality literature. Notable exceptions are the contributions of Tinsley and Lynch (2001), Timur and Getz (2008), and Wang and Krakover (2008). To fill the remaining gaps, this study focuses on the promotion of local and regional cooperation in the hospitality industry. Destination governance as a cooperative theory Some of the highlighted limitations of the hospitality literature may be overcome by applying the conceptual lens of destination governance theory. The destination governance approach deals with cooperation from a structural, destination network point of view. Thereby, the consideration of local and regional cooperation forms that go beyond B2B (such as public-private partnerships) becomes almost inevitable (Nordin and Svensson, 2007). Governance is often used to describe forms of “social coordination” (Wang and Xiang, 2007), or as Rhodes (1996) puts it: “Governance is about managing networks.” However, a more economic understanding of the governance concept is also sometimes applied. Used in the latter sense, governance refers to “economizing on transaction costs” (Williamson, 1999). This perspective demands that transactions be carried out in the most transaction cost-saving form. A widespread classification of such governance modes differentiates between markets, hierarchies and networks (Wang and Xiang, 2007). Their respective transaction costs are context-dependent; that is, they depend on the frequency, uncertainty, and specificity of a transaction. Both concepts of governance are relevant to the destination governance approach. From their combination derive two of its main characteristics: (1) the preference for the network as a structural framework for cooperation; and (2) its relativism and pragmatism. Both notions point out that destination governance is mainly engaged with structural and institutional conditions of cooperation, most notably with the mitigation of appropriation concerns due to opportunistic behavior (Brass et al., 2004). Regarding the first point, destination governance posits the network of partially autonomous players as the basic organizational model (Svensson et al., 2005; Raich, 2006). Although central steering possibilities are viewed critically, the hierarchy mode of governance is not completely dismissed. This is because structure is considered crucial to limit behavioral uncertainty and opportunism risks (Gulati, 1998). Second, destination governance is a relativistic approach to the promotion of cooperation. Since destination governance follows the context-dependency assumption of Williamson (1999) and holds the view that every destination is individual in some respects, it is highly skeptical towards the replicability of detailed sample solutions in the promotion of cooperation (Nordin and Svensson, 2007). Therefore, a priori it leaves unanswered the question about how to find the right solution in the specific case of a single destination and calls for tailored approaches.

Prior empirical research on cooperation From a theoretical point of view, therefore, the question of how to promote local and regional cooperation in practice remains unclear. Prior empirical research has touched on the issue only in a cursory manner, and the empirical evidence has been presented in a rather fragmented fashion. In regard to the practical promotion of cooperation in destinations, the related literature stresses, amongst other things: . shared objectives (Reid et al., 2008; Hu et al., 2009); . cultural congruence and attitude to cooperation (Wang and Fesenmaier, 2007; Lemmetyinen and Go, 2009); . character of leaders (Wang and Fesenmaier, 2007; Lemmetyinen and Go, 2009); . accepted and visionary tourism organizations, as well as broad involvement of stakeholders (Dredge, 2006; Wang and Fesenmaier, 2007; Reid et al., 2008); and . communication and joint knowledge creation (Breukel and Frank, 2009; Lemmetyinen and Go, 2009). Approaching the issue by simply combining these previous findings was deemed inadequate because they are situated on diverging levels of generalization and thus are difficult to match. Therefore, a theory-building, empiric study was undertaken to develop single-sourced propositions on how cooperation in destinations might be fostered by practical measures. GABEK and other methods in the study of cooperation In order to develop the required practical proposals, we applied an innovative qualitative research strategy: GABEK. GABEK’s methodological background involves phenomenological gestalt theory (Stumpf, 1939) and its expansion to a theory of linguistic gestalten (Zelger, 1999). This phenomenological orientation of GABEK requires the collection of perceptions through open qualitative interviews; then, on these raw data a keyword-based analysis is performed. By transforming statements into networks of interrelated keywords, GABEK allows the extraction of the main messages, as well as the capture of a phenomenon’s evaluation and its perceived causes and effects. While still relatively unknown to the wider international audience, GABEK has already been used in several business and social science studies (Buber, 2004; Mueller and Raich, 2005; Raich, 2008). Due to the complexities involved, qualitative research is generally considered suitable for the study of cooperation (Wang, 2008). Despite the consensus regarding the appropriateness of a qualitative approach, only a small minority of the performed qualitative studies presents a clear analytical method (Wang and Fesenmaier, 2007). Qualitative studies that focus mainly on data collection techniques while neglecting data analysis represent a lacuna that is evident in the whole tourism and hospitality field (Mehmetoglu and Altinay, 2006) and also common in management literature (Gephart, 2004). Therefore, the application of GABEK offers a rare opportunity to combine the ability of qualitative approaches to deal with complexity with a clear and traceable analysis process. The established qualitative methodologies employed in the extant hospitality and tourism research (see Mehmetoglu and Altinay, 2006) include ethnography (Bruner, 1995), phenomenology (Cohen, 1979), ethnomethodology (McCabe, 2007) and grounded

How to promote cooperation

929

IJCHM 24,6

930

theory (Glaser and Strauss, 1967). Among these, only grounded theory gives the investigator clear procedures of data analysis. Ethnography (for an application see Martin and Woodside, 2012), for instance, emphasizes the accuracy and comprehensiveness of qualitative data collection, whereas the data analysis phase remains under-specified. Phenomenology (for an application see Stierand and Do¨rfler, 2012) is underpinned by substantial philosophical thinking (Husserl, 1962), but has never really transformed its broad methodological assumptions into a tangible set of procedures. As with grounded theory, GABEK differs from these research strategies in that it provides the researcher with clear procedures of data analysis. Indeed, GABEK and grounded theory share numerous commonalities: both methods are systematically designed and embedded in an overarching research strategy; and both offer the respondents and investigators extensive freedom; but they also lay down clear rules, which facilitate computer implementation. Their differences can be summarized as follows: first, the modularity of GABEK’s analysis steps makes it more adaptable to the respective research purpose. Second, in GABEK respondents’ statements and interpretations by the investigators can be clearly distinguished. Third, the figurative output of GABEK produces an elegant visual representation of raw data and thus facilitates its interpretation. Fourth, the results generated by GABEK are suitable for deriving concrete recommendations, a feature which makes GABEK ideally suited for the research purpose of the present study. Indeed, the main rationale for preferring GABEK over grounded theory lies in its ability to generate practitioner-relevant knowledge in exploratory research attempts. 3. Methodology: qualitative case study research with GABEK Study setting and case selection To generate propositions for the practical promotion of cooperation in destinations, we selected a GABEK-based qualitative case study design. Particularly, the case study approach was deemed appropriate for supporting the operationalization of general notions for a concrete destination. For the case study location we opted for the South-Tyrolean (northern Italian) destination of Eppan-Kaltern-Tramin. The rationale behind this choice included the destination’s distinctive cooperation deficiencies. These are partly due to the administrative fragmentation of Eppan-Kaltern-Tramin, which is separated into three distinct municipalities. Although the three municipalities form a single destination from a demand point of view, the organization of the tourism supply largely follows the political partitions. The present case study provides value by examining a destination that suffers from political and administrative boundaries, which inhibit cooperation. Many destinations, particularly those situated on national borders, suffer from a similar mismatch of political and destination boundaries. However, the destination’s tourism actors have recognized the need for more cooperation, and are currently highly motivated to take concrete steps in that direction. Our study collects and systemizes their suggestions concerning the practical promotion of cooperation. These proposals are of particular interest because they are born out of adverse circumstances. Indeed, the high pressure to act present in the destination is likely to result in robust and concrete proposals.

Data collection We conducted 15 qualitative interviews, which lasted approximately one hour each and were audio taped and transcribed. The interviewees were chosen by purposeful sampling following the criteria of knowledgeability and diversity (Eisenhardt and Graebner, 2007; Raich, 2008; Pratt, 2009): We tried to interview informants who could offer a good insight into the situation in the tourist destination. On the other hand, to avoid accumulating biased information, the informants were selected in a way to ensure that diverse perspectives were represented in the interviews. Specifically, we strived to include in the sample all the major elements of the local tourism service bundle (see Table I) the: . diverse industries involved – represented by the heads of the professional associations; . local and regional DMOs – represented by the presidents or general managers of the seven DMOs operating in the area; and . public administration – represented by the respective persons responsible for tourism in the three municipal councils.

