Is this change personally beneficial for me as a
middle manager?
“The influence of change process antecedents on middle managers’ readiness
for change mediated by their personal valence”
Master Thesis MSc BA – Change Management University of Groningen, Faculty of Economics and Business
November 15th, 2012
D. Aalbersberg S1724916 Aquamarijnstraat 163 9743PD Groningen d.aalbersberg@student.rug.nl Supervisor University: Drs. H.P. van Peet Second Assessor University:Dr. C. Reezigt
Supervisors Field of Study: D. van Vliet
G. de Boer
Acknowledgements:
2
Is this change personally beneficial for me as a
middle manager?
“The influence of change process antecedents on middle managers’ readiness
for change mediated by their personal valence”
ABSTRACT
The aim of this study on middle managers’ perceptions of change was to explore the mediating effect of personal valence between change process antecedents and the readiness for change dimensions (i.e. general cognitive readiness and specific attitudinal readiness). Involvement in the change process (i.e. perceived communication and participation) and management support (split into top management – and change agent support) were gauged as change process antecedents. Data was gathered through distributing internet surveys among Dutch middle managers of a large organizational segment, who were involved in a large-scale change process. This resulted in 39 valid respondents. Regression analyses were used to test the hypotheses. The outcomes showed that middle managers’ personal valence mediated the relationship between the perceived involvement in the change process and general cognitive readiness for change. Accordingly, this study found that perceived personal benefit is a more proximal determinant of the general cognitive readiness for change than the perceived involvement in the change process. The limitations of this study, the implications of these results, and possibilities for future research are discussed.
3
TABLE OF CONTENTS
1. INTRODUCTION ...5
1.1 Organization under study... 6
Change Project. ... 6
1.2 Research focus ... 9
Research question... 10
1.3 Reader guide ... 10
2. THEORETICAL FRAMEWORK...11
2.1 Underlying frameworks: Four theoretical lenses on change... 11
Nature of change. ... 11
Level of change. ... 12
Positive vs. negative focus on change... 13
Research method. ... 14
2.2 Readiness For Change ... 15
Multidimensional view... 16
2.3 Change Process Antecedents ... 17
Participation. ... 17
Communication. ... 18
Involvement in the change process. ... 19
Management Support. ... 20
2.4 Mediator: Personal Valence ... 23
2.5 Conceptual Model ... 25
3. METHODS ...26
3.1 Data collection ... 26
3.2 Measures ... 28
Dependent variables: Readiness for Change dimensions... 29
Independent variables: Change process factors. ... 29
Mediator: Personal valence. ... 29
Finalization of survey... 30
3.3 Data analysis ... 31
Exploring data. ... 31
Factor analysis and reliability analysis... 31
Assumptions check for parametric tests... 37
4
4. RESULTS ...39
4.1 Descriptive and correlation analysis... 39
4.2 Regression analysis (hypotheses testing) ... 41
4.3 Summary of results ... 45
5. DISCUSSION ...46
5.1 Implications... 50
5.2 Limitations and future research ... 51
5.3 Conclusion... 54
6. REFERENCES...55
7. APPENDICES...64
Appendix A: Questionnaire including letter of introduction ... 64
Appendix B: Actions to boost the response rate... 67
Appendix C: Boxplots ... 68
Appendix D: Histograms and K-S test ... 69
Appendix E: Assumption checks of regression models... 70
5
1. INTRODUCTION
Change rules the world! Globalization, e-business, intensified competition, and the increased pace at which innovations are introduced have made it extremely essential but also tremendously challenging for organizations to implement changes in strategy, structure and culture (Burke & Trahant, 2000; Gordon, Stewart, Sweo & Luker, 2000). Many aspects have been identified and recommended to increase the success of the change implementation. Interestingly, with major organizational changes, new complex systems and structures have attracted management’s attention, whereas the people (soft) factor has often been neglected (Clegg & Walsh, 2004). Meanwhile, several studies state that if individuals in an organization are not ready or motivated to change, the organizational change is doomed to fail (George & Jones, 2001; Porras & Robertson, 1992; Schein, 1980). From this observation, readiness for change is one of the crucial stages that individuals in an organization need to go through in order to enable successful implementation of change (Armenakis, Harris & Mossholder, 1993; Cunningham, Woodward, Shannon, MacIntosh, Lendrum & Rosenbloom, 2002; Eby, Adams, Russell, & Gaby, 2000).
