Organizational Climate for Creativity: Exploring the Influence of Distinct Types of Individual Differences [1st ed.] 978-3-658-25240-3, 978-3-658-25241-0

Table of contents :
Front Matter ....Pages I-XV
Introduction (Christian Hoßbach)....Pages 1-3
Theoretical Foundations (Christian Hoßbach)....Pages 5-18
Method (Christian Hoßbach)....Pages 19-28
Results (Christian Hoßbach)....Pages 29-52
Discussion (Christian Hoßbach)....Pages 53-62
Conclusion (Christian Hoßbach)....Pages 63-64
Back Matter ....Pages 65-156

Citation preview

Christian Hoßbach

Organizational Climate for Creativity Exploring the Influence of Distinct Types of Individual Differences

BestMasters

Mit „BestMasters“ zeichnet Springer die besten Masterarbeiten aus, die an renom­ mierten Hochschulen in Deutschland, Österreich und der Schweiz entstanden sind. Die mit Höchstnote ausgezeichneten Arbeiten wurden durch Gutachter zur Veröf­ fentlichung empfohlen und behandeln aktuelle Themen aus unterschiedlichen Fachgebieten der Naturwissenschaften, Psychologie, Technik und Wirtschaftswis­ senschaften. Die Reihe wendet sich an Praktiker und Wissenschaftler gleichermaßen und soll insbesondere auch Nachwuchswissenschaftlern Orientierung geben. Springer awards “BestMasters” to the best master’s theses which have been completed at renowned Universities in Germany, Austria, and Switzerland. The studies received highest marks and were recommended for publication by supervisors. They address current issues from various fields of research in natural sciences, psychology, technology, and economics. The series addresses practitioners as well as scientists and, in particular, offers guidance for early stage researchers.

Weitere Bände in der Reihe http://www.springer.com/series/13198

Christian Hoßbach

Organizational Climate for Creativity Exploring the Influence of Distinct Types of Individual Differences

Christian Hoßbach Halle, Germany

ISSN 2625-3577 ISSN 2625-3615  (electronic) BestMasters ISBN 978-3-658-25241-0  (eBook) ISBN 978-3-658-25240-3 https://doi.org/10.1007/978-3-658-25241-0 Library of Congress Control Number: 2019930704 Springer Gabler © Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2019 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer Gabler imprint is published by the registered company Springer Fachmedien Wiesbaden GmbH part of Springer Nature The registered company address is: Abraham-Lincoln-Str. 46, 65189 Wiesbaden, Germany

Acknowledgements Many people and organizations have influenced and contributed to my work on this thesis. I would like to thank Prof. Dr. Anne-Katrin Neyer for providing valuable feedback. Her support, openness to new ideas, and network contributed to the success of this thesis. In addition to this, the opportunities that she created for me did not only impact my thesis, but also enabled me to continue to work within my domain of interest. I would also like to thank The Creative Problem Solving Group (CPSB) for supporting this research initiative and providing access to the instruments that I have used within this study. Especially Sarah from CPSB invested a lot of time and energy to help me collecting and preparing the data. Furthermore, this work would not have been possible without all the participants who invested their time and energy to give invaluable insights into how they actualize their creative potential and the critical work situations that are related to this. As the data was collected in different languages, I would like to thank Sabrina, Marjorie, and Maxim for their valuable help with the translations of the raw data. Writing this thesis in English posed some challenges to me and I am thankful for the help provided by all proofreaders, especially Sabrina and Scott, to overcome these challenges. I would also like to thank Luc, who first triggered my interest in the field of creativity and especially I would like to thank Scott for being a mentor and friend. The invaluable opportunities and experiences that he has created for me, had a significant impact on my life and also inspired me in working on this thesis. Above all, I would like to thank you, Sophia, for investing a tremendous amount of time and energy to support me during the entire writing process. You helped me to overcome blockades (sometimes even before I knew that I had one) and structured my thoughts. When I was working late at night you stayed awake to support me and, in the end, you helped me to deliver the thesis on time by helping me with all the formatting. You also compensated the time that I could not invest in other parts of our family life and, even though you endured this for almost four months, you still wanted to become my wife shortly after I submitted this thesis. Your love is the central part of my life and a constant source of energy for me. Christian Hoßbach

Content 1

2

Introduction............................................................................................................ 1 1.1

Problem Statement ......................................................................................... 1

1.2

Objective and Research Questions ................................................................ 1

1.3

Structure of the Thesis.................................................................................... 3

Theoretical Foundations ........................................................................................ 5 2.1

Organizational Climate for Creativity .............................................................. 5

2.1.1

Defining Organizational Climate............................................................... 5

2.1.2

Climate as an Intervening Variable .......................................................... 6

2.1.3

Dimensions of Creative Climate ............................................................... 8

2.2

Linking Individual Differences and Creative Climate ...................................... 9

2.2.1

3

Problem-Solving Style ............................................................................ 10

2.2.1.1

Conceptual Foundations.................................................................. 10

2.2.1.2

Dimensions ...................................................................................... 12

2.2.1.3

Impact on Perceptions of Creative Climate ..................................... 13

2.2.2

Gender ................................................................................................... 16

2.2.3

Other Sources of Individual Differences................................................. 17

2.2.3.1

Age .................................................................................................. 17

2.2.3.2

Life Phase........................................................................................ 18

2.2.3.3

Creative Self-Efficacy ...................................................................... 18

Method ................................................................................................................ 19 3.1

Research Design .......................................................................................... 19

3.2

Sampling ....................................................................................................... 20

3.3

Data Collection ............................................................................................. 21

3.4

Measures ...................................................................................................... 21

3.4.1

Problem-Solving Style ............................................................................ 21

3.4.2

Climate for Creativity .............................................................................. 23

3.5

Data Analysis ................................................................................................ 23

3.5.1

Quantitative Data Analysis ..................................................................... 24

3.5.2

Qualitative Data Analysis ....................................................................... 26

VIII

4

Content

Results ................................................................................................................ 29 4.1

Quantitative Results...................................................................................... 29

4.1.1

Sample ................................................................................................... 29

4.1.2

Descriptive Statistics .............................................................................. 29

4.1.2.1

Individual Differences ...................................................................... 30

4.1.2.2

Climate for Creativity ....................................................................... 32

4.1.3

Differences between Best- and Worst-Case Climate Perceptions ......... 33

4.1.4

Impact of Individual Differences on Climate Perceptions ....................... 34

4.1.4.1

Exploring Correlations ..................................................................... 34

4.1.4.2

Comparison of Climate Scores between different Groups .............. 38

4.1.4.3

Identifying the best Predictor Variables ........................................... 41

4.1.5 4.2

Qualitative Results ........................................................................................ 43

4.2.1

Conceptual Space .................................................................................. 43

4.2.2

Impact of Distinct Types of Individual Differences ................................. 44

4.2.2.1

Gender Differences ......................................................................... 45

4.2.2.2

Problem-Solving Style Differences .................................................. 46

4.2.3 5

Summary of Qualitative Results ............................................................. 51

Discussion ........................................................................................................... 53 5.1

6

Summary of Quantitative Results........................................................... 42

Interpretation of Results................................................................................ 53

5.1.1

OC: Different Conditions for Incremental and Radical Creativity ........... 53

5.1.2

MP: Preferred Levels of Interaction........................................................ 56

5.1.3

WD: Preferred Levels of Trust in Personal Relationships ...................... 57

5.1.4

Gender: Interaction of Surface and Deeper Level Diversity ................... 58

5.2

Implications for Practice................................................................................ 60

5.3

Implications for Future Research .................................................................. 61

5.4

Limitations..................................................................................................... 62

Conclusion........................................................................................................... 63

References................................................................................................................. 65 Appendix .................................................................................................................... 75

List of Figures Figure 1: A climate-centric Model for Organizational Creativity ................................... 7 Figure 2: A Model of Problem-Solving Style .............................................................. 11 Figure 3: Most salient positive and negative Forces impacting Creativity ................. 44

List of Tables Table 1: Definitions of Creative Climate Dimensions ................................................... 9 Table 2: Descriptive Statistics for VIEW Dimensions................................................. 31 Table 3: Descriptive Statistics for Best- and Worst-Case Climate Perceptions ......... 33 Table 4: Differences between Perceptions of Best- and Worst-Case Climates ......... 34 Table 5: Correlations with Best-Case Climate Perceptions ....................................... 35 Table 6: Correlations with Worst-Case Climate Perceptions ..................................... 37 Table 7: Correlations with Differences in Climate Perceptions .................................. 38 Table 8: Impact of OC and Subscales on Climate Perceptions ................................. 39 Table 9: Summary of Multiple Regression Results .................................................... 41 Table 10: Summary of Quantitative Results .............................................................. 42

List of Abbreviations Abbreviation

Meaning

ANOVA

Analysis of Variance

CCQ

Creative Climate Questionnaire

CHI

Challenge and Involvement

CON

Conflict

CPS

Creative Problem Solving

CPSB

The Creative Problem Solving Group Inc.