How to promote cooperation

931

The semi-structured interviews were kept as guided as necessary and as open-ended as possible (seven predetermined, guiding questions; follow-up questioning allowed), so that the actors could convey their relevance structures and formulate the problems and solutions as they conceptualized them. This openness in data collection is prescribed by qualitative theory development in general (Gephart, 2004) and the GABEK research strategy in particular (Zelger, 2000). Data analysis As applied in our study, the GABEK analysis pursues the main goal of portraying the qualitative interviews’ content in a style that has a reduced complexity, but is still Category

Number of interviewees

Professional associations Hoteliers and restaurateurs association Farmers association Trades association

3 1 1

Tourism organizations Local Tourism Association – Eppan Local Tourism Association – Kaltern Local Tourism Association – Tramin Regional Tourism Association Winepoint Kaltern Wineroute Association Provincial Tourism Association

1 1 1 1 1 1 1

Public administration Municipal Council of Eppan Municipal Council of Kaltern Municipal Council of Tramin Total

1 1 1 15

Table I. The sample: major elements of the local tourism service bundle

IJCHM 24,6

932

comprehensive. Since it is a cross-interview analysis, it does not bother so much about the who but rather about the what, grouping answers by topics, not by respondents. The basic idea is to represent these answers as networks of interrelated keywords by taking into account semantics and (to a lesser degree) syntax (see also Pask, 1976). To achieve this objective, the steps listed in Table II were performed with the aid of the WinRelan software. Roughly overviewed, Step 1 (coding) provides the complexity-reduction of the statement system, and Step 2 (data processing and interpretation) its restructuring. Looked at it in more detail, the procedure starts with the subdivision of the transcribed interviews into meaningful statements called “text units” (Step 1a). These resulting text units are coded with regard to the contained keywords (1b), evaluations (1c), and suggested cause-and-effect relationships between those keywords (1d). Moreover, the keywords are classified as basic values, goals, measures, or basic conditions/constraints (1e). For instance, an interview statement such as “it is a problem: drug abuse often leads to petty crime” might be coded through the keywords “drug abuse” and “petty crime”, which would both be marked as negatively evaluated and as being causally interrelated (drug abuse ! petty crime). Drug abuse might be considered a basic condition/constraint. Step 2, the data processing and subsequent interpretation, is aimed at restructuring the content of the interviews in order to present it in a more transparent manner (i.e. as graphs of keywords) and interpret it accordingly. Data processing in WinRelan is performed semi-automatically. Causal net graphs form a major automatically generated output of this analysis (2a), and they comprehensively display the cause-and-effect relationships between the keywords as the respondents perceive them. Generally, each connection in a causal net graph is based on statements taken directly from the interviews. Whereas GABEK places a special emphasis on coding and data organization, and accordingly provides relatively clear rules for its execution (see Table II), interpretation is less prescriptive. As stated above, GABEK differs from other qualitative analysis techniques in that it strictly separates coding from interpretation: the coding should be performed as inductively as possible, whereas research questions and theory come in only at the interpretation stage. At the stage of interpretation, however, GABEK considers assessing patterns by grouping the keywords and assigning theoretically meaningful constructs. We based interpretation primarily on causal net graphs because the perceived cause-and-effect-relationship nets clearly display the respondents’ suggestions regarding the amelioration of cooperation (2b). Furthermore, the graphical distinction among basic values, goals, measures, or constraints/basic conditions makes the causal net graphs a very useful instrument from an implementation point of view. 4. Results Promoters of cooperation and underlying dimensions Figure 1 shows a causal net graph, which displays the perceived cause-and-effect relationships around the selected keyword cooperation. Arrows stand for direct relationships; circular ending connecting lines for inverse relationships. The above-mentioned categorization of the keywords, which should facilitate the implementation process, is also pictured in the figure: the shapes of the text boxes allow differentiation between basic values (rectangles), goals (rounded corner rectangles), measures (hexagons), and basic conditions and constraints (ellipses).

No Low

No

No

Low to medium

1. Coding (a) Definition of text units

(b) Coding of keywords

(c) Coding of evaluations

(d) Coding of causal relations

(e) Coding of relevancy

Underlying rules

Coding done without interpretation Manual definition of text units; display of each Length of the text units depends on semantic text unit on an index card meaningfulness and the number of keywords used within a statement (limited to nine) Manual definition of keywords; display of the Text units coded with a set of three to nine keywords that identified keywords on the same index card as semantically can stand on their own (each) and (as a set) are the underlying text unit; a list of previously suited to represent the core message of the text unit; defined keywords helps avoid synonyms and keywords inductively developed and remaining within the homonyms respondents’ language; synonyms and homonyms avoided Display of the evaluated keywords on the same Respondents’ articulated evaluations identified: positively index card as the underlying text unit; for each or negatively evaluated keywords marked as such keyword, it is possible to manually assign positive (þ) or negative (2) evaluations (binary coding) Display of the causally coded keywords on the Implied influences and causal relationships in the same index card as the underlying text unit; for respondents’ statements identified: stated direct and inverse each pair of keywords within a text unit, it is relationships between keywords marked as such possible to code direct and inverse relationships (binary coding on a matrix of keywords) Number of (positive/negative) evaluations and Keywords categorized as basic values, goals, measures, or (ingoing/outgoing) causal relations is displayed basic conditions and constraints; several formal and for each keyword as a support to content-related rules exists for each category (for a full list categorization, which must be done manually of rules, see Buber and Kraler, 2000) Interpretation clearly separated from coding Interactive generation and display of causal net Keywords with coded causal relations function as nodes, graphs based on the coding of causal relations the coded causal relations define the link and the direction (links) and the coding of relevancy (type of of the relationships; selection of keywords based on the nodes); all interviews are thrown together research questions and the number of underlying text units No No predefined rules; good practice to define some higherorder categories and assess patterns based on the research questions

Support by software

Notes: This basic analysis procedure followed here differs from the full GABEK cycle proposed by Buber and Kraler (2000); Several options and additional features of WinRelan are not mentioned

(b) Interpretation of causal nets No

2. Data processing and interpretation (a) Display of causal nets High

Degree of automation

Step of analysis

How to promote cooperation

933

Table II. The performed process of analysis in WinRelan and the underlying rules (based on Buber and Kraler, 2000)

IJCHM 24,6

934

Figure 1. Influence variables on cooperation according to the respondents (all relationships are supported by at least three text units)

Using the causal net graph in Figure 1, we can make an in-depth examination of the respondents’ opinions regarding the preconditions and driving factors of cooperation (the keywords taken from the figure are italicized). As a first step of our data interpretation we reordered the obtained set of influence variables and grouped them manually into four subsets: (1) inhibitors of cooperation; (2) psychological and cultural promoters; (3) procedural promoters; and (4) organizational promoters of cooperation. For Group 1, the inhibitors of cooperation, the perceived inhibiting effect is graphically underlined by the circular end of the connecting line. We may observe that the respondents evaluated certain attitudes (such as envy, egoism, rivalry, or parochialism) as being detrimental to cooperation in their destination. The perceived negative relationship between the size of the cooperative space and the ease of

cooperation is also illustrated by Figure 1: according to the respondents, the greater the distance (spatial, socio-cultural, etc) between the single actors, the more difficult it is to establish functioning cooperation. However, the prevalence of oval-shaped text boxes (which stand for constraints and basic conditions) among the inhibitors suggests that little immediate influence potential exists in this group. The latter affirmation holds also for Group 2, the psychological and cultural promoters, which highlight the importance of the cultural congruence as well as the specific character of the single leaders. Regarding this area, variables available for active intervention (graphically represented by a hexagonal text box) could be the sensitization efforts and the definition of shared objectives. A different picture emerges if we focus on Group 3, the procedural promoters. This group is dominated by measures (hexagonal text boxes), and seems therefore to be the area where cooperation supporting action may be undertaken. Among the respondents’ suggestions grouped in this subset figure the involvement and the representation of everybody who is potentially affected by the outcomes in the respective decision-making process (“stakeholder involvement”). This leads to the insight that board composition of tourism organizations appears to be a critical issue. However, according to the interviewees, besides the procedural design of the decision-making process, Group 3 highlights that the opportunities for such cooperative decision-making should also be created. Occasions may include, for example, organizing events, conducting projects, forming (inclusive) workgroups, or promoting farm holidays. Group 3 also draws our attention to the perceived cooperative effect of artificially created proximity between actors. Proximity, as we have already seen, is suggested as a promoter of cooperation in general terms, and can be established by having shared offices or creating personal unions. The latter refers to a single person holding multiple positions among collective actors, for example, a member of the municipal administration who is simultaneously president of the local DMO. The fourth subset of influence variables in Figure 1, the organizational promoters, reveals that some collective actors have a positive influence on the cooperation level in the case study destination. The three cooperation-promoting organizations mentioned in Figure 1 qualify as tourism organizations (DMOs) in the broadest sense of the concept. The results outlined so far lead to a first proposition: P1.

Cooperation in destinations could be enhanced by paying special attention to the procedural and organizational dimensions of cooperation. These areas may be where tangible interventions to improve cooperation can be most easily effected.

Organizational promoters An in-depth look at one of the three DMOs, the Winepoint Kaltern, reveals some very interesting insights regarding how a DMO should be constructed to foster cooperation in the destination. The Winepoint Kaltern can be classified as a tourism organization in the wider sense. As both the umbrella organization of the local wine industry and simultaneously closely integrated with the local tourism association, its scope is to promote cooperation between the tourism and wine industries. Figure 2 shows the perceived effects caused by the Winepoint Kaltern. As already seen in Figure 1, the

How to promote cooperation

935

IJCHM 24,6

936

Winepoint Kaltern is evaluated as promoting cooperation. Yet because the majority of its other effects are themselves also related to cooperation, Figure 2 can be interpreted as showing how and why the organization is perceived to positively influence cooperation. Again, these causes are grouped into subsets: efficiency, stakeholder involvement, and evidence of performance. Efficiency was explicitly mentioned by the respondents, and also supported by the nomination of reference person; that is, someone who speaks for the organization with a competent voice and is entitled to negotiate binding contracts. In addition, respondents mentioned that Winepoint Kaltern promotes conversations and the participation of all relevant persons – traits both subsumable under the term “stakeholder involvement”. Finally, the DMO is viewed as standing for success obtained in the past, and has therefore assumed the status of a brand; that is, a trusted symbol that signals quality. P2.