Madsen, Miller, & John (2005) state that although much has been done in the general area of change, little research has been executed in the management area related to readiness for change at the individual level. Neves (2009) states that empirical studies measured readiness for change as a one-dimensional construct (Cunningham et al., 2002; Jones, Jimmieson, & Griffiths; 2005), and these studies found encouraging results. More specifically, these studies assumed readiness for change as a cognitive attitude of the change recipients towards a change effort. However, a few years ago Holt, Armenakis, Field and Harris (2007) and Armenakis, Bernerth, Pitts, and Walker (2007a) developed scales to measure readiness for change. Their results suggest that readiness for change is; in fact, a multidimensional construct.
6
One of the purposes of this study is to assess change recipients’ readiness for change during an organizational change project by using the instrument of Bouckenooghe et al. (2009).
Hereafter, the organizational change project under this study will be described, being considered as pivotal for this study. Subsequently, the research focus will be provided.
1.1 Organization under study
This study was performed at KPN, a large Dutch organization (± 30,000 FTE) situated in the telecommunications industry. The strategy of KPN strives for continuous improvement of cost/productivity and for acceleration of the catching up in terms of quality and customer experience. This is a major challenge for the operational teams (KPN N.V., 2012a).
At the same time, KPN stated a few barriers for achieving their strategy within the residential part of the consumer market (± 4,500 FTE): This segment ‘Consumer residential’ consists of all the fixed connections within the consumer market, including the fiberglass chain and the multiplay chain. One of these barriers was that the management of residential focused too much on the ‘what’ instead of the ‘how’ during their controlling. A second barrier was that the translation of desired output / results to good KPI’s did not work (KPN N.V., 2012a).
KPN thought that the lack of a coherent “Way of Working” and the lack of an End-to-End change approach on the various teams, supply chains and business units were important barriers for further improvement (KPN N.V., 2012a).
Change Project.
Therefore, KPN has launched project WoW (Way of Working) as a change approach that… • ..allows the new strategy, values and customer promises to become reality.
• ..allows the organization itself to focus on the ‘how’ by monitoring the appropriate KPI’s. • ..combines internal and external best practices into a new WoW toolkit.
• ..is implemented team per team within entire Residential.
• ..facilitates intensive coaching for middle management by internal change agents (Navigators) that implement the WoW elements team per team (KPN N.V, 2012a).
Project WoW focuses both on end-to-end (e2e) improvements and on the roll out of a standard way of working by implementing the WoW Toolbox. The e2e improvements are formulated by using five lenses. For each lens, there are tools from the WoW toolbox that will be implemented during the change. These lenses and associated WoW tools are described below (KPN N.V. 2012b):
1) Voice of the Customer: The customer is pivotal in everything we do:
• Work values derived from customer values
2) Process efficiency: Simplicity in processes, uniformity in execution; regular heartbeat for changes:
7 • Standard work instructions (sharing best practices)
3) Performance Management: Objectives are translated into “How-KPI’s”; consistency across the
various functions in the chain: • Week start and day start • Visible How-KPI’s
4) Organization and Skills: Clear control and roles in the chain”
• Team problem solving (kaizen) • Certified meeting structure
5) Mindsets and Behavior: Model behavior, coaching and commitment by managers; focus on
employees’ strengths.
• One-to-one performance coaching (team manager – employees) • Team barometer
During the execution of this study, the first WoW steps were already made in the fiberglass chain, and the first successes of this WoW change approach became clear. At that time, KPN wanted to implement WoW in the whole scale of Residential. The expansion of WoW was done by a further rollout in several chains and departments of Residential. For this further rollout the group of navigators was enhanced from 5 to 15 navigators early 2012, with further expansion to 30 navigators mid-2012 (KPN N.V., 2012b). These navigators serve as internal change agents/personal coaches for the middle managers (chain- and line managers) that undergo the WoW change as recipients. The current ambition in 2012 was to implement WoW in a large part of Residential (± 2900 FTE in 2012), and in 2013 whole KPN Residential will be changed according to the WoW approach.
Consequently, KPN Residential found and still finds itself in a major change that has an impact on both the structure (new procedures and processes) and culture (mindset change: More controlled and more focused on team work) of the chains and departments of Residential.