CSE

Creative Self-Efficacy

DEB

Debate

FRE

Freedom

IDS

Idea Support

IDT

Idea Time

KAI

Kirton Adaption-Innovation Inventory

LP

Life Phase

MANOVA

Multivariate Analysis of Variance

MP

Manner of Processing

NV

Novelty

OC

Orientation to Change

P-E

Person-Environment

PLH

Playfulness and Humor

RIT

Risk-Taking

SA

Structure and Authority

SD

Standard Deviation

XIV

List of Abbreviations

SOQ

Situational Outlook Questionnaire

SPSS

Statistical Package for the Social Sciences

SS

Search Strategy

TRO

Trust and Openness

VIEW

VIEW: An Assessment of Problem Solving Style

VUCA

Volatile, Uncertain, Complex and Ambiguous

WD

Ways of Deciding

WinRelan

Windows Relation Analysis

List of Symbols Symbol

Meaning

rwg

Coefficient for Interrater Agreement

R2

Coefficient of Determination

dz

Cohen`s standardized Effect Size of Difference Scores

a

Cronbach’s Coefficient Alpha

df

Degrees of Freedom

h2

Eta-squared

N

Frequency

p

Level of Significance

X

Mean

D

Mean Difference between Best- and Worst-Case Situation

b

Multiple Regression Coefficient

F

Value of F-Test

t

Value of t-test

L

Wilk’s Lambda

1 Introduction Nowadays, organizations have to cope with volatile, uncertain, complex and ambiguous (VUCA) environments (Bennett & Lemoine, 2014). Globalization and international interconnectedness made the marketplace more and more complex, competition harder, product life cycles shorter and the pace of the global economy faster (Isaksen & Tidd, 2006). This leads to a growing need for innovation and change for organizations. Most scholars agree that creativity, in its basic sense, refers to the generation of novel and useful ideas, whereas innovation refers to the implementation of creative ideas (Amabile, 1988; Woodman, Sawyer, & Griffin, 1993). In this perspective, creativity is the prerequisite of any innovation, or, as Isaksen and Tidd (2006) frame it more practically: “You can have creativity without innovation, but you can’t have innovation without creativity.” (p. 266). Therefore, developing and nurturing creativity is a key challenge for organizations. Based on the notion that contextual factors within the work environment can promote or inhibit creativity (Amabile, Conti, Coon, Lazenby, & Herron, 1996; Ekvall, 1996), hiring people with creative talent is not sufficient for organizations, because people’s willingness to use their creative potential depends on the presence of enabling conditions within their work environment (Mumford & Gustafson, 1988). One such condition within the work environment is the climate for creativity on which this thesis is focusing. 1.1 Problem Statement A lot of research has been done on the work environment (e.g. Amabile, 1988; Amabile et al., 1996; Ford, 1996; Oldham & Cummings, 1996), climate (Anderson & West, 1998; Ekvall, 1996; Isaksen, Lauer, Ekvall, & Britz, 2001; Scott & Bruce, 1994) and culture (Drazin, Glynn, & Kazanjian, 1999; Hurley & Hult, 1998; Martins & Terblanche, 2003) that is conducive to creativity. However, most of this research has not been done from a person-environment (P-E) fit perspective. Yet, there is evidence for an interaction of individual and contextual factors promoting creativity (Oldham & Cummings, 1996; Scott & Bruce, 1994) and that a fit between the actual and preferred work environment for creativity is associated with a higher creative performance (Choi, 2004; Puccio, Talbot, & Joniak, 2000). Further investigation is required to determine which factors in the work environment promote or inhibit creativity depending on individual differences. A better understanding of this interaction would allow organizations to take a more individualized approach in creating a work environment conducive to creativity and to make a more productive use of diversity. 1.2 Objective and Research Questions This thesis seeks to explore the influence of distinct types of individual differences on the perception of organizational climates for creativity. As there are many contextual variables impacting creativity (Isaksen, 2017), this thesis will focus on the climate for creativity as one particular aspect within the broader work environment. The focus shall © Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2019 C. Hoßbach, Organizational Climate for Creativity, BestMasters, https://doi.org/10.1007/978-3-658-25241-0_1

2

Introduction

be on organizational climate as it is more visible and more amenable to change than organizational culture (Ehrhart, Schneider, & Macey, 2014; Schneider, Brief, & Guzzo, 1996). Furthermore, there is some empirical evidence that climate functions as an intervening variable between factors in the work environment, i.e. leadership behavior, and creative performance (Ekvall & Ryhammer, 1998; Isaksen & Akkermans, 2011). There are also many different individual factors impacting creative behavior (Treffinger, Selby, & Isaksen, 2008). Based on the notion that even though to different degrees, almost any job requires at least some creativity (Shalley, Gilson, & Blum, 2000) and that everyone has at least some creative ability (Boden, 1994; Richards, 2007), but may prefer to use it in different ways (Isaksen, 2004; Isaksen & Dorval, 1993; Kirton, 1978), this study does not include an independent outcome measure to assess the level of creativity in order to examine what manifestations of climate are associated with the most or least creative outcomes. Rather, it seeks to shed more light upon the kind of climate in which different people perceive that they are most supported or inhibited to actualize their individual creative potential. Therefore, the main focus of this thesis will be on problem-solving style differences, which is related to the manner in which people prefer to express their creativity to solve complex and ambiguous problems (Selby, Treffinger, Isaksen, & Lauer, 2004). Within this area, a line of empirical research explored the impact of problem-solving style on the perceptions of the current climate for creativity. Very few quantitative differences were found, suggesting that there is, if at all, only a weak relationship between these constructs (Isaksen, 2009; Isaksen & Kaufmann, 1990; Isaksen & Lauer, 1999; Scott & Bruce, 1994). However, there is some preliminary evidence that people with different problem-solving styles have different perceptions of creative climate in bestand worst-case working situations (Isaksen & Aerts, 2011). Following this line of research, this thesis is supposed to gain a better understanding of Isaksen and Aerts (2011) preliminary findings and address some of their limitations. In contrast to this prior study, this thesis will take a multi-method approach to assess the climate for creativity from a quantitative and qualitative perspective. Furthermore, this study will also deliberately investigate the impact of gender on the perception of organizational climates for creativity. Baer and Kaufman (2008) concluded that differences in creative productivity were mainly caused by environmental factors, as prior studies on gender differences in creativity did not provide sufficient evidence that there are gender differences in creative ability. Moreover, there is evidence for style-related gender differences, e.g. in risk-taking (Byrnes, Miller, & Schafer, 1999), conflict management (Brewer, Mitchell, & Weber, 2002) or decision-making (Venkatesh, Morris, & Ackerman, 2000), that may be relevant for what environmental conditions men and women perceive to be helpful or hindering for using their creative ability. Gaining a better understanding of this interaction may also add new aspects to the current debate among practitioners on gender intelligence and how to manage gender diversity in organizations (Annis & Merron, 2014).

Structure of the Thesis

3

In this way, this exploratory, descriptive and multi-method study examines the interaction between individual and contextual factors influencing creativity by answering the guiding research question: In what ways do individual differences in problem-solving style and gender impact the perceptions of organizational climates for creativity? In addition to this guiding question, several other questions shall be answered within the course of this thesis. Which other sources of individual differences may exert a potential impact on the perception of creative climate? Which dimensions of creative climate are most impacted by which types of individual differences? Which other factors in the work environment are perceived as most salient positive or negative by different kinds of people and are there some conflicting factors that may be perceived positively and negatively by different people? 1.3 Structure of the Thesis In order to answer these questions, the next chapter will outline the theoretical foundations of the key constructs that are included in this study. Based on this, the constructs will be operationalized and a research model as well as an approach to analyze the data will be developed within the course of the third chapter. The quantitative part of this thesis is aimed at exploring the impact of distinct types of individual differences on the perception of the different dimensions of creative climate, whereas the qualitative part shall then help to put the quantitative findings into a more meaningful context and explore possible reasons why different facets of climate might be perceived differently. The quantitative and qualitative results of this study will be presented in chapter four. This is followed by a discussion of the key findings of this study and their implications for research and practice in chapter five. Finally, the last chapter will draw a short conclusion and provide an outlook for future research in this area.

2 Theoretical Foundations The following chapter will start with defining the key construct of organizational climate for creativity as well as explaining its intervening nature. After this, the operationalization of creative climate that is applied in this study will be described. This is followed by a discussion of the potential impact of distinct types of individual differences on the perception of creative climate, whereby the main focus will be on problem-solving style and gender differences. 2.1 Organizational Climate for Creativity Research on organizational climate has a long tradition and dates back to the early work of Lewin, Lippitt, and White (1939), who first used the term “social climate” (p. 271) to describe the impact of the group atmosphere, that was deliberately created by different leadership styles, on the aggressive behavior of young boys. This early research has shown that the group atmosphere, or climate, has an impact on the group members’ behavior and that it can be deliberately influenced. Over the years, two different approaches for studying organizational climate emerged in the literature: The molar climate approach, which is aimed at understanding how the general context of an organization is perceived by employees, and the focused climate approach, which is aimed at understanding specific aspects of climate in relationship to specific organizational outcomes (Ehrhart et al., 2014). Even though both approaches are still present in current research, the major attention was, and is still, on the focused climate approach (Ehrhart et al., 2014). This is partly a reason of Schneider and Reichers (1983) early critique that research should focus on studying a “climate for something” (p. 21). In their overview of the current state of the research on organizational climate, Ehrhart et al. (2014) describe the connection between these two approaches in the way that the molar climate can also be thought of as a “climate for employee well-being” (p. 82) which forms a basis for more focused climates to emerge. An example of such a focused climate is the climate for creativity which will be the main focus of this thesis. 2.1.1 Defining Organizational Climate There is no single definition of organizational climate that scholars agreed upon. From an ontological perspective, some researchers consider climate in a subjectivist sense and define it from a sense making perspective as the shared meaning that employees attach to workplace characteristics (Ehrhart et al., 2014; Schneider, 1975), whereas other researchers consider climate in an objectivistic sense and define it as typical behavioral patterns that exist independently of individual perceptions (Ekvall, 1996; Forehand & Gilmer, 1964; Isaksen, 2017). However, even though this distinction is important from a theoretical perspective, both lines of research are mostly using the same approach to measure climate by aggregating individual perceptions (Ekvall, 1987). © Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2019 C. Hoßbach, Organizational Climate for Creativity, BestMasters, https://doi.org/10.1007/978-3-658-25241-0_2

6

Theoretical Foundations

Besides this, it is also important to consider the scope of the climate construct. The prevailing paradigm in current research is to see climate and culture as closely related, but yet distinct constructs, and consider the deeper underlying values, beliefs and norms as an antecedent of the more observable behavioral patterns that form the climate of an organization (Ehrhart et al., 2014). This thesis is based on the objectivistic perspective of Ekvall (1996), who defined climate as “(…) a conglomerate of attitudes, feelings and behaviors which characterize life in the organization, and exists independently of the perceptions and understandings of the members of the organization” (Ekvall, 1996: 105), and makes a clear distinction between organizational climate, culture and other factors within the work environment. Ekvall (1983) was also one of the first researchers who deliberately studied organizational climate in relationship to creativity and proposed climate as an intervening variable. 2.1.2 Climate as an Intervening Variable One of the earliest references of the term „climate for creativity“ that can be found in the literature is Cummings (1965), who explored the impact of situational factors on the utilization of creative talent in organizations. However, even though his article has the title “organizational climates for creativity”, he focused more on factors in the broader work environment as he proposed, among others, a flexible and flat organizational structure and a non-controlling leadership style as characteristics of a creativity conducive organization (Cummings, 1965). Another early research study on the climate for creativity was provided by Abbey and Dickson (1983), who studied the work climate in research and development departments using the general climate questionnaire developed by Pritchard and Karasick (1973), and found, among others, that a reward system that recognizes excellent performance was positively related to innovative performance. However, these elements would be considered antecedents of climate in current research (Ehrhart et al., 2014; Isaksen, 2017). Hence, it is important to make some distinctions between climate, its antecedents and outcomes. Agreement can be found among researchers, that organizational creativity, and subsequently innovation, requires a work environment which is conducive to creativity. Ekvall (1996) demonstrated that the climates for creativity measured by the Creative Climate Questionnaire (CCQ), an earlier version of the Situational Outlook Questionnaire (SOQ) which will be used in this study, were, indeed, very different in organizations that were characterized as innovative, stagnated and average based on measures of product innovation. The model in figure 1 is based on the early work of Ekvall (1983, 1996) that has been further developed by Isaksen et al. (2001), and Isaksen (2013, 2017) and illustrates climate as an intervening variable. The basic notion of this model is that factors in the broader work environment, such as leadership behavior, determine the climate, which, in turn, impacts organizational and psychological processes, such as the way people

Organizational Climate for Creativity

7

communicate or solve problems, and subsequently organizational outcomes such as creative performance and well-being (Isaksen, 2017).