Figure 2. How Winepoint Kaltern influences cooperation

Regarding the organizational dimension, cooperation in destinations could be enhanced by establishing network management organizations (e.g. DMOs), which may foster cooperation among businesses.

Procedural promoters and GABEK as a decision-making tool Finally, Figure 3 illustrates an example of how GABEK might support the practical implementation process by expanding the causal net to include more concrete measures. To construct the figure, two arbitrarily chosen cooperation-influencing variables were taken from Figure 1; namely, personal union and comprehension (of the importance of cooperation), and the causal net was extended to catch their most important influencing factors. Thus, the fusion of organizations can be detected as a possible measure to establish personal unions. However, as the illustration shows, various side effects of merging DMOs should be considered before putting the measure into practice. Even if a fusion would, in the respondents’ perceptions, reduce administration costs, some fears remain that this may also bring in its wake increased

How to promote cooperation

937

Figure 3. Larger causal net, based on cooperation and its two influence variables comprehension and personal union

IJCHM 24,6

938

distance from single businesses, and thereby reduce their involvement. Furthermore, in a larger unit, this degree of consideration of diversity and peculiarity of each locality or firm cannot occur. Summing up, a measure may cause collateral damage, even though at first glance it seems unproblematic. Equally fruitful is a deeper examination of the comprehension case in the lower part of Figure 3; because, as the rectangle shape shows, it is classified as a basic value, whose defining condition is the difficulty in influencing it directly. This implies that, for any intervention ambitions, a search for more indirect actions is necessary. Thereby, for example, it can be seen that inter-municipal projects such as hiking maps or bike paths are perceived to indirectly promote cooperation by making its benefits clear. The same holds for farm holidays, or other structurally interdependent business forms such as small-scale enterprises, because they highlight the necessity of cooperation in order to survive competition. Furthermore, comprehension of the importance of cooperation can, according to the respondents, also be affected by education, formation and consultation. In contrast, hierarchical governance is seen as having repressive effect on this comprehension. P3.

Regarding the procedural dimension, cooperation in destinations could be enhanced by applying GABEK-like destination governance tools that facilitate the management of collective knowledge and the finding of pragmatic, commonly shared solutions for specific problems.

Discussion Promoters of cooperation and underlying dimensions Our results suggest that cooperation is facilitated by specific basic conditions (e.g. certain psychological dispositions of actors), which, unfortunately, are difficult to influence directly. Nonetheless, as set out in P1, apparently considerable opportunities remain to enact tangible measures in the areas of procedural and organizational design of cooperation. By overlooking these potential measures, two fundamental dimensions emerge: (1) legitimacy; and (2) efficiency. Legitimacy relates to social acceptance obtained by stakeholder involvement in cooperative decision-making; efficiency relates to the perceived cost-benefit ratio of cooperative outcomes. The proposed measure of personal unions (see Figure 1) suggests an inverse relationship between the two dimensions. The potentially positive effect of personal unions on cooperation is theoretically explainable using the concept of transaction costs (Beritelli et al., 2007). Personal unions – the same as shared offices and increased proximity in general – reduce transaction costs in the cooperation process, and tend to increase its efficiency (Williamson, 1979). Nonetheless, personal unions have the disadvantage of power concentration and may lead to diminished social acceptance or legitimacy. To sum up, personal unions highlight the trade-off between transactions costs and power distribution, which is often analogous to those between hierarchy and its opposite, heterarchy, as well as those between efficiency and legitimacy. Viewed from a pragmatic, normative standpoint, a compromise reconciling these dimensions is likely to provide a good basis for successful cooperation.

On the basis of this discussion, another proposition may be added: P4.

Cooperation in destinations could be enhanced if the two dimensions of efficiency and legitimacy and hierarchy and stakeholder involvement are balanced.

Organizational promoters As outlined in P2, a clue to the successful management of the dilemma may lie in the organizational promoters of cooperation (see Figure 1). Note that organizations can be regarded as institutionalized procedural rules. Consequently, by guaranteeing endurance to endorsed procedures, such organizations tend to reduce the individual risks of engaging in cooperation. However, as Figure 2 (and partly also Figure 3) show, in order to be perceived as promoting cooperation, according to the respondents, tourism organization should fulfill three conditions: work efficiently, involve all stakeholders and provide some evidence of performance. The simultaneous presence of all three elements is sometimes subsumed under a stakeholder-oriented concept of good corporate governance (Pechlaner et al., 2011). Furthermore, the three conditions coincide largely with the two above-posited dimensions of efficiency and legitimacy; managing tourism organizations in a manner corresponding to the principles of good corporate governance could be a means to balance the dimensions of efficiency and legitimacy, and thus may promote cooperation in the destination. Procedural promoters and GABEK as a decision-making tool Finally, P3 conveys our argument that GABEK can be another means to promote good cooperation. Recall that the procedural promoters category primarily calls for stakeholder involvement (Figure 1). Given that stakeholder involvement is normally manifested in the form of commonly shared decisions, collective knowledge management tools and decision-making tools (or “destination governance tools”) become crucial (Breukel and Frank, 2009). Figure 3 shows that GABEK can generate outcomes that are so rooted in practice as to justify its application as a destination governance tool. The following characteristics permit the actual research methodology of GABEK to glean such valuable insights from a practitioner point of view. First, the practical relevance is due to GABEK’s ability to code practitioner-relevant information explicitly (e.g. improvement suggestions in form of goals and measures). More generally, GABEK uses a democratic approach to knowledge generation by assuming that problems are best understood and resolved by ordering existing collective knowledge and extracting its most relevant parts, as well as by assigning more importance to more frequently named phenomena. These traits lead to the situation where neither the problem definitions nor suggested solutions generated by GABEK ever leave the practitioners’ realm of experience and comprehension, “which makes them easily acceptable for those affected” (Zelger, 2000). 6. Conclusions and implications This paper advances the hospitality literature with respect to the promotion of local and regional cooperation. This research focus is highly relevant, because in spite of the recognized benefits of cooperation, the willingness of a single hospitality business to cooperate is anything but sure in practice. Given that the hospitality literature mostly

How to promote cooperation

939

IJCHM 24,6

940

neglects this issue, we proposed using the destination governance framework as a partial remedy. This framework advocates the importance of the institutional, structural and procedural conditions for promoting local and regional cooperation. While this consideration is in line with some mainstream approaches in the research on interorganizational relationships (Williamson, 1979; Gulati, 1998), other scholars suggested strategic interdependence to be a much more important driver of cooperation (Pfeffer and Salancik, 1978; Casciaro and Piskorski, 2005). Due to a lack of related empirical studies, the dispute regarding the degree to which each of these types of conditions influences local and regional cooperation has remained open as far as hospitality is concerned. That prompted us to propose an inductive qualitative study to develop a set of propositions concerning the promotion of interorganizational cooperation in hospitality industry. The four propositions that could be identified have valuable implications for research as well as for practical application. First, the study enabled us to develop a conceptual framework that divides the cooperation promoting environment into three groups: psychological and cultural promoters, procedural promoters, and organizational promoters. Second, the study’s results give some indication that these groups differ especially in their accessibility for direct intervention initiatives: Tangible interventions may be most easily effected in the context of the procedural and organizational promoters of cooperation – an important implication from an implementation perspective. Third, our findings add to hospitality research by proposing that the promotion of regional cooperation can be operationalized via strengthening legitimacy and efficiency among these procedural and organizational promoters. With respect to the latter, this aim may be achieved by establishing network management organizations (e.g. DMOs) that operate in congruence with the concept of good corporate governance. With respect to the procedural promoters, cooperation in destinations could be enhanced by applying GABEK-like destination governance tools that facilitate the management of collective knowledge and the finding of commonly shared solutions. However, the major implication refers to the discussion about the respective roles of structural and procedural conditions on the one hand, and strategic interdependence on the other hand, in the context of promoting local and regional cooperation in hospitality. By comparing the inductively developed propositions against the comprehensive list of cooperation-promoting conditions that Oliver (1990) suggests, a good match with respect to the more structural and procedural-oriented conditions becomes apparent: Efficiency and legitimacy are prominent among the procedural promoters in our findings, and what Oliver refers to as stability is covered by the organizational promoters. In contrast, necessity, asymmetry and reciprocity, which point to the relevance of strategic interdependence, are largely absent from our results. Necessity and asymmetry are related to a coercive concept of cooperation and enforced exchange, and thus are understandably less relevant to the voluntary forms dominant in tourist destinations. More surprisingly, the reciprocity dimension, that is, cooperation based on equal exchange of scarce resources, is barely registered in the findings. This may be partly due to our focus on cooperation within destinations, where establishing reciprocity is less challenging than it is in international cooperation. The absence of reciprocity from our findings may be also related to the specific situation in the case study destination, where the congruency of the participants was not a major issue.