The rollout of this WoW change is done in multiple waves; in every wave the same change process is used to implement WoW top-down in whole residential. See table 1 for a detailed specification of these waves:
TABLE 1
Specification of the KPN WoW waves
Aug ’11 – Jan '12 Feb ’12 – May ‘12 Jun ’12 – Sep ‘12 Oct ’12 – Jan ‘13
Waves Wave 1 Wave 2 Wave 3 Wave 4
8
Meanwhile, KPN already evaluated and measured the outcomes of Wave 1. They indicated successes due to the change of WoW. However, they also found some points of improvement for making the next waves more effective; most important, the readiness for change of the chain- and line managers (middle managers) of Residential that undergo the WoW change could be enhanced at the start of the change implementation. KPN believes that the change readiness of the Residential middle managers is of major importance, because these middle managers will be the WoW communicators and drivers for their subordinates later in the change process (during the implementation of the WoW tools in the departments).
Besides, the internal change agents (Navigators) play a crucial role during the WoW changes within the particular departments. This role is especially important at the beginning of the rollout of WoW. One Navigator is responsible for one or two departmental rollouts of WoW.
Procedure of a WoW wave rollout.
This rollout lasts for ± 16 weeks. In these 16 weeks, the responsibility of ‘driving the change’ is being handed over from Navigator to the middle manager. See figure 1 for a visual presentation:
FIGURE 1
Who is driving the change of WoW?
At the beginning of a WoW wave, the navigators – together with the WoW program management - communicate the change message towards the middle managers, and the navigators try to involve these middle managers for participating during the WoW implementation. After 7 weeks a commitment session takes place, in which the middle manager presents his or her level of commitment towards the new WoW (tools). From this moment, the implementation of the WoW tools starts and the recipient manager will serve as change agent for his or her subordinates.
KPN wants to assess the middle managers’ change readiness after the introduction of the WoW change and prior to the implementation of WoW improvements and toolkit. This moment falls in the 7th week (the commitment week). In this study, the focus was on the 53 chain- and line managers
9
factors (e.g. support by the navigator) positively influence the readiness for change of the middle managers that undergo the WoW change.
1.2 Research focus
By combining both the abovementioned theoretical and practical introduction, the readiness for change of the managers is considered as central in this study. This study distinguishes readiness for change as a triadic attitude that comprises the emotional readiness dimension, the intentional readiness dimension and the cognitive readiness dimension.
Oreg et al. (2011) recently published a 60-year review of quantitative studies on change recipients’ reactions to organizational change, including readiness for change. They unraveled a model of the tridimensional reactions to change, reaction antecedents, and change consequences. Oreg et al. (2011) identified five primary antecedent categories for readiness for change:
(a) Change recipients characteristics, (b) internal context, (c) change process, (d) perceived benefit/harm, and (e) change content. The importance of these five antecedent categories for readiness for change has been widely acknowledged (e.g. Armenakis & Harris, 2002; Bouckenooghe et al., 2009; Judge, Thorsen, Pucik & Welbourne, 1999; Self, Armenakis & Schraeder, 2007).
Although change researchers studied a plethora of antecedent variables, the analysis of their interrelationships, including mediation and moderation effects, requires much greater attention (Armenakis et al., 2007a; Oreg et al., 2011). In fact, Oreg et al. (2011) specifically recommend future research on the potential mediating role of perceived benefit/harm in the relationship between change process antecedents and recipients’ readiness for change. In other words, perceived benefits/harm can be a more proximal determinant of change recipients’ reactions than the change process antecedents. They draw this conclusion tentatively, since only few studies explored this possibility (e.g. Johnson, Bernhagen, Miller & Allen, 2006) and more research is necessary. For measuring this tentative conclusion of Oreg et al. (2011) in this study, change process antecedents need to be assessed. Below, the change process antecedents that will be assessed in this study will be given.
10
crucial role in change projects and are prevalent change process variables. (3) These change process antecedents were present in the KPN change project WoW and could be influenced by the change agents during the following change phases and WoW waves.
Moreover, this study tests whether or not the perceived benefit/harm by the change recipients acts as mediator between change process antecedents and recipients’ readiness for change, for the reason Oreg et al. (2011) recommended research on this potential mediation relationship. Furthermore, personal valence – as developed by Holt et al. (2007) - will be used for measuring the extent to which the change is perceived as beneficial or harmful.
Ultimately, this study aims to establish a better understanding of readiness for change by assessing that the reason for which the change process may eventually influence change recipients’ readiness for change can be because of the influence that the change process may have on recipients’ perceived benefits or harms from the change.
The change recipients in this study are middle managers of the organization segment under study. Middle managers may suffer from a negative reputation and are considered to have a negative impact on organizational changes (Balogun, 2003; Huy, 2001). However, several studies indicate that middle managers play a crucial role during organizational changes (Balogun & Johnson, 2004; Balogun & Johnson, 2005). This study focuses on the middle manager’s perception on the change process, personal benefit, and the readiness for change.