Figure 1: A climate-centric Model for Organizational Creativity Source: Isaksen, 2017: 133

Among these various factors, leadership is considered as one of the most important factors because leadership behavior does not only have a strong direct impact on climate, but also an indirect impact through its ability to influence the other antecedents of climate (Isaksen, 2017). In some empirical studies, climate completely (Ekvall & Ryhammer, 1998) or partially (Isaksen & Akkermans, 2011) mediated the relationship between leadership and creative performance. Other studies found moderating effects of creative climate, e.g. in the relationship between transformational leadership and innovative performance (Jung, Chow, & Wu, 2003). A (partially) mediated relationship between leadership and outcomes was also reported in studies focusing on other focused climates such as the climate for service (Chuang & Liao, 2010; Salanova, Agut, & Peiró, 2005; Schneider, Ehrhart, Mayer, Saltz, & Niles-Jolly, 2005), or ethical climate (Mayer, Kuenzi, & Greenbaum, 2010) supporting the intervening nature of the climate construct. Besides the stream of research on creative climate, which is based on the pioneering research of Ekvall (1983), another stream of research gained considerable attention in the literature, which is based on the early work of Amabile (1988) and focuses more on the work environment. This rather broad construct includes elements that are similar to aspects of creative climate, such as freedom or challenging work, but also elements

8

Theoretical Foundations

that are considered antecedents of climate, such as organizational and supervisory encouragement (Amabile et al., 1996). This thesis will follow the first line of research and will be guided by the climate-centric model discussed above for several reasons. As mentioned earlier, there is ample empirical evidence supporting climate as an intervening variable. Moreover, among the various other models in the literature that include climate, or aspects of climate, as a variable (e.g. Amabile et al., 1996; Anderson, De Dreu, & Nijstad, 2004; Woodman et al., 1993), this model is among those that are most current, but it is also well-researched as it dates back to the seminal work of Ekvall (1983). Moreover, as Isaksen (2017) outlined, each element of the model is supported in the creativity literature. In addition, the model demonstrated evidence of its content validity as Isaksen (2013) was able to code more than 5,000 phrases derived from almost 1,000 participants from five organizations concerning elements in their current work environment that helped or hindered their individual creativity against the elements of this model. Finally, as this thesis is aimed at exploring the relationships between individual differences and the climate for creativity, it seeks to make a clear distinction between climate, its antecedents and outcomes. In order to do this, the next section will give an overview of the dimensions that are included in the operationalization of creative climate that is applied in this study. 2.1.3 Dimensions of Creative Climate As there is no single definition of climate that scholars agreed upon, there is also no single set of dimensions of creative climate that can be found in the literature. As outlined above, this thesis is based on the seminal work of Ekvall (1983, 1996), who proposed ten dimensions of creative climate that were based on a review of the literature as well as empirical evidence derived from the factor-analytic structure of the CCQ as his measure of creative climate. Later, the CCQ has been renamed to the SOQ and in the process of its further development continuous empirical support was only found for nine of the original ten dimensions (Isaksen, Lauer, & Ekvall, 1999). Hunter, Bedell, and Mumford (2007) conducted a meta-analytic study and examined the impact of different climate dimensions on creative behavior and found that standardized measures such as the SOQ produced higher effect sizes than other measures of creative climate that were only developed for specific studies. Therefore, the SOQ will be used in its unaltered and validated form as a measure of creative climate in this study. The psychometric properties of the instrument will be discussed in more detail in the next chapter. Definitions of the nine dimensions, that are applied to operationalize the climate for creativity in this study, are provided in table 1.

Linking Individual Differences and Creative Climate

9

Table 1: Definitions of Creative Climate Dimensions Dimension

Definition

Challenge and Involvement (CHI)

The degree of emotional involvement, commitment, and motivation in the operations and goals.

Freedom (FRE)

The level of autonomy, discretion, and initiative in behavior exerted by individuals to acquire information, make decisions etc.

Trust and Openness (TRO)

The degree of emotional safety, and openness found in relationships.

Idea Time (IDT)

The amount of time people can use (and do) for elaborating new ideas.

Playfulness and Humor (PLH)

The display of spontaneity, ease, good natured joking, and laughter.

Conflict (CON)

The presence of personal and emotional tensions or hostilities.

Idea Support (IDS)

The degree to which new ideas and suggestions are attended to and treated in a kindly manner.

Debate (DEB)

The expressing and considering of many different view-points, ideas and experiences.

Risk-Taking (RIT)

The tolerance of ambiguity and uncertainty.

Source: Adapted from Isaksen et al., 1999: 668

Isaksen (2017) demonstrated, that the impact of all of these nine dimensions of creative climate on creative behavior is empirically supported in the creativity literature. However, even though researchers seem to agree that situational factors such as climate influence creative behavior, only a few studies deliberately investigated how the perceptions of these dimensions may vary dependent on individual differences. Therefore, this thesis seeks to shed more light upon this interaction of individual and situational factors impacting creative behavior in organizations and the next section of this chapter will give an overview of the different sources of individual differences that are examined in this study. 2.2 Linking Individual Differences and Creative Climate The notion that individual behavior is influenced by an interaction of individual and situational factors dates back to the beginning of the last century and the early work of Lewin (1938). However, the predominant parts of the creativity literature that focused on gaining a better understanding of individual and contextual factors impacting creative behavior developed separately from each other, lacking to examine their interaction (Isaksen, Puccio, & Treffinger, 1993). Nevertheless, some studies (Oldham & Cummings, 1996; Scott & Bruce, 1994) provided empirical support for the call of other researchers to take a more interactionist approach to creativity research (Harrington, 1990; Isaksen et al., 1993; Mumford & Gustafson, 1988; Woodman & Schoenfeldt, 1990). Implied in this interaction is also the notion of P-E fit which is associated with

10

Theoretical Foundations

positive outcomes such as lower degrees of stress or higher degrees of satisfaction (Edwards, 1996; Pervin, 1968). This thesis also falls into the conceptual domain of P-E fit and seeks to explore the impact of distinct types of individual differences on the perception of organizational climates for creativity. The last part of this chapter focuses on the different sources of individual differences that are examined in this study and how they may relate to the operationalization of creative climate on which this thesis is built. The main focus of this thesis will be on the construct of problem-solving style, which is directly related to creativity (Selby et al., 2004). 2.2.1 Problem-Solving Style Isaksen (1995) described the relationship between creativity and problem-solving in the way that especially ambiguous and complex problems need creativity to be solved as opposed to clearly defined and structured problems. In a broad sense, problemsolving can be considered as overcoming the barriers that emerge in working towards a desired goal or outcome, so that it can be applied in almost any domain (Treffinger et al., 2008). Problem-solving styles are defined as “(…) consistent individual differences in the ways people prefer to plan and carry out generating and focusing activities, in order to gain clarity, produce ideas, and prepare for action.” (Selby et al., 2004: 222). Implied in this definition is also the relationship to the constant alternation of divergent and convergent thinking as essential element of the different phases of the method Creative Problem Solving (CPS) (Isaksen, Dorval, & Treffinger, 2011). 2.2.1.1 Conceptual Foundations As illustrated in figure 2, the model of problem-solving style that is applied in this study, links creativity and problem-solving with individual differences and operationalizes problem-solving style in three independent dimensions called orientation to change (OC), manner of processing (MP), and ways of deciding (WD) with two opposing styles respectively, that are based on the underlying constructs of learning style, psychological type and cognitive style.

Linking Individual Differences and Creative Climate

11


Figure 2: A Model of Problem-Solving Style Source: Isaksen, Kaufmann, and Bakken, 2016: 270

As it is also shown in figure 2, the model is mainly based on research and insights generated within the cognitive styles project, which refers to a line of research initiated at the International Center for Studies in Creativity at the State University College in Buffalo that was aimed at understanding how people approach the creative process differently and was summarized by Isaksen (2004). Building on the findings in different areas, i.e. learning style, cognitive style, and psychological type, VIEW: An Assessment of Problem Solving Style (VIEW) (Selby, Treffinger, & Isaksen, 2002) was developed that integrates these different streams of research and measures the three dimensions of problem-solving style. Given its conceptual soundness and its use in current research (Isaksen et al., 2016; Main, Delcourt, & Treffinger, 2017), this model is applied in this thesis as an operationalization of problem-solving style. An important assumption that is made in the creativity literature suggests that the level of creativity is conceptually independent from the preferred style of creativity, suggesting that people with different style preferences can be equally creative (Isaksen, 2004; Isaksen & Dorval, 1993; Kirton, 1978, 1989; Martinsen & Kaufmann, 1999). Moreover, there is also empirical evidence that measures of style did not correlate with measures of level of creativity supporting this theoretical assumption (Goldsmith, 1987; Houtz & Selby, 2009; Woodel-Johnson, Delcourt, & Treffinger, 2012). However, a recent study suggested that training which involved how people of different problem-solving styles can best collaborate in groups demonstrated a significant impact on problem-solving performance (Main et al., 2017).

12

Theoretical Foundations

In another recent study, Isaksen et al. (2016) examined the relationships between problem-solving style and personality by linking the VIEW styles with the Sixteen Personality Factor Questionnaire and found clear relationships in the expected directions providing support for the assumption that problem-solving style is anchored in personality traits. However, in contrast to criticism that style is a redundant construct as it is almost completely reflected in personality (von Wittich & Antonakis, 2011), they found that personality explained only some of the variance in problem-solving styles providing support for the assumption that problem-solving style is related to personality but nevertheless an independent construct (Isaksen et al., 2016). In the next section of this chapter, the dimensions of problem-solving style and their underlying relationships with personality will be described in more detail. 2.2.1.2 Dimensions The OC dimension describes people’s preferred degrees of novelty as well as their preferred way of dealing with structure and looking for alternatives when they are solving problems (Selby et al., 2004). The explorer style is associated with a tendency to a more radical approach to problem-solving emphasizing the novelty aspect of creativity, considering imposed structure as limiting and preferring to look broadly for different solutions, as opposed to the developer style that is associated with a tendency to a more adaptive approach to problem-solving emphasizing the usefulness aspect of creativity, considering structure and boundaries as enabling and preferring to look narrowly for a small number of promising solutions (Selby et al., 2004). Also, explorers tend to be more open to change, abstract, energetic, socially bold, dominant and outgoing as well as less rule conscious and perfectionist as compared to developers (Isaksen et al., 2016). As the OC dimension is a rather broad construct, it has three subscales called novelty (NV), structure and authority (SA), and search strategy (SS) that are related, but can also be interpreted independently, which allows for a deeper insight into the cognitive strategies that are usually applied in solving problems creatively (Selby, Treffinger, & Isaksen, 2014). Furthermore, this dimension is directly related to the adaption-innovation theory distinguishing between people with a preference for an incremental or radical kind of creativity and problem-solving (Kirton, 1976). Strong correlations were found between the OC dimension of VIEW and the Kirton Adaption-Innovation Inventory (KAI) as an operationalization of the adaption-innovation theory (Treffinger, Isaksen, & Selby, 2014a), suggesting that there is a strong conceptual overlap between both style constructs. The MP dimension concerns the level of interaction that people prefer and how they typically manage information and share their thinking (Selby et al., 2004). People with an internal style prefer an individual approach to problem-solving that allows them to reflect on their ideas first before they share their thinking with others, whereas people with an external style prefer to collaborate with others early in the process and tend to discuss their ideas with others before they are fully developed (Selby et al., 2004).