Therefore, when it comes to promoting cooperation in local and regional networks of hospitality businesses, the present study supports the position held by the theory on destination governance, which asserts that the institutional, structural and procedural framework is critical to promoting cooperation in destinations – relatively more critical than strategic interdependence. Although resource dependency and even resource exchange are common in such localized contexts, the majority of possibilities does not materialize as cooperative behavior, and strategic interdependence may consequently be considered a necessary but not sufficient condition for engaging in local and regional interorganizational cooperation in hospitality (see also Gulati, 1998). Rather, the risk of opportunistic behavior seems to be perceived as the major obstacle to cooperation, and therefore the measures that aid mitigating it – such as appropriately designed structures and processes – appear particularly valuable for its promotion. As to methodological conclusions, the advantages of using GABEK over the other qualitative approaches mentioned in the introduction can be summarized as follows: GABEK provides a rigorous and traceable data analysis procedure, and thereby mitigates the focus on the data collection phase in ethnography and phenomenology. GABEK is usable in mixed-method approaches (Tashakkori and Teddlie, 2003), because it is compatible with the positivist or post-positivist traditions of quantitative research. Moreover, it facilitates interpretation and presentation of qualitative data through its innovative output in the form of graphs of keywords. Finally, GABEK is ideally suited for use in applied research – or even in real-life situations – that demand an application of knowledge management tools. In sum, the research undertaken provides some insights on how researchers interested in practice-oriented and mixed-method techniques could profit from GABEK because of its ability to display practitioner-relevant results both parsimoniously and comprehensively, which can ultimately facilitate the development of systematic and testable propositions. Limitations and future research Although this study suggests employing the listed four propositions as guiding principles for promoting local and regional cooperation in hospitality in practice, we recognize that they are based on research conducted in only one case study destination. The modest number of interviewees may further limit the transferability (external validity) of the conclusions drawn. Nonetheless, a limited analytical generalization seems feasible (Polit and Beck, 2010). The research approach undertaken has further limitations. In particular, the conclusions may be valid only for more community-based destination models. Thus, a special need exists for further research to crosscheck the results in destinations managed by business corporations. Moreover, the traceability and rigidness of the GABEK procedure may have introduced some losses on the side of flexibility and richness of data. However, in our view, the gains in robustness, parsimony and generalizability compensate for these weaknesses. References Altinay, L. (2006), “Selecting partners in an international franchise organisation”, International Journal of Hospitality Management, Vol. 25, pp. 108-28.

How to promote cooperation

941

IJCHM 24,6

942

Babiak, K. (2007), “Determinants of interorganizational relationships: the case of a Canadian nonprofit sport organization”, Journal of Sport Management, Vol. 21 No. 3, pp. 338-76. Beritelli, P., Bieger, T. and Laesser, C. (2007), “Destination governance: using corporate governance theories as a foundation for effective destination management”, Journal of Travel Research, Vol. 46, pp. 96-107. Brass, D.J., Galaskiewicz, J., Greve, H.R. and Tsai, W. (2004), “Taking stock of networks and organizations: a multilevel perspective”, Academy of Management Journal, Vol. 47 No. 6, pp. 795-817. Breukel, A. and Frank, M.G. (2009), “Knowledge-based network participation in destination and event marketing: a hospitality scenario analysis perspective”, Tourism Management, Vol. 30, pp. 184-93. Bruner, M.E. (1995), “The ethnographer/tourist in Indonesia”, in Lanfant, M.F., Allcock, J.B. and Bruner, E.M. (Eds), International Tourism: Identity and Change, Sage, London, pp. 225-41. Buber, R. (2004), “The job image of young sales assistants: introspection and young consumers’ assessment”, in Gadner, J., Buber, R. and Richards, L. (Eds), Organising Knowledge: Methods and Case Studies, Palgrave Macmillan, Houndmills, Basingstoke, pp. 153-67. Buber, R. and Kraler, C. (2000), “How GABEK and WinRelan support qualitative research”, in Buber, R. and Zelger, J. (Eds), GABEK 2. Zur qualitativen Forschung. On qualitative research, Studien-Verlag, Innsbruck, Wien, Mu¨nchen, pp. 111-37. Burt, R.S. (1992), Structural Holes: The Social Structure of Competition, Harvard University Press, Cambridge, MA. Candela, G. and Figini, P. (2010), “Destination unknown: is there any economics beyond tourism areas?”, Review of Economic Analysis, Vol. 2, pp. 256-71. Casciaro, T. and Piskorski, M.J. (2005), “Power imbalance, mutual dependence, and constraint absorption: a closer look at resource dependence theory”, Administrative Science Quarterly, Vol. 50 No. 2, pp. 167-99. Chathoth, P.K. and Olsen, M.D. (2003), “Strategic alliances: a hospitality industry perspective”, International Journal of Hospitality Management, Vol. 22, pp. 419-34. Chathoth, P.K., Mak, B., Sim, J., Jauhari, V. and Manaktola, K. (2011), “Assessing dimensions of organizational trust across cultures: a comparative analysis of US and Indian full service hotels”, International Journal of Hospitality Management, Vol. 30 No. 2, pp. 233-42. Cohen, E. (1979), “A phenomenology of tourist experiences”, Sociology, Vol. 13, pp. 179-201. Crotts, J.C., Coppage, C.M.A. and Andibo, A. (2001), “Trust-commitment model of buyer-supplier relationships”, Journal of Hospitality and Tourism Research, Vol. 25, pp. 195-208. d’Angella, F. and Go, F.M. (2009), “Tale of two cities’ collaborative tourism marketing: towards a theory of destination stakeholder assessment”, Tourism Management, Vol. 30, pp. 429-40. Dredge, D. (2006), “Policy networks and the local organisation of tourism”, Tourism Management, Vol. 27, pp. 269-80. Ebers, M. (1997), “Explaining interorganizational network formation”, in Ebers, M. (Ed.), The Formation of Interorganizational Networks, Oxford University Press, Oxford, pp. 3-40. Eisenhardt, K.M. and Graebner, M.E. (2007), “Theory building from cases: opportunities and challenges”, Academy of Management Journal, Vol. 50 No. 1, pp. 25-32. Fyall, A. and Garrod, B. (2004), Tourism Marketing: A Collaborative Approach, Channel View Publications, Cleveland. Galaskiewicz, J. (1985), “Interorganizational relations”, Annual Review of Sociology, Vol. 11, pp. 281-304.

Gephart, R. (2004), “Qualitative research and the Academy of Management Journal”, Academy of Management Journal, Vol. 47 No. 4, pp. 454-62. Glaser, B.G. and Strauss, A.L. (1967), The Discovery of Grounded Theory: Strategies for Qualitative Research, Aldine, Chicago, IL. Gulati, R. (1998), “Alliances and networks”, Strategic Management Journal, Vol. 19 No. 4, pp. 293-317. Higher Education Funding Council for England HEFCE (1998), Review of Hospitality Management, HEFCE, London. Hu, C. and Racherla, P. (2008), “Visual representation of knowledge networks: a social network analysis of hospitality research domain”, International Journal of Hospitality Management, Vol. 27, pp. 302-12. Hu, M.L.M., Horng, J.S. and Sun, Y.H.C. (2009), “Hospitality teams: knowledge sharing and service innovation performance”, Tourism Management, Vol. 30, pp. 41-50. Husserl, E. (1962), The Crisis of European Sciences and Transcendental Phenomenology, Northwestern University Press, Evanston, IL. Lee, K. (2008), “Issues for international franchising: lessons from the case of a Poland-based restaurant operator”, Cornell Hospitality Quarterly, Vol. 49, pp. 454-7. Lemmetyinen, A. and Go, F.M. (2009), “The key capabilities required for managing tourism business networks”, Tourism Management, Vol. 30, pp. 31-40. McCabe, S. (2007), “The beauty in the form: ethnomethodology and tourism studies”, in Ateljevic, I., Pritchard, A. and Morgan, N. (Eds), The Critical Turn in Tourism Studies: Innovative Research Methodologies, Elsevier, Oxford, pp. 227-43. Martin, D. and Woodside, A. (2012), “Structure and process modeling of seemingly unstructured leisure-travel decisions and behavior”, International Journal of Contemporary Hospitality Management, Vol. 24 No. 6, pp. 855-72. Mehmetoglu, M. and Altinay, L. (2006), “Examination of grounded theory analysis with an application to hospitality research”, International Journal of Hospitality Management, Vol. 25, pp. 12-33. Mueller, C. and Raich, M. (2005), “The ambiguous relationship of leadership and intellectual capital: understanding how intellectual capital is developed”, The Electronic Journal of Knowledge Management, Vol. 3 No. 1, pp. 35-44. Nordin, S. and Svensson, B. (2007), “Innovative destination governance: the Swedish ski resort of ˆ re”, Entrepreneurship and Innovation, Vol. 8 No. 1, pp. 53-66. A Novelli, M., Schmitz, B. and Spencer, T. (2006), “Networks, clusters and innovation in tourism: a UK experience”, Tourism Management, Vol. 27, pp. 1141-52. Oliver, C. (1990), “Determinants of interorganizational relationships: integration and future directions”, Academy of Management Review, Vol. 15 No. 2, pp. 241-65. Pask, G. (1976), Conversation Theory, Elsevier, Amsterdam. Pearsall, J. (Ed.) (2001), The Concise Oxford Dictionary, 10th ed., Oxford University Press, Oxford. Pechlaner, H. and Tschurtschenthaler, P. (2003), “Tourism policy, tourism organisations and change management in Alpine regions and destinations: a European perspective”, Current Issues in Tourism, Vol. 6 No. 6, pp. 508-39. Pechlaner, H., Raich, F. and Kofink, L. (2011), “Elements of corporate governance in tourism organizations”, Tourismos: An International Multidisciplinary Journal of Tourism, Vol. 6 No. 3, pp. 57-76.