In order to investigate these relations between middle managers’ perceptions of change process antecedents, personal benefits or harms, and readiness for change the following research question is proposed:
Research question
“To what extent do change process antecedents influence the multiple dimensions of middle managers’ readiness for change and does personal valence mediate this relationship?”
1.3 Reader guide
The subsequent chapter of this study starts with describing the underlying assumptions of this study. It contains a theoretical framework and the hypotheses of this study. Additionally, a conceptual framework for analyzing the research question is presented.
11
2. THEORETICAL FRAMEWORK
This chapter sets out the variables and gives a high level of literary context to this study. At first, the four basic assumptions of this study will be given. Subsequently the dependent variable - readiness for change - will be discussed. This will be followed by a discussion of the independent variables: the change process antecedents. Finally, personal valence – perceived benefit/harm – will be discussed, leading into the hypotheses of this study and an overview of the studied relations presented in a conceptual model.
2.1 Underlying frameworks: Four theoretical lenses on change
Bouckenooghe (2010) set forth a map that contains four ways of looking at change, so called, lenses, representing four dualities. These four lenses are (a) the nature of change (i.e., planned or episodic change vs. emergent or continuous change; Porras & Silvers, 1991; Weick & Quinn, 1999); (b) the level of change (i.e., individual level or collective level; Aktouf, 1992; Bray, 1994; Pettigrew, Woodman, & Cameron, 2001; Poole & Van de Ven, 2004); (c) the positive versus negative focus on change (i.e., negative problem-solving view vs. positive potential view; Abrahamson, 2004a, 2004b; Seo, Putnam, & Bartunek, 2004); and (d) the research method (i.e., variance or process methods: Mohr, 1982; Van de Ven & Poole, 2005).
Below, these four mentioned dualities are explained in order to provide the basic assumptions underlying this study (positioning of this study).
Nature of change.
The nature of change or how change evolves over time is one of the most important dualities in change research (Porras & Silvers, 1991). Weick and Quinn (1999), on the one hand, discern change as episodic, planned, discontinuous, and, on the other hand, change as continuous, evolving, emerging, and incremental. Planned change is an intentional intervention plan for changing an organization and is characterized as deliberate, with a purpose and is systematic (Burnes, 2009; Lippit, Watson & Westley, 1958; Tenkasi & Chesmore, 2003). Continuous or emergent change, though, is a change that aim to be evolving and ongoing (Burnes, 2009).
In terms of the process theories of change defined by Van de Ven and Poole (1995), planned change reflects the teleological theory. According to Van de Ven and Poole (1995), the teleological theory states that organizations are driven by a purpose or objective, they adapt and change to achieve their purpose. These changes follow a programmatic step by step sequence.
12
Level of change.
Organizations and change are multi-level phenomena (Klein & Kozlowski, 2000; Pettigrew et al., 2001). A growing group of scientific authors states that change theories should include multiple levels and clarify relationships among these levels (Poole & Van de Ven, 2004). Besides, change is perceived differently at each level when it rolls out in the organization (Caldwell, Herold, & Fedor, 2004).
The level of change can be split in the individual level and collective level focus. According to Bouckenooghe (2010) the individual level encompasses interventions aimed at individuals, whereas the collective level refers not only to interventions aimed at team level but also system wide interventions aimed at the whole organization. The individual-centered level brings the individual into the focus of the change. This individual-centered approach assumes that organizational change is only possible under the condition that recipients are ready to change their attitudes regarding to the current situation, and therefore, a key element of success of organizational changes is individual readiness for change (Bouckenooghe, 2010).
Besides this difference between individual level and collective level, another difference exists in the hierarchical level of the change recipients.
Change recipients: Middle managers.
In this study, the change recipients are the middle managers (or line managers) of the organization. There is no general definition of a middle manager; however, a frequently used definition in (middle) management literature is Likert’s definition (1961): Middle managers are linking pins. In other words, the role of the middle manager is a tie between top management and operational workers (Vogler, 2007).
Middle managers are now often expected to be key drivers of change (Caldwell, 2003). Thus, middle managers have to implement corporate change initiatives, while also confirming that the organization is maintaining its core business (Balogun, 2003). As a result – which is underlined by numerous studies – middle managers play a crucial role during organizational changes (Balogun & Johnson, 2004; Balogun & Johnson, 2005; van Vuuren & Elving, 2008; Elving & Bennebroek Gravenhorst, 2005; Hales 2005; Huy, 2002). Namely, the middle managers are linking pins between top management level; in which change is initiated, and the individual level; in which the emotional needs of the work force are recognized and balanced (Balogun & Johnson, 2005).