Linking Individual Differences and Creative Climate

13

Correspondingly, internals tend to be more self-reliant, apprehensive and perfectionist as well as less socially bold, energetic, dominant, outgoing and emotionally stable in their underlying personality structure than externals (Isaksen et al., 2016). The WD dimension describes where people typically have their priorities in decisionmaking and how they make trade-offs when they are confronted with complex problems (Selby et al., 2004). A task-oriented way of deciding is associated with a preference for objective and verifiable criteria as well as a tendency to seek the solution that is most logical or rational, whereas a person-oriented way of deciding is associated with a preference for subjective criteria as well as a tendency to seek the solution that provides the best fit for all people that are involved in the process (Selby et al., 2004). In their personality structure, people tending towards the person style also tend to be more outgoing and warm, sensitive and energetic as well as less perfectionist, private and rule-conscious in contrast to people tending towards the task style (Isaksen et al., 2016). 2.2.1.3 Impact on Perceptions of Creative Climate The different ways how people prefer to express their creativity and the different approaches they prefer to take in solving problems creatively imply that their creativity is also supported or inhibited in different contexts. The notion of trait activation theory (Tett & Guterman, 2000) suggests, that certain traits need situational opportunities to be expressed and, dependent on individual differences, certain situations are stimulus to different behaviors dependent on individual differences. Even though problem-solving styles are distinct from personality traits, they demonstrated high levels of stability over time (Treffinger et al., 2014a) which is in line with the notion of considering the construct as a source of consistent individual differences as it is explicitly stated in its definition mentioned above. Therefore, it is assumed that different problem-solving styles are also triggered by different situational opportunities that allow for their expression, and, consequently, people of different styles have different requirements to a work environment that provides the best fit to behave according to their style preference. In line with the notion of P-E fit, some researchers found empirical support, that a fit between the preferred cognitive style and the style demands of the workplace is associated with positive outcomes such as a lower intention to leave (Chan, 1996) and an increased self-rated creative performance (Puccio et al., 2000). Furthermore, there is empirical evidence that a general fit between actual and desired environmental conditions has a positive impact on affective outcomes such as commitment and satisfaction (Choi, 2004; Livingstone, Nelson, & Barr, 1997) as well as self-rated creative performance (Choi, 2004). Even though these studies point out the positive impact of having a fit between individual and situational factors on creative behavior and other outcomes in an organizational context, they did not investigate how problem-solving style and other individual differences impact perceptions of creative climate.

14

Theoretical Foundations

This study follows a line of empirical research that was aimed at addressing this research question and will be summarized briefly in the next sections. Kirton and McCarthy (1988) suggested that groups that share different cognitive style characteristics develop different cognitive climates that impact the overall organizational climate and that less personal tension will be found in groups that are more homogenous. Isaksen and Kaufmann (1990) were among the first researchers investigating the relationship between creative climate and problem-solving style directly by linking the KAI inventory with the CCQ in a large quantitative study with 634 participants. They found that people with an adaptive style perceived higher levels of the CHI dimension, whereas people with an innovative style perceived higher levels of the CON dimension in their current climate. However, these relationships were only weak and no other differences were found (Isaksen & Kaufmann, 1990). This study started an academic debate concerning the research methodology and the importance of examining this question. Clapp and Kirton (1994) argued that both domains are conceptually independent linking it to the level-style distinction in the creativity literature, as, in their view, the CCQ represents a measure of organizational progressiveness. Further, they interpreted the findings of the previous study in the way that the CHI and CON dimension form a continuum that describes the psychological state how participants reacted to change rather than the organizational environment. This critique was followed by a response of Isaksen and Lauer (1999) who reanalyzed their previous data and replicated their previous findings. The correlational analysis, which was not reported in their original study, also indicated a relationship between cognitive style and the DEB and RIT dimension, suggesting that innovators perceived higher levels of both dimensions. Interestingly, this relationship could only be found between female innovators and female adaptors revealing some potential gender differences that might be important to consider in future research. In responding to the critique mentioned above, they argued that the CCQ is not a direct measure of progressiveness, but rather, as described earlier in this chapter, a measure of a creativity supporting climate. This may be associated with progressiveness in terms of innovation, but they noted in their response that there is no general climate score that will directly lead to more progressiveness and suggested that this will depend on the organizational context. Furthermore, they pointed to the factor-analytic structure of the CCQ supporting the nine-dimensional climate construct and argued that CHI and CON are not psychological states, but can be treated as related but independent dimensions of creative climate. Nevertheless, they agreed that the few and weak relationships, that were found, provide more evidence for the fact that there is no, or, if at all, only a weak impact of cognitive style on the perception of organizational climates for creativity (Isaksen & Lauer, 1999). This line of research was followed by Isaksen (2009), who first used VIEW as a three dimensional construct of problem-solving style and linked it to the current version of

Linking Individual Differences and Creative Climate

15

the SOQ (Isaksen, 2007) that takes a multi-method approach in assessing the organizational climate for creativity. He reported no quantitative relationships between problem-solving style and perceptions of the current climate for creativity at all and, therefore, concluded that these constructs are independent. However, in analyzing the responses to three open-ended questions asking people to describe factors in their current work environment that support and hinder their creativity as well as the actions they would take to improve their immediate work environment he did find qualitative differences that are in line with the theoretical foundations of the different dimensions of problem-solving style that were described above and suggest that people with different problem-solving styles have different requirement to an environment that supports their creativity. As these findings were very exploratory, he called for future research to further investigate this relationship and outlined the value of taking a multimethod approach (Isaksen, 2009). The fact that prior studies, if at all, found only weak relationships between problemsolving style and the perceptions of current organizational climates for creativity supports the objectivistic perspective on climate considering it as an organizational characteristic that exists independently of individual perceptions (Ekvall, 1987). However, a new research design applied by Isaksen and Aerts (2011) that asked people to make an appraisal of the best- and worst-case situation in which they felt most or least creative and to assess the climate in these situations instead of asking them to assess the current climate revealed significant differences that will be summarized briefly in the next paragraph. This exploratory study used VIEW as a measure of problem-solving style and a short form of the SOQ as a measure of creative climate and found that explorers perceived higher levels of the FRE dimension in their best-case and less idea support in their worst-case situation than developers. The authors concluded that these findings were consistent with the theoretical foundations of the OC dimension in the way that explorers tend to see imposed structure as limiting. Also, their tendency to propose ideas that challenge the existing paradigms is likely to require more support. If they cannot find this support in their work environment, they are more likely to associate such an environment with their worst-case situation. Furthermore, task-oriented people perceived higher levels of the TRO and DEB dimensions in their best-case climate than person-oriented people. The authors suggested that these findings might result from the tendency of task-oriented people to separate people and their ideas so that they are more likely to engage in higher levels of debate and assume that there is a sufficient level of trust as opposed to person-oriented people who tend to think more holistically and, therefore, might avoid open discussion as, in their perceptions, this involves not only idea tension but also personal tension. In addition, a global climate score for the best and worst-case situation was used as the dependent variable in a multiple regression analysis. Gender, age and the WD dimension were found as best predictor variables for the best-case climate, suggesting that women, older as well as task-oriented participants perceived higher levels of the global climate score. In the worst-case

16

Theoretical Foundations

situation, the OC and WD dimensions of VIEW demonstrated the best predictive power among the variables that were examined suggesting that developers and participants tending to the person style in deciding perceived a higher global climate score (Isaksen & Aerts, 2011). Even though the new research design of this exploratory study to examine the relationship between problem-solving style and creative climate showed up to yield promising results, the study had several limitations that shall be addressed in this thesis. Firstly, Isaksen and Aerts (2011) used only a short form of the SOQ that included just one item to measure each dimension of creative climate. This might have confounded the results in the way that participants only indicated their perceptions of certain facets of climate and that not all aspects of creative climate were measured reliably. Secondly, in addition to the potential reliability issues, Isaksen and Aerts (2011) took a solely quantitative approach to assess the climate for creativity, which is why their study lacked to include meaningful contextual data as it is usually provided by the SOQ (Isaksen, 2007). Furthermore, as Isaksen (2009) reported meaningful qualitative differences how people with different problem-solving styles perceived their current climate and work environment, it is assumed that further differences will be found when applying the best-and worst-case research design. Therefore, this study will take a multi-method approach to assess the climate for creativity. Thirdly, Isaksen and Aerts (2011) did not investigate the different facets of the OC dimension of problem-solving style, but rather solely looked at different perceptions of explorers and developers on a global level. As Gilson, Lim, D’Innocenzo, and Moye (2012) called for a deeper investigation of factors that support or hinder incremental or radical kinds of creativity, this thesis seeks to investigate the OC dimension of problem-solving style in more detail by examining each of the three subscales separately. Finally, Isaksen and Aerts (2011) reported in their regression analysis, that gender and age were the best predictor variables of the best-case climate. However, they calculated a general climate score for their analysis, so that an interpretation of these findings concerning any specific dimension of climate was not possible. As these findings need further investigation, this thesis also examines age and gender differences in the perception of organizational climates for creativity, instead of treating these sources of individual differences merely as control variables, whereby the focus will be on gender differences. 2.2.2 Gender Prior studies investigating gender differences in creative behavior barely found significant differences, and, if so, mostly contrasting findings, which is why Baer and Kaufman (2008) conclude in their meta-analytic study that the differences that were found in some studies may be primarily caused by environmental influences. Furthermore, there is also empirical evidence for gender differences in areas related to creative climate such as conflict management (Brewer et al., 2002), risk-taking behavior (Byrnes et al., 1999), decision-making (Venkatesh et al., 2000), trust (Maddux & Brewer, 2005) or social support (Olson & Schultz, 1994). Taking this further, it seems reasonably to assert that men and women have different needs concerning a work environment that

Linking Individual Differences and Creative Climate

17

allows them to use their creative potential most effectively and, therefore, that there are some potential differences in the way men and women perceive a creativity supporting or inhibiting organizational climate. Related to this, a topic that gained recently attention among practitioners is the notion of gender intelligence (Annis & Merron, 2014). For instance, Annis and Merron (2014) argue that based on genetic and sociological differences men and women have different behavioral tendencies, such as typical thinking patterns or different strategies for managing conflict, and, therefore, they assert that men and women have different needs concerning their work environment in order to work most effectively. However, there are only few deliberate investigations of gender differences in the literature that showed contrary findings. Kwaśniewska and Nęcka (2004) for instance indicated, that female managers perceived less autonomy in their current climate compared to male managers. In contrast to this, correlations of the nine SOQ dimensions with gender in a large aggregated sample suggest that women tended to have more positive perceptions of their current climate, including autonomy, than men (Isaksen, 2007). Yet, the relationships reported in these studies were only weak which supports the objective nature of climate described above. Given the significant impact of gender on the perception of the climate for creativity which was most preferred by participants reported in the study of Isaksen and Aerts (2011), this study seeks to further explore this potential impact of gender beyond differences in problem-solving style. 2.2.3 Other Sources of Individual Differences Even though this thesis focuses on the impact problem-solving style and gender on the perception of creative climate, other sources of individual differences, i.e. age, lifephases and creative self-efficacy, are examined. Their potential impact on climate perceptions will be discussed briefly within the next section of this chapter. 2.2.3.1 Age The correlations between age and the nine climate dimensions measured by the SOQ that were reported by Isaksen (2007) were of very small magnitude suggesting that age has, if at all, only a weak impact on the perception of current climates for creativity which is consistent with the lack of findings concerning the impact of the other sources of individual differences on the perception of current climates for creativity mentioned above. However, the results reported by Isaksen and Aerts (2011) suggested that older people tended to perceive a stronger positive creative climate, measured by an omnibus climate score, in their best-case situation than younger people. This would be consistent with Hunter et al. (2007), who found in their meta-analytic study some evidence that climate had a stronger impact on the creativity of older people and assumed that their sensitivity to climate might be due to their broader experience and corresponding frames of reference. In order to further explore the impact of age on perceptions of creative climate, it is included as an independent variable in this study.