How to promote cooperation

943

IJCHM 24,6

944

Pechlaner, H., Volgger, M. and Herntrei, M. (2012), “Destination management organizations as interface between destination governance and corporate governance”, Anatolia: An International Journal of Tourism and Hospitality Research. Pfeffer, J. and Salancik, G.R. (1978), The External Control of Organizations, Harper and Row, New York, NY. Polit, D.F. and Beck, C.T. (2010), “Generalization in quantitative and qualitative research: myths and strategies”, International Journal of Nursing Studies, Vol. 47, pp. 1451-8. Pratt, M.G. (2009), “For the lack of a boilerplate: tips on writing up (and reviewing) qualitative research”, Academy of Management Journal, Vol. 52 No. 5, pp. 856-62. Raich, F. (2006), Governance ra¨umlicher Wettbewerbseinheiten. Ein Ansatz fu¨r die Tourismus-Destination, DUV, Wiesbaden. Raich, M. (2008), “Basic values and objectives regarding money: implications for the management of customer relationships”, International Journal of Bank Marketing, Vol. 26 No. 1, pp. 25-41. Reid, L.J., Smith, S.L.J. and McCloskey, R. (2008), “The effectiveness of regional marketing alliances: a case study of the Atlantic Canada Tourism Partnership 2000-2006”, Tourism Management, Vol. 29, pp. 581-93. Rhodes, R.A.W. (1996), “The new governance: governing without government”, Political Studies, Vol. 44, pp. 652-67. Stierand, M.B. and Do¨rfler, V. (2012), “Reflecting on a phenomenological study of creativity and innovation in haute cuisine”, International Journal of Contemporary Hospitality Management, Vol. 24 No. 6, pp. 946-57. Stumpf, C. (1939), Erkenntnislehre: Band 1, Johann Ambrosius Barth, Leipzig. Svensson, B., Nordin, S. and Flagestad, A. (2005), “A governance perspective on destination development – exploring partnerships, clusters and innovation systems”, Tourism Review, Vol. 60 No. 2, pp. 32-7. Tashakkori, A. and Teddlie, C. (Eds) (2003), Handbook of Mixed Methods in Social and Behavioral Research, Sage, Thousand Oaks, CA. Timur, S. and Getz, D. (2008), “A network perspective on managing stakeholders for sustainable urban tourism”, International Journal of Contemporary Hospitality Management, Vol. 20 No. 4, pp. 445-61. Tinsley, R. and Lynch, P. (2001), “Small tourism business networks and destination development”, International Journal of Hospitality Management, Vol. 20, pp. 367-78. Wang, Y. (2008), “Collaborative destination marketing: understanding the dynamic process”, Journal of Travel Research, Vol. 47 No. 2, pp. 151-66. Wang, Y. and Fesenmaier, D.R. (2007), “Collaborative tourism marketing: a case study of Elkhart County, Indiana”, Tourism Management, Vol. 28 No. 3, pp. 863-75. Wang, Y. and Krakover, S. (2008), “Destination marketing: competition, cooperation or coopetition?”, International Journal of Contemporary Hospitality Management, Vol. 20 No. 2, pp. 126-41. Wang, Y. and Xiang, Z. (2007), “Toward a theoretical framework of collaborative destination marketing”, Journal of Travel Research, Vol. 46, pp. 75-85. Wang, L. and Zajac, E.J. (2007), “Alliance or acquisition? A dyadic perspective on interfirm resource combinations”, Strategic Management Journal, Vol. 28 No. 13, pp. 1291-317. Williamson, O.E. (1979), “Transaction-cost economics: the governance of contractual relations”, Journal of Law and Economics, Vol. 22 No. 2, pp. 233-61.

Williamson, O.E. (1999), “Strategy research: governance and competence perspectives”, Strategic Management Journal, Vol. 20, pp. 1087-108. Xiao, Q., O’Neill, J.W. and Wang, H. (2008), “International hotel development: a study of potential franchisees in China”, International Journal of Hospitality Management, Vol. 27, pp. 325-36. Yang, J.-T. (2010), “Antecedents and consequences of knowledge sharing in international tourist hotels”, International Journal of Hospitality Management, Vol. 29, pp. 42-52. Zelger, J. (1999), “Wissensorganisation durch sprachliche Gestaltbildung im qualitativen Verfahren GABEK”, in Zelger, J. and Mair, M. (Eds), GABEK. Verarbeitung und Darstellung von Wissen, Studien-Verlag, Innsbruck, Wien, Mu¨nchen, pp. 41-87. Zelger, J. (2000), “Twelve steps of GABEKWinRelan: a procedure for qualitative opinion research, knowledge organization and systems development”, in Buber, R. and Zelger, J. (Eds), GABEK 2. Zur qualitativen Forschung. On Qualitative Research, Studien-Verlag, Innsbruck, Wien, Mu¨nchen, pp. 205-20. Corresponding author Harald Pechlaner can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

How to promote cooperation

945

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0959-6119.htm

IJCHM 24,6

946 Received 8 September 2011 Revised 14 December 2011 Accepted 21 December 2011

RESEARCH IN BRIEF

Reflecting on a phenomenological study of creativity and innovation in haute cuisine Marc B. Stierand Academy of Hotel and Facility Management, NHTV Breda University of Applied Sciences, Breda, The Netherlands, and

Viktor Do¨rfler Department of Management Science, Strathclyde Business School, University of Strathclyde, Glasgow, UK Abstract Purpose – This paper aims to present and reflect on a phenomenological research process used to elucidate the nature of creativity and innovation in haute cuisine. Design/methodology/approach – In-depth unstructured interviews and field notes capturing subjective experiences were employed to elucidate the experiences of 18 top chefs from the UK, Spain, France, Austria and Germany with regards to creativity and innovation. Findings – The findings are twofold: first, an empirical sample finding is presented in order to contextualize the type of findings obtained; second, key methodological findings are presented explaining the process of elucidating the nature of creativity and innovation through iterative learning from the descriptions of the interviewees and the subjective experiences gathered. Research limitations/implications – The underlying phenomenological study is limited to male haute cuisine chefs in five European countries. Future research is planned including female and male chefs from other countries in order to learn whether similar empirical findings can be obtained. Practical implications – The paper presents a research process for elucidating cognitive and nebulous phenomena such as creativity and innovation to make them accessible to managers, researchers, students and policy makers. Originality/value – The findings explain the process of elucidating the nature of creativity and innovation through iterative learning from the descriptions of the interviewees and the subjective experiences gathered. Further conceptual and methodological development emerges from investigating interviewees representative of the notion of the extraordinary. Keywords Phenomenology, Interviews, Chefs, Creativity, Innovation, Research methods, Cooking, Creative thinking, Western Europe Paper type Research paper

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 pp. 946-957 q Emerald Group Publishing Limited 0959-6119 DOI 10.1108/09596111211247254

Introduction The purpose of this paper is to report and reflect on a phenomenological research process elucidating the nature of creativity and innovation in haute cuisine. Unstructured in-depth interviews and field notes capturing the subjective experiences of the interviewer were employed to elucidate the experiences of 18 top chefs from the UK, Spain, France, Austria and Germany. We argue that our approach helps avoiding some of the difficulties associated with more traditional approaches and sheds light on

phenomena that otherwise may remain under-explored. In management and business research there is a tendency of applying micro-perspectives when studying creativity and innovation, which has deprived understanding of the phenomenology of creativity and innovation. The reason for this tendency may well be that both phenomena are cognitive and nebulous concepts and therefore are difficult to access by more traditional and dominant research approaches. Phenomenology is not such a dominant approach in management and business research, but it has a long-standing tradition in psychology aiming for elucidation of complex cognitive and nebulous phenomena that may not be accessible by more objective and quantitative research approaches. The context of haute cuisine was chosen for three reasons. First, haute cuisine is one of a few sectors where expert judgement about creative and innovative performance is formalized and publicly available through institutions like Michelin’s Guide Rouge or the Gault Millau restaurant guide. Second, haute cuisine restaurants have attracted little systematic research (Lane, 2010). And third, the interviewer has professional experience as chef in the haute cuisine sector, which facilitated elucidation of the interviewees’ experiences and contributes to an understanding of the nature of creativity and innovation in the context of haute cuisine. Furthermore, the interviewees were chosen, because they can be defined as being representative of “the extraordinary”, a notion derived from recognized studies by Maslow (1970, 1971), Gardner (1993), Csı´kszentmiha´lyi (1997) and Nakamura et al. (2009). These authors define the extraordinary as an individual who is self-actualizing and able to destroy or fundamentally alter existing structures within a domain or create a new domain. In order to alter existing domain structures or create a new domain, creativity and innovation is required and therefore the underlying study focused only on extraordinary chefs. There is, however, a degree of subjectivity involved in selecting who counts as extraordinary and thus the interviewees had to be listed in Michelin’s Guide Rouge and/or the Gault Millau guide, because these two restaurant guides are considered to be the two formal authorities for identifying chefs at the apex of culinary practice and creativity. 1. Philosophical framing The philosophical underpinnings of the underlying study are derived from the ideas of Edmund Husserl’s philosophical phenomenology. Phenomenology is not only the study of what people experience but also the study of the ways in which these experiences can be understood and how these experiences develop a worldview. This means that phenomenology is both a philosophy and a research approach. Lived experience, life-world and intentionality Finlay (2009) stresses that essential to all phenomenological research is a rich description of lived experience, which can be understood through the concept of the life-world being the context of experiences. A person’s lived experience of a phenomenon is only understandable in the context of this person’s life-world, because in this world individual perception and experience are altered by others, who also bring a wider range of perspectives through their opinions and experiences (Husserl, 1936). Husserl stressed that a person must engage in a dialogue with the world in order to grasp the subjective dimensions of experiences, which are vital to come as close to the experience as possible (see Laverty, 2003). Smith (2007) adds to this that these