13
Middle managers’ key challenge is the ability to manage these roles simultaneously (Bryant & Stensaker, 2011). According to Bryant and Stensaker (2011), expecting middle managers to apply these roles simultaneously may be placing new, and/or unfair, expectations upon middle managers as individuals. In times of change initiatives, these individuals may also be faced with personal challenges towards the change content. If planned organizational change is initiated at the top of the organization, such as restructuring, it demands middle managers to act on decisions made by top management. In essence, this makes the middle manager a change recipient; the change has to make sense to him/her (Balogun and Johnson, 2004; Caldwell, 2009). For instance, middle management may be implementing and communicating changes without genuinely believing in or supporting the change initiative (Bryant & Stensaker, 2011). In this way, some middle managers may not align words and deeds, and may be inconsistent (Valentino, 2004). Therefore, middle managers will need to go through personal change before guiding others through change (Balogun, 2003).
Consequently, the role of middle management in undergoing and implementing planned change suggests a number of challenges related to the dual role of being change recipient and change agent. Moreover, middle managers may find themselves in a position in which they have to critically think by themselves whether or not they will be able to be ready and to support the planned change, and promote the planned change (Bryant & Stensaker, 2011).
The focus of this article will be on the role of the middle manager as being a change recipient and their readiness is assessed prior to their role of change implementer / communicator among his or her subordinates.
ASSUMPTION 2a: In this article, attitudes toward change are examined from an individual-centered perspective.
ASSUMPTION 2b: In this article, the change recipients are middle managers in an organization.
Positive vs. negative focus on change.
The dualism of positive vs. negative focus is derived from Seo et al. (2004). They described the dimension positive vs. negative focus as one of the critical characteristics of the change process, and identified where to focus attention on driving the energy of the organization to provoke change. The negative focus centers attention to approaches that stress negative aspects of organizations, such as current problems; however, the positive focus draws attention to the positive reasons for change, such as opportunities (Seo et al., 2004).
14
change as overcoming problems (i.e., resistance to change) in contrast to thinking about change as seizing opportunities (i.e., readiness for change).
For a long time, the negative view of human and organizational function has dominated in the field of organizational change (Cameron & Powley, 2008). However, a step in the direction of the positive focus is found by Luthans (2002) and Cameron, Dutton and Quinn (2003). In recent years, the positive focus received focused attention as is shown by a special issue in the Journal of Applied Behavioral Science (Cameron & Powley, 2008).
This dualism is also reflecting the distinction between the first generation Organizational Development approaches – e.g. action research and sensitivity training- from second and third generation OD approaches (i.e., learning organizations and appreciative inquiry). Third generation OD proponents often contrast their approach with the problem solving view of action research by arguing that action research pays too little attention to positive aspects as potential drivers of change (Cooperrider & Srivastva, 1987).
ASSUMPTION 3: Even though the negative focus has dominated in the field of change, attitudes toward change, in this article, are embedded in the positive psychology approach (i.e., readiness for change).
Research method.
Research into change distinguishes between two research strategies: (a) The variance strategy and (b) the process strategy (Bouckenooghe, 2010). The dominant literature and research on ‘attitudes toward change’ have adopted the variance strategy to determine how a set of independent variables positively influences the level of readiness for change (e.g. Bouckenooghe et al., 2009; Holt et al., 2007). Therefore, the variance strategy is also adopted in this study.
The variance strategy concentrates on variables that represent the important aspect of the subject under study and focus on cause-effect relationships. Variance research supports predictive models capable of explaining variation in such outcome measures as resistance to change or readiness for change (Bouckenooghe, 2008). Furthermore, variance research relies mainly on survey-based research designs that are grounded in the statistical general linear model (Van de Ven & Poole, 2005). The research method will be more elaborated in the methodology chapter.
15 2.2 Readiness for Change
A major interest in the existing literature is the identification of the variables that promote or facilitate change within organizations (e.g. Coch & French, 1948; Lewin, 1947). Early theorists such as Lewin (1947) examined how the social field surrounding individuals acted to obstruct or facilitate change within organizations. Nowadays, there is a large variety in perceptions, reactions and attitudes towards change, which is also reflected in the work of scholars using many ways for conceptualizing people’s reactions toward change (Oreg et al., 2011).