18

Theoretical Foundations

2.2.3.2 Life Phase In addition to examining merely age differences, this thesis also considers the current life phase (LP) of participants as another source of individual differences. Gerlmaier and Latniak (2016) studied characteristics of biographical heterogeneous innovation teams in an IT context and provided some evidence that people with children living in their households indicated that they worked less intensely than people without children, suggesting that they might be less involved in the activities and work of their team. The authors of this study also reported that the perceptions of the current team climate were roughly equal, but, however, they used only a four-item measure of innovative team climate, that merely tapped on elements that can be compared with the CON and IDT dimensions of the operationalization of creative climate on which this thesis is based (Gerlmaier & Latniak, 2016). This study seeks to examine the impact of LP on the perceptions of creative climate in best- and worst-case situations. LP will be operationalized based on the model of family life cycles developed by Carter and McGoldrick (1988), who proposed normative and non-normative life courses, whereby the latter case involves divorces and patchwork families. Accordingly, participants were asked to indicate whether or not they live in a relationship and if they have one or more children of different age groups to determine their current LP. 2.2.3.3 Creative Self-Efficacy Creative self-efficacy (CSE) is defined as “the belief one has the ability to produce creative outcomes” (Tierney & Farmer, 2002: 1138). Even though this thesis mainly focuses on problem-solving style rather than level or creative ability, the construct of CSE was added as a component that is associated with individual (Tierney & Farmer, 2004) and team (Shin & Zhou, 2007) creative performance. Robinson-Morral, Reiter-Palmon, and Kaufman (2013) examined the relationship between CSE, job requirements and creative performance and found, that people who were most creative had high self-perceptions of CSE and perceived that their job required creativity to a high degree. Taking this further, in this thesis it will be examined if individuals with higher self-perceived creative ability also perceive that they can actualize their creative potential most effectively in different climates for creativity. CSE will be measured by using a three-item scale that was developed by Tierney and Farmer (2002). In the further course of this study, the next chapter will describe in more detail how this question was addressed methodologically, before the results will be presented and discussed.

3 Method This chapter will provide an overview of the research design that was applied to answer the guiding research questions of this thesis. After this, the targeted sample and approach for the data collection will be described, followed by a brief summary of the psychometric properties of the key assessments applied in this study. Finally, the approach for the quantitative and qualitative analysis of the data will be explained. 3.1 Research Design For answering the guiding research questions, this thesis takes a multi-method approach and assesses the climate for creativity from a quantitative and qualitative perspective. The quantitative approach to assess the climate for creativity was chosen to get a standardized assessment of the different organizational climates for creativity that can be compared against each other. This approach also allows to determine, whether, and to what extent, the different types of individual differences had an impact on the perception of climate, and if so, what specific aspects of climate might reveal the largest differences. However, a solely quantitative design would only allow for a broad overview, but not for a deeper understanding of how and why certain aspects of climate could be perceived differently. To address this limitation, that Isaksen and Aerts (2011) identified in their study, a qualitative part was added, asking people to describe the most important factors that supported and inhibited their creativity in the context of their best- and worst-case situation. Also, as Ehrhart et al. (2014) argue, climate research, which had an historical focus on quantitative approaches, could benefit from integrating qualitative approaches as well, that were historically more often applied in research on organizational culture. The distinct types of individual differences, that were described in the previous chapter, will be treated as independent variables and the perceptions of the nine dimensions of creative climate will be treated as dependent variables. As the main focus of this study will be on individual differences in problemsolving style and gender, a quantitative and qualitative examination will only be conducted for these two variables. The impact of the other variables will be examined just quantitatively. As it is important to investigate the positive and negative forces impacting creative behavior in organizations (Amabile et al., 1996; Livingstone et al., 1997), participants were asked to reflect on their best and worst-case situation, in which they had positive and negative critical experiences concerning their own creative behavior in a workrelated context. This design was adapted from prior research on the climate for creativity, where it was already successfully applied (Isaksen & Aerts, 2011; Isaksen et al., 2001). The basic idea behind this design can be compared with the original notion of the critical incident technique (Flanagan, 1954), which is applied in qualitative research and also aimed at investigating critical situations in order to understand the positive or negative impact of certain behaviors. However, in contrast to the critical incident technique which asks people to describe their own behavior, the subject of inquiry in this © Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2019 C. Hoßbach, Organizational Climate for Creativity, BestMasters, https://doi.org/10.1007/978-3-658-25241-0_3

20

Method

thesis is to understand the context (i.e. the climate for creativity) in these critical situations. Therefore, all participants in this research study, were asked to respond to the SOQ twice to assess the climate for creativity in the context of their best- and worstcase situation. Given the objectivistic definition of climate on which this study is based, participants were not merely asked to describe a hypothetical positive or negative manifestation of climate in which they would like to work in or not, but rather to reflect on a real situation in which they felt most or least creative. 3.2 Sampling The objective of this study is to gain a deeper insight into the context (i.e. the climate) that is perceived to support or inhibit creativity by different kinds of people, which is why it is not expedient to study the perceptions of the current climate for creativity within a single organization. Therefore, people with work experiences in different organizations were targeted to be included in this research study. Also, given the exploratory nature of this study, no single industry but rather a high variance in the distinct types of individual differences was targeted. However, a necessary condition to be included was that they could clearly identify a positive and negative work situation, in which they felt very creative or uncreative respectively. As mentioned earlier in chapter 2.3.1, creativity is especially required for solving complex and ambiguous problems. Therefore, people who are likely to be frequently confronted with these kinds of problems in their daily work were targeted to be included in this study. With friendly support of The Creative Problem Solving Group Inc. (CPSB), a list of 634 people, who contracted with them for services such as facilitative leadership for CPS or different training programs, was acquired to invite them to participate in this research study. These people work in a variety of different positions in organizations in different countries. As they contracted with CPSB, it is assumed that they are more likely to be confronted with complex problems in their daily work that requires their creativity. However, as mentioned above, this thesis follows the notion that everyone has some creative potential that can be supported or inhibited by contextual factors such as the climate for creativity. Therefore, other people were invited to participate to increase the sample size of this study. Among these additional participants that were invited to participate, were a group of human resources managers and a group of engineers, technical consultants, and sales managers who are working in a variety of organizations in Germany and Austria that were clients of a German commercial agency. As this thesis also seeks to examine the impact of age on the perception of organizational climates for creativity, some younger individuals were also targeted to participate, including a group of Norwegian graduate students with some prior work experience who were enrolled in a course focusing on creativity in individuals, teams, and organizations, a group of alumni who completed a dual-track program in business administration at a German university and are now working in a variety of different positions, as well as some other young professionals.

Data Collection

21

3.3 Data Collection Participants were invited by receiving an e-mail in which they were asked to participate in this research study. They were informed about the scope and relevance of the project as well as the different opportunities concerning their way of participation. They were then invited to register via an online questionnaire and should either create an anonymous participant code or indicate their names and e-mail addresses if they were interested in receiving an individual feedback report about their preferred problemsolving style in return for their participation. This questionnaire also contained questions concerning their LP and CSE using the scales that were mentioned above. After this, participants were invited to complete VIEW as a measure of their problem-solving style and to respond to the SOQ twice to assess the climate for creativity in their bestand worst-case situation. Furthermore, participants were able to choose if they want to respond to the questions in English or German. Finally, the responses of the different questionnaires were matched together, and only complete data sets were included in the sample for the subsequent analyses. Given the relatively comprehensive questionnaires and the amount of time which was required from participants, they were offered feedback about their individual problemsolving style and a summary of the key insights of this study in return for their participation. In line with the ethical research guidelines and code of conduct of the American Psychological Association (2002, 2010), people participated with informed consent and were ensured of their anonymity. If they indicated their names because they were interested in the feedback, they did so voluntarily and were ensured that their results will be held strictly confidential and will be anonymized before they will be included in this research study. 3.4 Measures As the measures of problem-solving style and organizational climate for creativity were used in its original form to achieve a reliable and valid assessment of the key variables that were examined in this study, their psychometric properties will be described in more detail in the next section of this chapter. 3.4.1 Problem-Solving Style The construct of problem-solving style is measured by using the VIEW assessment (Selby et al., 2002). The purpose of this section is to give a brief overview of how the assessment works and its psychometric properties, as a detailed description of the dimensions, that are applied as an operationalization of problem-solving style in this thesis, was already provided in the previous chapter. A comprehensive overview of the development and construction of the instrument was provided by Selby et al. (2004), on which the following summary is based. VIEW is a web-based assessment with 34 items that are scored on a seven-point bipolar scale.

22

Method

The OC dimension is assessed by 18 items, of which five items belong to each of the three subscales NV, SA, and SS. The additional three items are used to assess some general aspects of the OC dimension. Furthermore, the MP and WD dimension are assessed by eight items respectively. Each item refers to the general statement “When I am solving problems, I am a person who prefers…” and offers two opposing descriptions to continue this statement. Respondents shall indicate in each case on the sevenpoint scale, which of these two descriptions better applies to their preferred approach of solving problems. In each case, both statements are written in a positive way in an attempt to reduce response bias due to social desirability. The item scores, that can each range from one to seven, are then added up to calculate the overall score in each dimension. Scores in the OC dimension can range from 18 to 126 with a theoretical mean of 72. Scores below the mean are indicating a tendency towards the explorer style while scores above the mean are indicating a tendency towards the developer style. In each of the three subscales of OC, scores can range from five to 35 with a theoretical mean of 20. Again, lower scores are indicating an explorer and higher scores are indicating a developer preference. In the MP and WD dimension, scores can range from eight to 56 with a theoretical mean of 32. Scores below the mean are indicating a preference for the external or person style respectively, and scores above the mean are indicating a preference for the internal or task style respectively (Selby et al., 2004). VIEW is a well-researched and validated assessment and demonstrated ample evidence of its reliability and validity, including stability over time, sufficient degrees of internal consistency, a coherent factor structure, as well as correlations with related constructs, which is summarized by Treffinger et al. (2014a). Given its psychometric properties and its direct relationship to creativity and CPS, VIEW was chosen to be used as a measure of problem-solving style in this research study. As the assessment is proprietary and its use requires certificated training, the items are not published in any publicly available source. However, the author completed a qualification program and got the permission to use the assessment for this research study. In order to allow for a more reliable assessment and limit misinterpretations due to language issues, VIEW was translated into several languages and a summary of the reliability of the different versions can also be found in Treffinger et al. (2014a). Even though comprehensive data for assessing the quality of the German translation is not yet available, it is assumed that the translation is of sufficient quality to be used for this research study, given the thorough translation methodology, in which the author was involved and which followed accepted guidelines for the cross-cultural translation of assessments (Brislin, 1970; Geisinger, 1994). Nevertheless, Cronbach’s Alpha (a) will be conducted for the current sample to ensure a sufficient degree of reliability of the assessment of participant’s problem-solving style.