Reflecting on a study of creativity 947

IJCHM 24,6

948

subjective dimensions can only be accessed through introspection (i.e. self-observation) and therefore can only be described in subjective terms. Then, in order to grasp lived experiences they have to be contextualised within the life-world. This notion is called intentionality, which Husserl inherited from his mentor Franz Brentano, who explained it as “reference to a content, the directedness toward an object or the immanent object quality” (cited in Spiegelberg, 1971, p. 39). Bracketing pre-understandings Being a chef interviewing other chefs means by definition that one cannot refrain from pre-understandings and therefore one must try to bracket these pre-understandings. The process of bracketing is often misunderstood as the need to be unbiased and objective, but instead it is a process that can be seen as a “dialectic movement between bracketing preunderstandings and exploiting them reflexively as a source of insight” (Finlay, 2009, p. 13). This effect can be explained by the phenomenological concept of qualia, which is the subjective component of people’s experiences (see, e.g. Jackson, 1982). Qualia cannot be transferred to others, only experienced subjectively. However, if two people (e.g. two chefs) have experienced qualia of the “same” phenomenon, they can talk about these qualia. So, being an interviewee with a chef background enabled the discussion of the qualia of culinary creativity and innovation, immensely enriching the data of the study. Investigating the extraordinary As aforementioned, the interviewees were chosen, because they can be described as being representative of “the extraordinary” as defined in recognized studies by Maslow (1970, 1971), Gardner (1993), Csı´kszentmiha´lyi (1997) and Nakamura et al. (2009). The authors have discussed elsewhere several reasons for investigating the extraordinary (see Do¨rfler and Stierand, 2009). The main reason, though, is that the extraordinary seems to be more representative of the phenomenon (e.g. creativity and innovation) than of the sample or population (e.g. chefs). Hence, the underlying study is based on the premise that there are extraordinary chefs and examining those helps to generate a better understanding of the cognitive and nebulous phenomena of creativity and innovation in haute cuisine, because these chefs are representative of the creativity and innovation in haute cuisine. This means that creativity and innovation in themselves are “extraordinary phenomena”, because they are the result of “extraordinary actions”. Essentially, this premise follows a logic outlined by Maslow (Maslow, 1971, p. 7): If we want to know how fast a human being can run, then it is no use to average out the speed of a “good sample” of the population; it is far better to collect Olympic gold medal winners and see how well they can do.

Investigating extraordinary people has the effect that the knowledge that can be gained from the direct and descriptive research findings is likely to be of a more transferrable quality and therefore easier to use for building theory, which numerous authors claim is as vital for qualitative as it is for quantitative research (e.g. Campbell and Stanley, 1966, Gibbert, 2006). Description and interpretation One of the most widely used phenomenological methods is that of Giorgi (e.g. 1985, 1994), who in the 1970s developed his conception of descriptive phenomenology in the

field of psychology by building on Husserl’s ideas. Giorgi aims at the conscious essences of phenomena and their essential structures and his method is said to be particularly suitable for analyzing text such as transcribed interviews. However, one beneficial yet complicating factor of the underlying study is that the interviewer has professional experience as chef in the haute cuisine sector, which facilitated elucidation of the interviewees’ experiences, but required rigorous reflection. Hence, the philosophical stance of this research is interpretivist using both description and interpretation (see Tsoukas, 1989) and thus can be seen as an extension of Giorgi’s original method. Giorgi’s descriptive phenomenology was chosen over other phenomenological variants, because the research question of the underlying study is ‘What is the lived experience of creativity and innovation in haute cuisine?’ and Giorgi’s approach is suitable for these types of research questions. Table I provides a comparison between some of the key phenomenological variants (see Finlay, 2008).

Reflecting on a study of creativity 949

2. Research process Interviewee identification and selection As aforementioned, interviewees were identified based on their mentioning in the Michelin and/or the Gault Millau guides, because these two restaurant guides are Variant

Typical research question

Aim

Descriptive

What is the lived experience of creativity and innovation in haute cuisine?

To identify the essential or general structures underlying the phenomena of creativity and innovation in haute cuisine

Heuristic

What is my experience of creativity and innovation in haute cuisine?

To produce a composite description and creative synthesis of the experience of creativity and innovation in haute cuisine

Lifeworld

To focus on existential themes such as What is the lifeworld of one who is engaged with creativity and innovation the person’s sense of self-identity and embodied relations with others when in haute cuisine? experiencing creativity and innovation in haute cuisine

IPA

What is the individual experience of creativity and innovation in haute cuisine?

Critical narrative

What story or stories does a person tell To produce a narrative (perhaps from just one person) and to show how the of their experience of creativity and narrative was co-created in the innovation in haute cuisine? research context

Relational

What is it like to be a creator and innovator in haute cuisine?

Source: Based on Finlay (2008)

To capture individual variations between co-researchers. Thematic analysis would involve some explicit interpretation on the part of both coresearcher and researcher

To focus on the co-researchers’ selfidentity and “creative adjustment”

Table I. Comparison of phenomenological variants

IJCHM 24,6

950

considered to be the two formal authorities for identifying chefs at the apex of culinary practice and creativity. The interviewees were then selected based on a number of other data sources: interviewer’s personal experience as chef in the haute cuisine sector, articles from the trade press, websites of chefs, cookery books, and an interview with the German Chef Harald Wohlfahrt (www.traube-tonbach.de; 3 * Michelin). This range of data facilitated the identification and justification of chefs who have experience with the phenomena under investigation and would be useful for the purpose of the research. The interview with Chef Wohlfahrt offered rich insight, but also helped to encourage other chefs to participate, because he signed a letter of support. Originally written in German, this letter was then translated into French and English and sent via email first to chefs he had personally recommended. Because of Wohlfahrt’s strong reputation in the sector, the responses were mainly positive. After these positive answers, additional emails were sent out now also listing the names of the new participants, which again encouraged more chefs to participate. In total 35 chefs from the UK, France, Spain, Austria and Germany were contacted of which nine did not reply, seven refused, and 19 agreed to participate. One chef, unfortunately, had to cancel the interview appointment because of TV commitments. This meant that at the end a total of 18 chefs was interviewed comprising two chefs from the UK, four chefs from France, three chefs from Spain, two chefs from Austria and seven chefs from Germany. This selection does not suggest any professional or personal preferences, but is merely the result of a convenience sampling strategy. Table II lists all participating chefs, names of their restaurants and locations. Data collection The data collection part of the research process (see Table III) has been conducted through unstructured interviews (Van Enk, 2009, p. 2) and relied only on some pre-planned topics, because the aim was to achieve at an emergent dialogue that could help elucidate and better understand the interviewees’ experiences with the phenomena (see Kwortnik, 2003, Melissen and Stierand, 2010). This interviewing style was possible due to the interviewer’s previous chef experience, which eased the building of trust and in-depth conversations between professionals. With hindsight this style would not have been so rewarding without this previous experience, which became particularly apparent when discussing complex situations, because of the natural use of metaphors, specialist terms of the profession, and examples from practice. However, this previous experience required the interviewer to rigorously self-reflect after each interview. This process of self-reflection was maintained by critical debates within the research team and by actively practising the process of bracketing through listening and looking at the interviews as openly as possible (see Giorgi, 1994). These two practises helped distinguishing between those parts of the pre-understanding that might blur and those parts that help developing a deeper insight and understanding of the interviewees’ accounts (see Finlay, 2009). Data analysis The interviews were first analyzed using Giorgi’s (1994, 1985) descriptive phenomenological approach. Later, a second layer of analysis was applied offering a

Chef

Michelin/Gault Millau ranking (at time of research)

Restaurant

UK Fergus Henderson Raymond Blanc

1 */ – 2 */ –

St John’s, London Le Manoir aux Quat’Saisons, Great Milton

France Jean-Georges Klein Michel Troisgros Michel Bras Se´bastien Bras

3 */18 3 */19 3 */19 3 */19

L’Arnsbourg, Baerenthal Maison Troisgros, Roanne Bras, Laguiole Bras, Laguiole

Spain Andoni Luis Aduriz Joan Roca Ferran Adria`

2 */ – 2 */ – 3 */ –

Mugaritz, Errenteria El Celler de Can Roca, Girona El Bulli, Roses

Austria Heinz Reitbauer Roland Trettl

2 */19 1 */18

Steirereck, Vienna Ikarus im Hangar-7, Salzburg

Germany Harald Wohlfahrt Dieter Mu¨ller Nils Henkel

3 */19 3 */19 3 */19

Heinz Winkler Hans Haas Joachim Wissler Juan Amador

3 */19 2 */18 3 */19 3 */17

Schwarzwaldstube, Baiersbronn Dieter Mu¨ller, Bergish Gladbach Dieter Mu¨ller (now Gourmetrestaurant Lerbach), Bergish Gladbach Venezianisches Restaurant, Aschau Tantris, Munich Vendoˆme, Bergisch Gladbach Amador, Langen

more interpretive account of what was learned about the nature of culinary creativity and innovation from analyzing the interviewees’ accounts. The reason for adding a second layer of analysis started intuitively, because it seemed that the interview data and the subjective experiences gathered during the field trip could offer a much richer and broader understanding of the nature of creativity and innovation. This richer and broader understanding can be called the meta-level of the findings, which may be defined as a pattern laying beyond the descriptive findings, or as the essence, structure, loosely coupled associations, or specific implications of the descriptive findings. By using this rather unusual two-layer analysis it became clear that the ongoing debate about whether a phenomenological study should be descriptive or more interpretive should change into a debate about how the analysis can be made more transparent so that others can recover how the findings emerged and why they form a convincing story. Following this train of thought, the first layer of analysis was descriptive and was particularly important for the classification of meaning units from the raw interview data and the generation of themes. In turn, these first-layer themes themselves form a framework that was used as a starting point for the second layer of analysis that was interpretative explaining what was learned about the nature of culinary creativity and innovation from analyzing the interviewees’ accounts. The entire research process is summarized in Table III.