Readiness for change, presented as the opposite of resistance, was first mentioned by Jacobson (1957). Although readiness for change was introduced in 1957, a comprehensive and theoretically sound conceptualization of the variable was only defined in the 1993 by Armenakis, Harris and Mossholder. Surprisingly, between 1957 and 1993 the number of studies on readiness for change is scant in comparison with the work published on resistance to change (Bouckenooghe, 2008). Besides, in addition to both readiness for change and resistance to change, many other constructs have entered the stage (Oreg et al., 2011). As a consequence, below, clarity will be created in the middle of this maze of concepts by defining readiness for change and its multiple dimensions.
Empirical research states that recipient’s readiness for change is an important driver of change success – if there is no belief among the recipients that change is necessary, or if the recipients think the organization will not be able to change then change initiatives may fail (Armenakis et al., 1993; Cunningham et al., 2002; Eby, Adams, Russell, & Gaby, 2000).
As mentioned before, Armenakis et al. (1993) created a theoretically sound conceptualization of readiness for change. A conceptualization that is still the most well accepted definition in literature (e.g., By, 2007; Chonko, Jones, Roberts, & Dubinsky, 2002; Eby et al., 2000; Holt, Armenakis, Harris, & Feild, 2007; Jones, Jimmieson, & Griffiths, 2005; Madsen et al., 2005; Weeks, Roberts, Chonko, Lawrence, & Jones, 2004) and they formulated readiness for change as: “Individuals’
beliefs, attitudes, and intentions regarding the extent to which changes are needed and the organization’s capacity to successfully undertake those changes” (Armenakis et al., 1993, p. 681).
Furthermore, Armenakis et al. (1993) describe that when readiness for change exists, an organization is prepared to embrace change and resistance to change is reduced.
16
Despite the salience of the cognitive component in the readiness for change literature, among others Bouckenooghe et al. (2009) emphasize the importance of covering all three basic components in the assessment of readiness for change. For this reason, the recipients’ reactions toward change should be captured along at least three dimensions; affect/emotion, cognition, and intention (Piderit, 2000; Oreg et al., 2011).
Multidimensional view.
Bouckenooghe et al. (2009) stress that readiness for change is conceived as a multifaceted concept that comprises an emotional dimension of change, a cognitive dimension of change, and an intentional dimension of change. It is a more accurate view of reality in acknowledging that recipients’ change reactions involve affections, cognitions and intentions rather than just perceiving readiness as one-dimensional phenomenon (Bouckenooghe et al., 2009; Oreg et al., 2011). According to Bouckenooghe (2008), Oreg et al. (2006), and Piderit (2000) this tridimensional view is essential for the reason that the ways in which affective, cognitive, and intentional responses become manifest coincide. Bouckenooghe (2008, p. 24) gives an example: “people may exhibit feelings in support of
change (affect), however their risk-benefit analysis of the change outcome (cognition) might inhibit their behavioral intentions.”
Bouckenooghe et al. (2009) derive this multi-dimensional view from Piderit (2000). Piderit (2000) initiated the idea to measure reactions to change on these three dimensions. She grounded this idea on the early attitude theorists Katz (1960) and Rosenberg and Hovland (1960). They argue that attitudes are structured along three dimensions that are labeled as cognitive, emotional, and intentional.
This conception is also known as the tripartite view of attitudes (Ajzen, 1991), which is another argument for the adoption of the tridimensional view in reactions to change; researchers and practitioners are able to easily ground this tridimensional view in the Theory of Planned Behavior (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975). This Theory of Planned Behavior provides a designation of the relationships and dynamics between the core components of readiness for change, and explains how these key dimensions are affected by individual level (i.e., efficacy) and context level (i.e., social pressure). Azjen (1991) states that people’s intentions are determined by their feelings and thoughts, social pressure, and efficacy.
The three components are defined as follows: Emotional readiness for change is “the affective
reactions toward change” (Oreg, 2006; Piderit, 2000). Cognitive readiness for change is “the beliefs and thoughts people hold about the change”. For instance, what are the benefits or disadvantages
caused by the change? Intentional readiness for change is the extent to which employees are prepared
17
Moreover, a review study on measurement of readiness for change by Holt, Armenakis, Harris and Field (2007b) concluded that readiness for change in some cases is conceived as an attitude towards a specific change project (Armenakis et al., 1993; Backer, 1995), whereas in other cases it is provided as a general cross-situational climate variable of organizations (Beckhard & Harris, 1987; Stewart, 1994). Accordingly, Bouckenooghe et al. (2009) distinguishes readiness for change both as a change specific state (i.e. emotional and intentional readiness for change) and a general characteristic (i.e. cognitive readiness for change).