Data Analysis

23

3.4.2 Climate for Creativity As described in the previous chapter, the SOQ was chosen to be used as a measure of organizational climate for creativity in this study. The SOQ is based on the foundations of one of the first measures of creative organizational climate that can be found in the literature (Ekvall, 1983, 1996) and has been continuously developed since, which makes it a very well-researched instrument with good psychometric properties, including sufficient scale reliabilities and a coherent internal factor structure, that are summarized in (Isaksen & Ekvall, 2015). A comprehensive overview of the construction and measurement approach of the current version of the SOQ was provided by Isaksen (2007), which will be summarized briefly in this paragraph. The SOQ is a web-based questionnaire that assesses the nine dimensions of creative climate, that were described above, with 53 items, whereby it includes one distraction item that is not related to any of the nine dimensions. Considering climate in an objectivistic sense, items are worded in the way that they refer to observable patterns of behavior, treating respondents as observers of their work environment. Each item is scored on a four-point scale ranging from zero to three on which respondents shall indicate to what extend they perceive each statement as applicable to the patterns of behavior that they observe and experience in their work environment. Item scores in each dimension are then added up, divided through the number of items in each dimension and, for an easier comparison, multiplied with the factor 100 to obtain the overall score in each dimension. The SOQ is constructed as a multi-method tool, as it also includes open-ended questions in addition to the quantitative part described above. Three questions ask respondents to indicate the factors in their work environment that are most helpful and most hindering to their creativity as well as the most important actions they would take to improve the climate for creativity in their immediate work environment. This additional qualitative data can be used to put the quantitative climate scores into a more meaningful context (Isaksen, 2007). Similar to the VIEW assessment, the SOQ is also proprietary, and its use requires certificated training, which is why the items are not published in any publicly available source. The author is qualified and was permitted to use the SOQ for the purpose of this research study. With the involvement of the author, the German translation of the SOQ was also approached recently applying a similar methodology as in the German translation of VIEW. As a group of German participants were invited to participate in this research study, they were free to choose if they want to complete the assessment in English or German. However, to ensure a sufficient reliability of the scales, Cronbach’s coefficient Alpha (a) will be computed for the current sample of this thesis. 3.5 Data Analysis After this chapter provided an overview of the research design, targeted sample as well as approaches for collecting the data and measuring the key constructs, the last section of this chapter will focus on the way in which the data will be analyzed from a

24

Method

quantitative and qualitative perspective in order to answer the guiding research question of this thesis. 3.5.1 Quantitative Data Analysis The quantitative analysis was done with the help of the statistic software Statistical Package for the Social Sciences (SPSS) version 24. In a first step, the data of the different questionnaires was matched together in one database by using the participant codes or names. Incomplete datasets from people who provided only one climate assessment of their best- or worst-case situation were excluded from the subsequent analysis. After structuring the data, descriptive statistics were conducted to give an overview of the distributions of the examined variables (see section 4.1.2). Also, as the sample of this study consists of participants from varied backgrounds, working in a variety of organizations and countries, the rwg-indicator was calculated to ensure that there was a sufficient degree of interrater agreement (James, Demaree, & Wolf, 1984; LeBreton & Senter, 2008) in their assessments of their best- and worst-case climates for creativity. In a next step of the analysis, it was examined if participant’s perceptions of the climate in their best- and worst-case situation were significantly different from each other. As the same participants provided data from both situations, a test of equality of means for dependent samples (t-test) was conducted and effect sizes (Cohen’s dz) were calculated by dividing the mean differences (D) between both situations in each dimension through its standard deviation (SD) (Schäfer, 2016). After this, correlational analysis was applied as a first step to explore the relationships between individual differences and climate perceptions. Also, as all the perceptions in all climate dimensions were significantly different in the best- and worst-case situation, the relationship between the individual differences and the magnitude of the perceived differences of the climate dimensions between these two opposing situations were also examined. The Pearson correlational coefficient was computed for all metric variable pairs, including the dimensions of problem-solving style and creative climate as well as CSE. As the characteristics of a dummy variable, which is coded (0;1), can be regarded as quasi-metric, the Pearson correlation can also be applied for this specific combination of nominal and metric variables (Cleff, 2015). Therefore, gender was included as a variable in the correlational analysis. Although age was measured as a metric variable, it was recoded into a dummy variable that distinguishes between participants that are older and younger than 40, due to its non-normal distribution in this sample. Moreover, as only few participants indicated data concerning their LP and, therefore, not all theoretical LP’s were represented in this sample, this variable was also recoded into a dummy variable that distinguishes between participants with and without children. Correlational analysis assumes that there is a linear relationship between two variables and there might be misleading interpretations of correlations in the case of a non-linear relationship (Cleff, 2015). Therefore, a graphic analysis was applied using scatter plots

Data Analysis

25

to ensure a linear relationship and no evidence of a non-linear relationship was found. Another assumption of correlational analysis that is important for the test of significance is the normal distribution of each pair of variables, however, the test is relatively robust against violations of this assumption (Cleff, 2015). Furthermore, this thesis is of exploratory nature and the main target is not to confirm that there is a relationship in the general population, but rather to explore what types of individual differences might have a greater influence on the perception of organizational climates for creativity in order to form a foundation for further research. Therefore, comprehensive statistical tests such as the Kolmogorov-Smirnov test (Janssen & Laatz, 2017) were not conducted and, instead, the distribution of all variables was examined visually by plotting it against a theoretical normal distribution. Except for age, which was dummy coded as described above, it is assumed that the variables follow the patterns of a normal distribution to a sufficient extend. Moreover, correlations are a necessary but not sufficient condition for assuming a causal relationship between two variables (Cleff, 2015). Therefore, the correlational analysis was just a first step to explore possible relationships between individual differences and climate perceptions. It was followed by several analyses of variance to examine the impact of each source of individual differences in more detail. An analysis of variance (ANOVA) examines differences concerning a dependent metric variable (in this case the climate dimensions) between nominal groups determined by an independent variable (in this case the individual differences) (Schäfer, 2016). In addition, a multivariate analysis of variance (MANOVA) takes the interaction of several dependent variables into account (Mertler & Reinhart, 2017). Therefore, a MANOVA was conducted first to examine the overall impact of each source of individual differences on climate perceptions in the best- and worst-case as well as the perceived difference between both situations. Wilk’s Lambda (L) was used as an indicator to interpret the overall quality of each MANOVA, whereby the coefficient can theoretically range from zero to one with smaller values indicating a stronger impact of the examined independent variable (Mertler & Reinhart, 2017). As Mertler and Reinhart (2017) suggest, the analysis would not go any further if the MANOVA impact was not statistically significant. However, given the exploratory nature of this study, subsequent ANOVAs were conducted to examine the single impact of all individual differences on each climate dimension to explore areas in which potential differences might be found, even though the overall impact of a certain source of individual on climate was not significant in this sample. This is also supported by the fact, that the level of significance is sensitive to the size of the sample (Schäfer, 2016), so that more significant results might be found in a larger sample. Effect sizes of potential differences were then examined by calculating eta-squared coefficients (h2) that give an indication of the proportion of the variance in climate scores that was explained by each source of individual differences and the level of significance was interpreted by conducting F-tests (Schäfer, 2016).

26

Method

The (M)ANOVAs also assume a normal distribution of the examined variables, which was already discussed above in the section about the correlational analysis. In addition, Box’s test was conducted for each MANOVA to ensure that the assumption of homogeneity of covariance is met (Mertler & Reinhart, 2017). A further assumption of the subsequent ANOVAs is that the variance of the dependent variable, is homogeneous in all groups of the independent variable, which was examined in SPSS by conducting the Levene test in each case (Fromm, 2012). These tests showed no irregularities for the significant results that were interpreted in this study. To prepare for the analysis, in addition to the variables that were already dummy coded, different groups of contrasting problem-solving styles were formed (see appendix 1). Given the small number of participants who responded to questions concerning their LP and CSE, a MANOVA was not conducted in these cases. After exploring the correlations and applying the MANOVAs gave an overview of the potential impact of individual differences on climate perceptions, a multiple linear regression was conducted for those climate dimensions that were impacted by several individual differences in order to determine their joint impact on a certain dimension of creative climate. As the multiple regression was applied to find out the best predictor variables among those that demonstrated an impact in the previous analysis, a backward approach was chosen to enter the predictor variables in the model. This approach first includes all variables and then stepwise deletes those with a poor fit to the model in order to improve the overall quality of the model (Janssen & Laatz, 2017). The overall quality of the regression model was assessed by the adjusted coefficient of determination (R2), the level of significance was examined by conducting an F-test and the predictive power of each variable was assessed by interpreting the standardized regression coefficient (b) (Janssen & Laatz, 2017). In addition to an approximately normal distribution of the examined variables and a linear relationship between independent and dependent variables, the multiple linear regression assumes independence and normal distribution of the residuals, and only a limited degree of multicollinearity among the predictor variables (Janssen & Laatz, 2017). However, given the exploratory nature of this study, the former assumptions were only examined visually by plotting the residuals against a normal distribution and examining their variance in relationship to the dependent variable by using scatter plots (Janssen & Laatz, 2017), and it was controlled for multicollinearity by examining the tolerance estimate in SPSS. It is assumed that these assumptions were met to a sufficient degree to allow for an interpretation of the results within this study. 3.5.2 Qualitative Data Analysis After the quantitative analysis gave a broad overview of the potential impact of distinct types of individual differences, the qualitative analysis shall allow to put these results into a more meaningful context and explore some additional perceptual differences that might not have been revealed in the quantitative part. Two of the three open-ended questions of the SOQ asking participants to describe the factors that supported and

Data Analysis

27

hindered their creativity in their best- and worst-case situation, were subject of the qualitative analysis. Even though the third question concerning potential improvement actions was not included in the qualitative analysis of this thesis, it was chosen to collect the data, as this thesis is designed to form a foundation for future inquiry in this area that might also involve a reanalysis of some data included in this study. A first step to prepare for the qualitative analysis of the data, was to translate the German comments into English so that all narrative data could be analyzed in the same language. It was tried to keep the translation as close to the meaning of the original statements in order to ensure that a potential loss of information was kept at a minimum level (see appendix 2 for the translation). Even though the scope of the narrative comments is of a manageable size as they were just derived from open-ended questions in a mainly quantitative questionnaire, the number of participants and complexity of the research design including two opposing situations called for a more comprehensive approach for the qualitative analysis. Therefore, it was chosen to use the software Windows Relation Analysis (WinRelan) and to apply some aspects of the GABEK methodology (Zelger, 2000). The main reason for choosing this program was the fact that it allows not only to tag the different sources of individual differences in each narrative comment by using criteria, but also to add a positive or negative evaluation to each coded key term (Zelger, 2000). This way, the software allowed for a direct comparison between different groups of participants, such as explorers and developers, concerning their perceptions of their bestand worst-case situation. After the data was imported to the program by using a separate index card for each narrative comment and tagging the different sources of individual differences by using criteria, the first step of the analysis involved the open coding and constant comparison of key terms in each comment followed by an indication of the positive and negative evaluation of each key term in which the best- and worst-case situation were treated separately by using different evaluation lists in WinRelan (Zelger, 2002). This resulted in a list of coherent factors that were identified by participants that also mirrors the positive or negative connotations that participants associated with these factors. Other elements of the very comprehensive GABEK approach to qualitative analysis, were not applied given the nature of the narrative data included in this research study. Rather, as this study focuses on creative climate, the factors that were inductively derived from the narrative comments in the first step, were then coded against the nine climate dimensions applying a deductive approach (Mayring, 2015) in the second step of the analysis. As not all factors fell into the conceptual space of climate, the other factors were coded against the elements of the MOC, which was presented in figure 1 above. Definitions of these elements are provided in appendix 3. Finally, in a third step, similar to a force-field analysis (Lewin, 1951), it was analyzed what factors were perceived to be supporting or restraining creativity in the best- and

28

Method

worst-case situation and in what ways both situations were associated with the same or different positive and negative forces, by analyzing the evaluation profiles of the best- and worst-case situation in WinRelan and comparing the profiles of different groups of participants reflecting problem-solving style and gender differences against each other. Even though there was roughly a normal distribution among the style variables and a roughly equal number of men and women, the group sizes were not exactly equal, and, furthermore, the number of codes in each group was also dependent on if and how detailed participants responded to each of the open-ended questions. Therefore, the frequency of each identified factor was weighted with the total number of codes in each group to allow for a direct comparison between the respective groups. After the last section of this chapter introduced the approaches for the quantitative and qualitative analysis that were applied in this study, the next chapter will give an overview of the results of the different analyses.