Reflecting on a study of creativity 951

Table II. List of interviewees

IJCHM 24,6

Interviewee identification and selection Research methods employed

952

Table III. Summary of the research process

Identification based on mentoring in Michelin/ Gault Millau Selection using purposive and convenience sampling based on: † Interviewer’s chef background † Trade press articles, chefs’ websites, cookery books, etc. † Interview with chef Harald Wohlfahrt ( ! signed letter of support)

Data collection

Data analysis

Unstructured in-depth interviews and research notes capturing the subjective experience of the interviewer Only some pre-planned topics Aim was to achieve an emergent dialogue Interviewer’s chef background helped to create trust and enhanced understanding Interviewer’s chef background required enhanced need for selfreflection and bracketing of preunderstandings

Transcription of digitally recorded interviews First layer of analysis using Giorgi’s descriptive phenomenological approach to identify meaning units and themes in each interview followed by an identification of general themes considering all interviews together Second layer of analysis was the interviewer’s interpretation of what he learned about the nature of creativity and innovation in haute cuisine from analyzing the interviewees’ accounts and from taking the subjective experiences (captured in the research notes) into consideration

3. A sample finding: the concept of harmony In this section only the sample finding of “harmony” is presented to provide better contextual understanding for the remainder of the paper. As aforementioned, the first layer of analysis was descriptive and was particularly important for the classification of meaning units from the raw interview data and the generation of themes. Figure 1 shows those themes that relate to the concept of harmony as described by the chefs. Most of the chefs talked about harmony but in a variety of ways using metaphoric descriptions often with reference to the arts. We gathered the chefs’ descriptions about harmony and organized them in themes as illustrated in Figure 1. However, there was unavoidably a certain degree of subjectivity involved because the descriptions emerged from the interaction between interviewer and interviewee and the themes were organized first by the interviewer and then discussed with the research team. Hence, we used the notion of bracketing described above as good as possible. These first-layer themes, in turn, were used as a starting point for the second layer of analysis that was interpretative in nature. The reason why this second layer of analysis was interpretative can be understood by looking at the terms highlighted in bold in Figure 1. These terms became “search terms” when we re-visited the literature we used before, and when we reviewed new literature aiming to better understand the essence of the concept of harmony. From this second literature review we learned that harmony is typically mentioned in the field of music or architectural aesthetics and is often associated with beauty because “the beautiful is that which pleases universally without a concept” (Immanuel

Reflecting on a study of creativity 953

Figure 1. Themes related to the concept of “Harmony”

Kant cited in Naini et al., 2006, p. 277). Furthermore, the concept of proportion emerged as essential feature of harmony in the literature on art. Leonardo da Vinci, for example, stated that proportion is the ratio between the individual parts and the whole (Naini et al., 2006). The proportion of the so-called golden ratio can be found, for instance, in the paintings of Leonardo da Vinci and Piet Mondrian, in the Notre Dame Cathedral and in the Sagrada Familia. Also composers like Be´la Barto´k and Fre´de´ric Chopin have evidently used the power of the golden ratio by setting the climax of their compositions accordingly (Gross and Miller, 1997). The reason we have found in the literature for using the golden ration is that it “allows the minor element to occupy a portion of the whole that makes it maximally striking” without making any of the parts excessive (Berlyne, 1971, p. 232). This idiographic explanation is very much in line with the descriptive findings about the need for a harmonious composition and contrast of ingredients in a dish in terms of aesthetics, smell, taste, and even between the dish and the whole ambient of the restaurant in which it is served, including the weather outside, the music, etc. 4. Discussion and concluding reflections While we refer in our paper to a study elucidating the nature of creativity and innovation in haute cuisine, it is interesting to ask whether the research process presented here may also be useful for studying creativity and innovation in other

IJCHM 24,6

954

Figure 2. Research note example and elucidating meaning during interview

contexts and for studying other cognitive and nebulous phenomena. Thus it is important to define what “type of findings” can be achieved by following such a research process. Unlike dominant, more structured, and often quantitative approaches, the epistemological outcomes of a phenomenological study cannot be presented following the positivist ideal of objectivity, reliability and validity. It is, however, possible to reveal the trustworthiness of phenomenological findings. Wolcott (1996, p. 136), for example, says that while his work is not guided by validity, he is in search of critical themes and aims to write credible interpretations, which reflect his learning and understanding. In our study we applied a two-layer analysis starting with an idiographic description of the interviewees’ accounts and then progressing towards an idiographic explanation that emerged through the researchers’ iterative learning from both the interviewees’ accounts and the subjective experiences gathered. It is merely impossible to visualise such a cognitive and nebulous learning process, but maybe the following example of a research note (see Figure 2) including the subjective experiences of the interviewer and an example of a drawing negotiating meaning during one of the interviews can help to at least imagine how the data for the underlying study was gathered and analysed. On the left-hand side is an example of a research note following the template of Schatzman and Strauss (1973, pp. 95-107). The observational and conceptual notes, in particular, were helpful in re-experiencing the interview and provided little pieces of evidence supporting first the process of descriptive analysis and later the interpretive explanation. The right-hand side is a typical example of an attempt to derive meaning at the descriptive level. Based on this kind of meaning negotiation the meta-levels of

the findings were easier to derive, because the interviewees naturally revealed more of their lived experience and life-world through this negotiating with another chef. This makes it at least plausible that the research process on hand may also be useful for studying other cognitive and nebulous phenomena. Hence, a phenomenological study can offer alternative criteria than reliability or validity: It can offer an “Aha!” moment when presenting the essence of a phenomenon and when both researcher and reader have a sense of being “gripped by the phenomenon understood” in the way it is presented (Crotty, 1996, p. 169). This is also confirmed by Hayllar and Griffin (2005) by saying that this very moment resembles Buytendijk’s “phenomenological nod”, which Van Manen (1990, p. 27) describes as the moment when we can nod to a description, because it is “collected by lived experience and recollects lived experience.” As we hope we have demonstrated, a research approach like ours does not have to be less rigorous than the more widely accepted ones. In fact, our underlying intention is to challenge both the research and practice communities to critically re-think their entrenched fear of such approaches, because they are often considered as soft, non-transparent, beyond comprehension, against scientific conventions, or just not handy (Denzin and Lincoln, 2000; Lipshitz et al., 2001; Moore et al., 2011). There are, however, a number of limitations to the underlying study that need to be addressed. The study is limited to male haute cuisine chefs in five European countries, but future research is planned including female and male chefs from other countries in order to learn whether similar empirical findings can be obtained. There are also plans to continue this research in other contexts, such as scientific knowledge, music, painting, etc to investigate whether the findings can really be transferred to other contexts. Thus, this paper is also an invitation for other researchers to test and comment on the research approach presented here and to discuss their experiences whether their investigations are in the same or any other context and whether they are studying the same or any other cognitive and nebulous phenomena. References Berlyne, D. (1971), Aesthetics and Psychology, Appleton, Century and Crofts, New York, NY. Campbell, D. and Stanley, J. (1966), Experimental and Quasi-experimental Designs for Research, Houghton Mifflin, Boston, MA. Crotty, M. (1996), Phenomenology and Nursing Research, Churchill Livingstone, Melbourne. Csı´kszentmiha´lyi, M. (1997), Creativity: Flow and the Psychology of Discovery and Invention, HarperCollins, New York, NY. Denzin, N. and Lincoln, Y. (2000), “Introduction: the discipline and practice of qualitative research”, in Denzin, N. and Lincoln, Y. (Eds), Handbook of Qualitative Research, Sage, Thousand Oaks, CA. Do¨rfler, V. and Stierand, M. (2009), “Investigating the extraordinary”, paper presented at British Academy of Management Annual Conference, Brighton, 15-17 September. Finlay, L. (2008), “Introducing phenomenological research”, available at: www.lindafinlay.co.uk/ phenomenology.htm Finlay, L. (2009), “Debating phenomenological research methods”, Phenomenology & Practice, Vol. 3, pp. 6-25. Gardner, H. (1993), Creating Minds, Basic Books, New York, NY.