This study uses the readiness for change-concept of Bouckenooghe et al. (2009). This implies that readiness for change consists of three attitudinal dimensions: Emotional-, intentional-, and cognitive readiness for change. In their concept, the emotional and intentional readiness are focused on the specific change, and the cognitive readiness is focused on change in general.
2.3 Change Process Antecedents
The factors that influence readiness for change are manifold (Oreg et al., 2011). According to Oreg et al. (2011) these factors (or antecedents) are the reasons for the extent of recipient’s readiness for change. The focus of this study is on the change process antecedents and not on the contextual antecedents.
The change process is the “how” factor of change (Self et al., 2007). Oreg et al. (2011) state that the change process antecedents involved the manner in which change was implemented and are the most frequently studied category of antecedents of reactions to change. Forty-two of the seventy-nine studies in Oreg et al.’s (2011) review included variables that pertained to the change process. They classified these variables into five process categories. The three most prevalent categories are participation, communication and management support (during change). These three variables are subject to this research for their assumed influence on readiness for change, where participation and communication are combined as one variable and labeled as involvement in the change process. First, participation and communication will be examined separately and thereafter they are combined as one variable. After that, management support will be discussed.
Participation.
18
Such participation creates a contribution and control over the change (Armenakis & Bedeian, 1999). As a rule, change recipients who experienced high levels of participation tended to report higher readiness for change (Amiot et al., 2006; Armenakis et al., 1993; Coch & French, 1948; Coyle-Shapiro, 2002; Fishbein & Azjen, 1975; Oreg et al., 2011; Sagie & Koslowsky; 1994; Steel & Lloyd, 1988).
Moreover, without this participation sustainable change is unlikely (Armenakis & Harris, 2009). Cunningham et al. (2002) described that an employee’s perceived control over the change process is a necessary condition for creating readiness for change. Participation gives the recipients the opportunity to have an impact on change (Devos et al., 2007). According to McNabb and Sepic (1995), lack of recipient participation is a major cause of disappointing results of organizational renewal. For such renewal efforts to succeed, recipients have to believe that their opinions have been heard and considered (Reichers, Wanous, & Austin, 1997).
Lines (2004) describes participation as the extent to which staff members are involved in and informed about decisions that directly concern them; decisions about organizational change included. According to Bouckenooghe and Devos (2007), self-discovery through active participation in decision making, combined with the symbolic meaning of change leaders demonstrating their confidence in the wisdom of recipients, can produce an authentic sense of control over the organizational change and hence engender increased readiness for change. Therefore, participation is generally believed to increase the acceptance of proposed changes and they found that the recipients who actually participate during change are more ready for change than recipients who did not participate (Holt et al., 2007).
However, literature shows two different measurements of participation in the change process. Namely, the measurement between participants and non-participants (e.g. Holt et al., 2007) or measurement of change recipients’ perceptions about the participation in the change process (e.g. Bouckenooghe et al., 2009).
Communication.
19
Paulsen, & Tourish, 2004). Poorly managed change communication often results in scattered rumors and uncertainty (Schweiger & DeNishi, 1991), which often overstate the negative aspects of change and negatively influence readiness and build resistance to change. Accordingly, management should try to keep such rumors to a minimum and change recipients have to be informed correctly (Reichers et al., 1997).
In Oreg’s (2006) study; however, contrary to his beliefs, additional information about change coincided with negative evaluations of the change (Oreg, 2006). It is not merely the amount of communication and information that determines reactions to change but also the content and quality of this communication and information (Oreg et al., 2011).
Bouckenooghe & Devos (2007) focused on the quality of communication. Their analysis of 17 papers about a.o. process antecedents for readiness for change indicated quality of change communication as the most cited process factor of readiness for change. Communication assists recipients in making sense of the coming change, makes the coming change more salient, and helps reframe the recipients (Weick, 1995). Bouckenooghe et al. (2009, p. 599) define quality of change communication with the aid of Miller et al. (1994): “How change is communicated and the clarity, the frequency and openness
determine whether or not communication is effective”. The quality of communication contributes to
the awareness and justification of the need and reason for change. Accordingly, it reduces the change related uncertainty of the change recipients and positively shapes their readiness for change.