4 Results This chapter of the thesis will focus on analyzing the gathered data from a quantitative and qualitative perspective. The quantitative results will give a broad overview of the climate perceptions in the best- and worst-case situation and the potential impact of the different types of individual differences that were examined. After this, the qualitative results will add more depth by examining different aspects and manifestations of climate and some other factors in the work environment that were perceived to have a positive or negative impact on creativity. 4.1 Quantitative Results The first part of this chapter is dedicated for the quantitative results. A short overview of the sample is followed by the descriptive results among the individual differences and climate variables. After this, a comparison of climate perceptions in the best- and worst-case situation will give an overview of the perceived differences. The impact of individual differences on climate perceptions is then examined by exploring correlations as well as conducting several MANOVAs and multiple regression analyses. Finally, the most important quantitative findings will be summarized. 4.1.1 Sample A total number of 139 people participated in this research study. As some of them only completed the SOQ for the best- or worst-case situation, 16 incomplete data sets were excluded from the analysis, resulting in a sample size of 123 participants. These were working in a variety of different positions including consultancy, teaching or human resources and other areas such as management, sales or research and development. Appendix 4 shows a detailed listing of all current occupations that have been indicated by participants. There was roughly an equal number of male (61) and female (62) participants. The mean age of participants was 40.2 years with a standard deviation of 16 years, ranging from 21 to 70. As some participants declined to indicate their age, data concerning age is only available from 117 participants. The distribution deviates heavily from a normal distribution resulting in two groups of rather young and rather old participants and only few participants around the age of 40 (see appendix 5). Therefore, age was not treated as a metric variable in the subsequent analyses but was recoded into a dummy variable reflecting two groups of people up to an age of 40 (N=64) and people who are older than 40 (N=53). The general overview of the composition of the sample is now followed by a detailed description of the distribution of the results in all variables that were examined. 4.1.2 Descriptive Statistics This section will start with providing an overview of the distribution of results, scale reliabilities and inter-correlations among all individual differences variables that were examined. After this, climate perceptions in the best- and worst-case, the inter-rater agreement as well as inter-correlations among climate dimensions will be analyzed. © Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2019 C. Hoßbach, Organizational Climate for Creativity, BestMasters, https://doi.org/10.1007/978-3-658-25241-0_4

30

Results

4.1.2.1 Individual Differences A description of the distribution of scores in each of the three dimensions of VIEW as well as the subscales of the OC dimension is followed by the descriptive results for the CSE scale, and the composition of participants LP. After this, the inter-correlations among the individual differences will be examined. Problem-Solving Style Table 2 shows the results for each of the three dimensions of VIEW as well as the three subscales of the OC dimension for all 123 participants. Almost the full range of possible scores was obtained in each of the dimensions reflecting a diverse sample. The observed means were close to the theoretical means in all dimensions except for the WD dimension, in which the observed mean of 35.9 is slightly higher than the theoretical mean of 32, indicating that more participants in this sample had a preference for the task style. This is also reflected in the higher median and mode in this dimension. Whereas the median in the OC dimension, its subscales, and the MP dimension is close to the mean, the mode in the OC and MP dimensions deviates from it. In addition, all distributions are slightly skewed and have a moderately negative kurtosis, indicating that they are not perfectly symmetric. However, a deviation with a magnitude of less than one concerning skewness and kurtosis is regarded as acceptable to assume an approximately normal distribution (Bowen & Guo, 2012). Furthermore, the results of this sample were close to the results reported in the technical manual of VIEW for a sample of 44,802 respondents (Treffinger et al., 2014a), suggesting that the distribution of scores in the VIEW dimensions in this sample was close to the distribution in the general population. In addition to the descriptive results, a graphic comparison was conducted by plotting the distribution of scores in each dimension against a normal distribution (see appendix 6). The graphic comparison supports the assumption that the distribution of VIEW scores in all dimensions approximately follows the general pattern of a normal distribution. In all dimensions, the reliability coefficient (a) is well above the accepted standard of .70 (Nunnally & Bernstein, 1994), indicating a sufficiently high internal consistency of the scales. The high reliability of the scales suggests that participants had a quite similar understanding of the items even though they responded in different languages.

Quantitative Results

31

Table 2: Descriptive Statistics for VIEW Dimensions Mean

SD

Range

Median

Mode

Skewness

Kurtosis

a

OC

71.6

20.1

23-117

73

59

-.15

-.42

.93

OC: NV

18.9

6.2

6-33

19

22

.05

-.54

.87

OC: SA

19.1

6.2

5-34

19

18

.04

-.75

.76

OC: SS

21.1

6.5

7-34

22

24

-.26

-.87

.83

MP

31.1

10.5

8-54

33

38

-.11

-.78

.91

WD

35.9

8.8

16-56

36

39

.10

-.40

.88

Variable

a

N=123 Source: Own Survey.

CSE and LP Of the 123 participants, only 46 responded to the additional questionnaire and provided data concerning their LP and CSE. Almost the full range of possible scores from very low to very high self-ratings of CSE was obtained. The scale demonstrated a sufficient degree of reliability and the observed mean, median and mode were slightly higher than the theoretical mean of the scale, and the distribution has a moderately negative skewness and kurtosis, indicating a deviation from a normal distribution (see appendix 7). Given the low response rate, not all theoretical life phases were reflected in this sample (see appendix 8). As mentioned in the previous chapter, it was transformed into a dummy variable reflecting two groups of those who indicated that they have children (N=23) and those who indicated that they do not have children (N=23). Inter-Correlations among Individual Difference Variables Small but significant correlations were observed between the WD dimension and the OC and MP dimensions of VIEW, as well as gender, indicating that participants who had a preference for the task style were a little bit more likely to be men, and to have a preference for the developer style (especially concerning the SS aspect of OC), and the internal style. Also, significant negative correlations were observed between age and all aspects of the OC dimension, indicating that older participants were more likely to have a preference for the explorer style (especially concerning the SA aspect). In addition, a significant positive correlation was found between gender and the NV aspect of OC, indicating that women in this sample were more likely to have a preference for the developer style. However, the correlation with the other two aspects of OC and the overall OC dimension was not statistically significant but goes in the same direction. These findings concerning the VIEW dimensions were consistent with those reported in the technical manual (Treffinger et al., 2014a), even though the magnitude of the relationship between age and the OC dimension was stronger, which might have been affected by the non-normal distribution of age in this sample. Moreover, a significant negative correlation was found between CSE and the NV aspect of OC, indicating

32

Results

that participants who thought of themselves as more creative were more likely to have a preference for the explorer style. Finally, significant correlations were observed between LP and the SA aspect of OC as well as age, indicating that participants, who stated that they have children, were more likely to be older and to have a preference for the explorer style. A summary of the inter-correlations among individual differences variables is provided in appendix 9. 4.1.2.2 Climate for Creativity Table 3 shows the descriptive results of the perceptions of the climate for creativity of all 123 participants in the nine SOQ dimensions in their best- and worst-case situation. The mean scores in the eight generally positive climate dimensions were, as expected, higher in the best-case situation compared to the worst-case situation. The same expected pattern was observed for the negative CON dimension with higher mean scores in the worst-case situation compared to the best-case situation. This pattern is further illustrated by the range of scores that was generally located more in the upper sector of possible scores in the best-case situation and more in the lower sector of possible scores in the worst-case situation, except for the CON dimension. The distribution of scores in all dimensions for the best- and worst-case situation is displayed and plotted against a normal distribution in appendix 10 and 11. As it was graphically examined, the distribution of scores in all dimensions generally follows the patterns of a normal distribution, except the positive and negative skewness, which was mainly caused by the research design asking people to think of their best- and worst-case situation when they responded to the questionnaire. However, as mentioned in the previous chapter, the quantitative analyses that were conducted in this thesis are relatively robust against deviations from a normal distribution, which is why it is assumed that the skewness and slight deviation from a normal distribution did not affect the subsequent analyses to a large degree. As the sample of this exploratory study was very diverse, the rwg coefficient was conducted as a more profound measure of interrater agreement (James et al., 1984; LeBreton & Senter, 2008). All climate dimensions demonstrated strong (rwg > .70) to very strong (rwg > .90) agreement among raters (LeBreton & Senter, 2008), justifying the aggregation of individual climate perceptions to calculate mean scores for the climate in the best- and worst-case situation. The only exception, was the perception of the CON dimension in the worst-case situation that demonstrated only moderate (rwg > .50) agreement among raters (LeBreton & Senter, 2008). Even though it is acknowledged that this indicates a rather low level of interrater agreement, the CON dimension in the worst-case situation was not excluded from the subsequent analyses. Furthermore, all dimensions demonstrated an internal consistency (a) above .70 that was close to the values reported in the technical resources of the SOQ (Isaksen & Ekvall, 2015), suggesting that the measured climate scores for the best- and worst-case situation in this sample can be regarded as reliable.

Quantitative Results

33

Table 3: Descriptive Statistics for Best- and Worst-Case Climate Perceptions Best-Casea

Worst-Casea

Dimension

Mean

SD

Range

a

CHI

230.1

50.1

86-300

.86

FRE

201.9

57.0

50-300

TRO

197.2

60.1

IDT

169.9

PLH

rwg

Range

a

Mean

SD

rwg

.92

125.1

64.9

0-257

.89

.84

.83

.87

127.1

65.4

0-300

.84

.79

40-300

.77

.80

104.7

55.6

0-240

.74

.83

65.5

0-300

.88

.83

83.2

55.9

0-250

.85

.88

211.5

57.5

33-300

.84

.87

103.0

63.6

0-267

.89

.86

CON

81.3

61.7

0-267

.83

.82

187.1

78.9

0-300

.87

.58

IDS

200.7

63.2

20-300

.88

.85

89.1

55.8

0-200

.86

.89

DEB

206.8

57.0

17-300

.86

.88

128.2

63.2

0-283

.87

.85

RIT

168.9

63.9

0-300

.82

.80

84.2

53.7

0-260

.78

.86

a

N=123 Source: Own Survey.