Reflecting on a study of creativity 955

IJCHM 24,6

956

Gibbert, M. (2006), “Generalizing about uniqueness: an essay on an apparent paradox in the resource-based view”, Journal of Management Inquiry, Vol. 15, pp. 124-34. Giorgi, A. (1985), “Sketch of a psychological phenomenological method”, in Giorgi, A. (Ed.), Phenomenology and Psychological Research, Duquesne University Press, Pittsburgh, PA. Giorgi, A. (1994), “A phenomenological perspective on certain qualitative research methods”, Journal of Phenomenological Psychology, Vol. 25, pp. 190-220. Gross, S. and Miller, N. (1997), “The ‘golden section’ and bias in perceptions of social consensus”, Personality and Social Psychology Review, Vol. 1, pp. 241-71. Hayllar, B. and Griffin, T. (2005), “The precinct experience: a phenomenological approach”, Tourism Management, Vol. 26, pp. 517-28. Husserl, E. (1936), Crisis of European Sciences and Transcendental Phenomenology, Northwestern University Press, Evanston, IL. Jackson, F. (1982), “Epiphenomenal qualia”, The Philosophical Quarterly, Vol. 32, pp. 127-36. Kwortnik, R. (2003), “Clarifying ‘fuzzy’ hospitality-management problems with depth interviews and qualitative analysis”, Cornell Hotel and Restaurant Administration Quarterly, pp. 117-29. Lane, C. (2010), “The Michelin-starred restaurant sector as a cultural industry: a cross-national comparison of restaurants in the UK and Germany”, Food, Culture and Society, Vol. 13, pp. 493-519. Laverty, S. (2003), “Hermeneutic phenomenology and phenomenology: a comparison of historical and methodological considerations”, International Journal of Qualitative Methods, Vol. 2, pp. 1-29. Lipshitz, R., Klein, G., Orasanu, J. and Salas, E. (2001), “Taking stock of naturalistic decision-making”, Journal of Behavioural Decision-making, Vol. 14, pp. 331-52. Maslow, A.H. (1970), Religions, Values, and Peak Experiences, Penguin, New York, NY. Maslow, A. (1971), The Farther Reaches of Human Nature, Viking, New York, NY. Melissen, F. and Stierand, M. (2010), “Strolling into organisations of the future”, paper presented at European Group for Organizational Studies Colloquium, Lisbon, 28 June – 3 July. Moore, K., Smallman, C., Wilson, J. and Simmons, D. (2011), “Dynamic in-destination decision-making: an adjustment model”, Tourism Management, Vol. 33, pp. 635-45. Naini, F., Moss, J. and Gillc, D. (2006), “The enigma of facial beauty: esthetics, proportions, deformity, and controversy”, American Journal of Orthodontics and Dentofacial Orthopedics, Vol. 130, pp. 277-82. Nakamura, J., Shernoff, D. and Hooker, C. (2009), Good Mentoring: Fostering Excellent Practice in Higher Education, Jossey-Bass, San Francisco, CA. Schatzman, L. and Strauss, A. (1973), Field Research: Strategies for a Natural Sociology, Prentice Hall, Englewood Cliffs, NJ. Smith, D. (2007), Husserl, Routledge, London, New York, NY. Spiegelberg, H. (1971), The Phenomenological Movement: A Historical Introduction, Martinus Nijhoff, The Hague. Tsoukas, H. (1989), “The validity of idiographic research explanations”, Academy of Management Review, Vol. 14, pp. 551-61. Van Enk, A. (2009), “The shaping effects of the conversational interview: an examination using Bakhtin’s theory of genre”, Qualitative Inquiry, Vol. 15, pp. 1265-86.

Van Manen, M. (1990), Researching Lived Experience: Human Science for an Action Research Pedagogy, State University of New York Press, Albany, NY. Wolcott, H. (1990), “On seeking – and rejecting – validity in qualitative research”, in Eisner, E. and Peshkin, A. (Eds), Qualitative Inquiry in Eduation: The Continuing Debate, Teachers College Press, New York, NY.

Reflecting on a study of creativity

About the authors Dr Marc B. Stierand is Principal Lecturer in Strategy and Innovation at the NHTV Breda University of Applied Sciences. His research interests include creativity, innovation and cognition in organizational contexts. Marc B. Stierand is the corresponding author and can be contacted at: [email protected] Dr Viktor Do¨rfler is Lecturer and Director of the Management Development Program at the Strathclyde Business School. His research interests include personal knowledge and knowledge increase in organizational contexts.

957

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

IJCHM 24,6

958

Book review Research Themes for Tourism Edited by Peter Robinson, Sine Heitmann and Peter Dieke CABI Wallingford, UK 2011 ISBN: 9781845936846 £75.00/$145.00/e105.00 (hardback) 320 pp. Review DOI 10.1108/09596111211247263 This book provides an overview of a number of major themes on tourism studies nowadays, allowing readers to become familiar with a wide range of topics. Twenty chapters in this book and 20 issues with some common characteristics: they are all contemporary and under-researched (some of them still “emerging”), with good potential for generating new knowledge and for inspiring both academics and practitioners to do more. Of course the list of issues is not exhaustive – that would have been a Herculean task: a limited number of them had to be chosen, and the editors have undoubtedly made an interesting selection, although others could have been added. That would be an endless and not very fruitful discussion, although in such a young academic discipline (or indiscipline, as argued by Tribe and others) it is not easy to find themes that are over-researched. Nevertheless, I would classify the book’s chapters into two different types of content: (1) Some of them are transversal in character, addressing issues of general interest; they include the first seven chapters (mass tourism vs niche tourism, tourism development, tourist behavior and motivation, authenticity, semiotics, sustainability, and community-led tourism, plus the slow movement in food, cities and tourism, tackled in chapter nine). (2) Others focus on the analysis of diverse niches of the tourism industry: ranging from those of rural or cultural character to those based on events, festivals and arts, sports and adventures, heritage, films, gastronomy, and historical or religious sites; some niches are even based on particular types of tourist, such as the LGBT grouping (lesbians, gays, bisexuals and transgenders), and others on particular motivations for travel, such as health or medical treatments, and on a particular means of transportation, such as cruise vessels.

International Journal of Contemporary Hospitality Management Vol. 24 No. 6, 2012 pp. 958-960 q Emerald Group Publishing Limited 0959-6119

At the same time, there are evident interactions between some of the chapters: for instance, a type of tourism centered on arts, heritage or gastronomy is considered as cultural tourism; in other words, those are sub-segments of cultural tourism. The criteria used to select the chapters, and their sequence, are not apparent or clear enough to me. As a suggestion, its structure could have been organized in several blocks, according to the different nature of the topics under study.

In particular, I would have like to see some more chapters devoted to transversal research topics, such as knowledge management and innovation, strategy and competitiveness, or human resources management, among other possibilities, because tourism firms present clear and interesting peculiarities in these fields. With such an approach, the book’s structure would have been more balanced. However, the title does suggest to readers what they will find inside: a number of interesting research themes in contemporary tourism. Besides, given the layout and orientation provided to the chapters, with a mix of theory (with the corresponding literature reviews) and practice (including examples), the editors have tried to target two different audiences: researchers and professionals. The merit of this attempt is well worth recognition: it is always a difficult challenge to be addressed. The ultimate and inevitable question in these cases is whether both audiences will finally be satisfied. This work tends to fall between differing needs: researchers may feel some disappointment, given a certain lack of deeper insight into theoretical and, above all, methodological issues, that generally command most interest from this group; for their part, professionals would maybe like to see more concrete answers to their day-to-day concerns, and a selection of proven “best practices” that would make their work easier. But part of our job as academics is to present not only answers, but also questions for further discussion, new research projects and future advances. So, I would say that the prime target market for this book is the academic community, in a double sense: (1) The contents provide valuable working material for PhD courses and students; the latter will find a good starting point with regard to literature reviews and guidelines for future research in any of the 20 topics under study. This should make it much easier for them to identify the seminal publications and the gaps in existing research for their own work and future original contributions. (2) The book also constitutes valuable working material for teaching purposes, given the many examples (I do not consider case study to be the proper term for the real experiences encapsulated in the various chapters; case studies present real problems which demand decisions and possible solutions) and review questions included. From this material students, on both undergraduate and postgraduate courses, can see how the theoretical underpinnings are, or are not, reflected in real situations. Therefore, assuming that the main objective of this book is to present the state of the art of a number of current issues in the field of tourism, with both rigorous theoretical foundations and practical implications, the final outcome is satisfactory. However, this book cannot constitute an optimal approach for the two main target populations: researchers and practitioners. Moreover, the selection of topics and the sequence of presentation are debatable. This book is complementary to others more tightly focused on research methodologies: quantitative and qualitative, descriptive, explanatory or predictive, and underlying the interdisciplinary and cross-cultural profile of tourism as a field of study. Some of these readily come to mind: Tourism Research (by G. Jennings, 2010), “Planning Research in Hospitality & Tourism” (by L. Altinay and A. Paraskevas, 2008), Research Methods for Leisure & Tourism: A Practical Guide (by A.J. Veal, 2006), Tourism Research Methods (by B.W, Ritchie, P.M. Burns and C.A. Palmer, 2005).

Book review

959

IJCHM 24,6

960

Finally, I would definitely recommend this book to those with a research interest: students, faculties and managers. It is a valuable and relevant collection, to be complemented with other works, depending on the type of reader: academics (who will still need to consult books with a stronger orientation to research methodologies and, eventually, other works that address the general or specific topics of interest), and practitioners (who need works with more detailed managerial implications). Alfonso Vargas-Sa´nchez GEIDETUR, University of Huelva, Huelva, Spain