In addition, all climate dimensions demonstrated significant positive correlations in the best-case situation, except for CON that was negatively correlated with the other dimensions (see appendix 12). In the worst-case situation, a similar pattern was observed (see appendix 13), except for fewer significant correlations with CON which might be partially due to the high dispersion of scores in this dimension as mentioned above. These results indicate some degree of multicollinearity among the climate dimensions, and are generally consistent with those reported in the technical resources of the SOQ (Isaksen & Ekvall, 2015). 4.1.3 Differences between Best- and Worst-Case Climate Perceptions All participants responded to the SOQ twice and thereby assessed the climate for creativity in their best- and worst-case work situation, which they had identified when they responded to the questionnaire. As shown in table 4, the largest difference between the best- and worst-case situation was observed in the perception of IDS, whereas the smallest difference was observed in the perception of FRE. The means of the perceptions of climate in the best- and worst-case situation were all highly significantly different from each other. Furthermore, all differences demonstrated high effect sizes (Cohen, 1988; Schäfer, 2016). As the t-test reacts relatively robust to violations of the assumption that the variables are normally distributed (Schäfer, 2016), it is assumed that, for the purposes of this exploratory study, the distribution of scores in the climate dimensions is sufficiently normal to not affect the results of the t-test.

34

Results Table 4: Differences between Perceptions of Best- and Worst-Case Climates Dimensiona CHI

Mean Difference 105.0

FRE

74.8

TRO

92.5

IDT

86.7

PLH

108.5 b

CON IDS DEB RIT

105.8 111.5 78.6 84.7

SD 78.6 78.1 83.3 80.5 81.9 94.4 83.8 77.2 76.6

tc

dz ***

1.34

***

.96

***

1.11

***

1.08

***

1.32

***

1.12

***

1.33

***

1.02

***

1.11

14.81

10.63 12.31 11.94 14.70 12.43 14.75 11.29 12.27

a

N=123 The difference for the CON dimension was reversed. df = 122 *** p < 0.001 (2-tailed) Source: Own Survey. b c

After the last section of this chapter gave an overview of the composition of the sample and distribution of the results in the individual difference and climate variables that were examined, the next section focuses on the interaction of these variables and seeks to identify important quantitative connections between individual differences and perceptions of organizational climates for creativity. 4.1.4 Impact of Individual Differences on Climate Perceptions The first step in exploring connections between distinct types of individual differences and the perceptions of organizational climates for creativity was to conduct correlational analysis. 4.1.4.1 Exploring Correlations Table 5 shows the correlations between individual differences and the perceptions of the climate dimensions in the best-case situation. Among the 90 possible correlations, 18 were statistically significant and all of these relationships were negatively correlated. Most of the statistically significant correlations of climate dimensions were found with the OC dimension of VIEW, indicating that with a decreasing score on OC (an increasing preference for the explorer style), participants tended to perceive higher levels of the CHI, PLH, IDS, DEB, and RIT dimensions. The three subscales of the OC dimension reveal which specific aspects of the overall OC dimension might have been potential drivers of these relationships. Significant correlations with the CHI dimension were found with all three OC subscales. However,

Quantitative Results

35

the correlation with the SA subscale was lower than the other two correlations, suggesting that NV and SS might be more important aspects for the relationship between OC and CHI in the best-case situation. The same pattern was observed concerning the relationship between OC and IDS. In this case, the correlation of the SA subscale was not even significant. Concerning PLH, only the SS subscale demonstrated a significant correlation that was also stronger than the correlations with the other subscales and the overall OC dimension, suggesting that the SS aspect of OC might be more important for this relationship. On the contrary, the SS aspect demonstrated only a smaller and not significant correlation with the DEB dimension of climate, which shows that this relationship might have been more driven by the NV and SA aspects of OC. Finally, significant correlations between OC and RIT were found in all three subscales. However, the NV aspect of OC demonstrated the strongest correlation. Furthermore, a significant correlation was found between the MP dimension of VIEW and perceptions of the RIT dimension in the best-case situation, indicating that with a decreasing score on the MP dimension (an increasing preference for the external style), participants tended to perceive higher levels of RIT in their best-case situation. A significant and slightly stronger correlation was also found between gender and RIT, suggesting that men perceived more RIT in their best-case situation than women. No significant correlations were found between the WD dimension of VIEW, age, CSE, or LP with any of the climate dimensions in the best-case situation. Table 5: Correlations with Best-Case Climate Perceptions Best-Case Climate Dimensions Variable OC -

*

N

CHI

123

-.26**

-.06

-.10

-.15

-.19*

.06

-.21*

-.19*

-.26**

-.27**

-.05

-.13

-.13

-.15

.04

-.22*

-.20*

-.28**

-.19*

-.06

-.07

-.09

-.11

-.02

-.12

-.18*

-.23**

-.28*

-.08

-.05

-.14

-.24**

.11

-.19*

-.14

-.20*

NV SA SS

FRE

TRO

IDT

PLH

CON

IDS

DEB

RIT

MP

123

-.14

-.16

-.05

-.17

-.13

.02

-.17

-.07

-.19*

WD

123

-.09

-.03

-.12

-.03

-.16

.04

-.13

.02

-.09

Gender

123

-.12

-.10

-.01

-.16

-.07

-.02

-.11

-.16

-.24**

Age

117

.10

.01

-.07

.02

-.13

.08

-.04

-.03

.10

CSE

46

-.03

.03

-.01

-.04

-.02

.10

.06

.26

.15

LP

46

-.17

-.02

-.28

-.25

-.17

.28

-.21

-.23

-.06

p < .05 (2-tailed) ** p < .01 (2-tailed) Source: Own Survey.

36

Results

In the worst-case situation, fewer and weaker significant correlations were found compared to the best-case situation. As shown in table 6, the overall OC dimension demonstrated only one significant correlation with the TRO dimension, which shows that with an increasing score on OC (an increasing preference for the developer style) participants tended to perceive higher levels of TRO in their worst-case climate. Looking at the subscales of OC reveals that this relationship might be mainly driven by the NV aspect. The NV and SA aspects of OC demonstrated also a significant negative correlation with the CON dimension, indicating that with a decreasing score on NV and SA (an increasing preference for the explorer style), participants tended to perceive higher levels of CON in their worst-case climate. Even though this correlation is not significant for the overall OC dimension, the NV and SA aspects might have an impact on the perception of the CON dimension in the worst-case situation. Furthermore, a significant correlation was found between the MP dimension of VIEW and perceptions of the RIT dimension in the worst-case situation, suggesting that with a decreasing score on the MP dimension (an increasing preference for the external style), participants tended to perceive higher levels of RIT in their worst-case situation. This relationship has about the same magnitude and direction as the relationship of MP and RIT in the best-case situation. No significant correlations were found between the WD dimension of VIEW, age, gender, and CSE with any of the climate dimensions. However, LP demonstrated a significant negative correlation with RIT, which shows that participants, who did not have children, perceived higher levels of RIT in their worst-case climate. Nevertheless, these findings should be interpreted cautiously given the low sample size.

Quantitative Results

37

Table 6: Correlations with Worst-Case Climate Perceptions Worst-Case Climate Dimensions Variable OC -

N 123

NV SA SS

CHI

FRE

TRO

IDT *

PLH

CON

IDS

DEB

RIT

.02

-.01

.18

-.01

.16

-.16

.05

.11

.04

.01

.03

.18*

.00

.15

-.18*

.06

.09

.06

.01

.01

.14

.07

.10

-.20*

.07

.10

.03

.01

-.07

.12

-.10

.14

-.03

-.02

.09

.03

MP

123

-.11

-.07

-.04

-.02

-.09

-.08

-.03

-.07

-.19*

WD

123

.05

-.03

.11

-.01

.03

-.05

.03

.01

-.04

Gender

123

.15

-.05

.14

-.08

.11

-.03

-.02

.04

-.03

Age

117

.13

.03

-.02

.01

-.06

.10

-.05

-.02

.02

CSE

46

.00

.09

.03

-.07

-.03

.02

-.03

-.01

-.08

LP

46

.23

.19

.03

.09

.07

-.06

.03

.06

.37*

*

p < .05 (2-tailed) ** p < .01 (2-tailed) Source: Own Survey.

Another way to look at the correlations between individual differences and best- and worst-case climate perceptions is to examine the magnitude of the differences between the best- and worst-case situation. Table 7 shows the correlations between individual differences and the magnitude of the differences between the perceptions of climate in the best- and worst-case situation. The analysis revealed 19 statistically significant negative correlations. Most of these were, again, discovered with the OC dimension of VIEW, indicating that with a decreasing score on OC (an increasing preference for the explorer style), participants tended to perceive a larger difference between their best and worst-case climate concerning the CHI, TRO, PLH, IDS, DEB, and RIT dimensions. The results of the subscales indicate, that the NV aspect might have caused the relationship between OC and CHI, TRO, IDS, DEB, and RIT. The perception of the latter two dimensions might have also been driven by the SA aspect of OC, whereas the SS aspect was mainly correlated with the PLH dimension and demonstrated only a weaker correlation with the RIT dimension than the other two subscales of OC. Furthermore, significant correlations were found between gender and the perception of CHI and RIT, which shows that men perceived a larger difference in these dimensions between a very positive situation and a very negative situation that they have recalled when responding to the SOQ. The same pattern applied for the relationship between LP and RIT, suggesting that participants without children perceived a larger difference in this dimension. Again, this should be interpreted cautiously given the small number of people, who responded to the questions concerning their life phase.

38

Results Table 7: Correlations with Differences in Climate Perceptions Differences between Best- and Worst-Case Climate Dimensions Variable OC -

N 123

CHI

FRE *

-.19

*

NV SA SS

-.18

-.13

TRO

-.04 -.07 -.06

IDT *

-.20

*

-.21

-.14

-.11 -.11

PLH

CON **

-.26

*

-.22

-.18 -.17

IDS

DEB *

-.19

*

-.21

RIT *

-.24**

*

-.28**

*

-.23

-.22

-.12

-.16

-.16

-.14

-.22

-.22*

**

-.10

-.13

-.17

-.19*

-.16

.00

-.12

-.05

-.28

MP

123

.00

-.06

-.01

-.12

-.02

-.08

-.11

.00

-.03

WD

123

-.10

.01

-.16

-.02

-.13

-.07

-.12

.01

-.05

Gender

123

*

-.20

-.03

-.11

-.07

-.13

-.01

-.07

-.15

-.18*

Age

117

-.04

-.01

-.04

-.03

-.05

.03

.00

-.01

.07

CSE

46

-.02

-.05

-.02

.01

.01

-.05

.07

.17

.18

LP

46

-.28

-.18

-.20

-.27

-.17

-.24

-.18

-.20

-.32*

* p < .05 (2-tailed) ** p < .01 (2-tailed) Source: Own Survey.

The correlational analysis was just a first step to explore possible relationships between individual differences and climate perceptions. In the next section, several MANOVAs were conducted to shed more light upon the impact of individual differences on the perceptions of organizational climates for creativity. 4.1.4.2 Comparison of Climate Scores between different Groups The impact of each source of individual differences on the overall perception of the climate for creativity in the best- and worst-case situation as well as the perceived difference between these situations was examined by conducting a MANOVA in each case which was followed by several ANOVAs in order to examine potential differences in each of the nine climate dimensions. The results of all the different MANOVAs are displayed in appendix 14. As with the correlations, most of the statistically significant results were found in the OC dimension of VIEW. In fact, of all the MANOVAs that were conducted, the influence of the OC dimension on the overall perception of the climate in the best-case situation was the only one that reached statistical significance (L=.865; p