Interestingly, several studies on participation and quality of communication - perceived by change recipients during a change process - indicate that these two variables are closely related (e.g. Lines, 2004; Oreg et al., 2011). In fact, some studies really combine these two process factors into one variable (Axtell et al., 2002; Bouckenooghe et al., 2009). Bouckenooghe et al. (2009) label this variable involvement in the change process. Furthermore, they show that the change recipient perceives the communication and the participation as a whole, and does not make a distinction between these two change process factors.
Since the focus of this study is on the perceptions of the change recipient, the process factors communication and participation will be combined into one factor, in accordance with Bouckenooghe et al. (2009):
Involvement in the change process.
20
Furthermore, Axtell et al. (2002) describe the variable “exposure to change” which corresponds with the variable “involvement in the change process” (Bouckenooghe et al., 2009) in a way that they both include the perceptions of change recipients on participation and communication during the change process. According to Axtell et al. (2002), all too often, recipients’ first meaningful involvement in the change takes place when it is implemented. Axtell et al. (2002) mention this as exposure to change, which is a very important stage in any change process. Exposure enables familiarization with the change situation, which is also an important component of participative approaches (Sagie & Koslowsky, 2000). Besides participation, exposure may similarly involve greater information and better communication (Miller & Monge, 1985), and may offer the ultimate realistic preview through direct usage of the change goals and procedures for the change recipients (Scheiger & DeNishi, 1991). Hence, exposure to change both includes perceptions of the recipient towards participation and communication in the change process. Their result indicates that greater exposure will increase recipients’ openness to change (Axtell et al., 2002).
Concluding, the abovementioned substantiations on participation and communication indicate a positive effect of participation and quality of communication on readiness for change. Besides, Bouckenooghe et al. (2009) show that perceived participation and communication during the change process can be seen as one variable “involvement in the change process” and they indicate that greater involvement in the change process lead to more recipients’ readiness for change. Therefore:
Hypothesis 1 (H1): A change process that is characterized by more involvement of the change
recipients will influence readiness for change more positively than a change process that is characterized by less involvement of the change recipients.
Management Support.
During major changes the management of the organization is an important component in the change process, in particular, management’s behavior (Bouckenooghe, 2007). Establishing a need to change is one of the first important steps to follow in implementing change (Armenakis and Bedian, 1999; Armenakis and Harris, 2001; Bernerth, 2004; Galpin, 1996; Judson, 1991). Management support is key for this process (Kotter, 1995; Lakshman, 2005).
Holt et al. (2007) define management support as the extent to which the recipient feels that the organization’s leadership and management are or are not committed to and support or do not support implementation of the prospective change. Such support is distinct from a general supportive atmosphere, often discussed as trust in management, which is a context factor.
21
regardless of the intervention model. These five domains are combined to shape recipients’ readiness for change. They define principal support as the support from change agents and opinion leaders who affect organizational change (Armenakis & Harris, 2002; Armenakis et al., 2007a; Oreg et al., 2011). The term ‘change agent’ may refer to the global (i.e., CEO of the organization) and/or local change agent (i.e., the immediate supervisor or change project manager). Do both of these principals genuinely support the change? (Armenakis et al., 2007b).
Principal/management support is the belief that change agent(s), organizational leaders, one’s direct manager and one’s respected peers demonstrate to the recipients that they support the organizational change and are motivated to see it through to success (Armenakis & Harris, 2002). In this study, the focus is not on the peers as principals.
The rationale for management support is found in the social-information processing (Salancik and Pfeffer, 1978) and social-learning theories (Bandura, 1986). According to Armenakis et al. (2007b), these theories describe the role of interpersonal networks within an organization in influencing the attitudes and beliefs of employees. Evidence may be offered that people have or have not bought-in to the change.
A common phrase associated with the management support is “walking the talk”. Simons (2002) labeled this as ‘behavioral integrity’, which is, alignment or misalignment of words and deeds. In other words, the perception of recipients whether a change leader’s behavior is consistent with his or her spoken word (Armenakis et al., 2007b).
Furthermore, according to Oreg et al. (2011), some studies demonstrated the effect of management support during a particular change on change recipients’ reactions to change (e.g. Amiot et al., 2006; Eby et al., 2000). For instance, Armenakis et al. (1999) and Logan and Ganster (2007) associated management support during change with readiness to change and lower perceived negative effects of the change.
Bouckenooghe et al. (2009) made a distinction between top management support and immediate supervisor support during change.
Top management support.
Organizational members should have the feeling that top management cares for their well-being and is supportive of their concerns about the change (Eisenberger, Huntington, Hutchinson, & Sowa, 1986). Thus, the perceived top management support during change may impact someone’s reaction to the particular change such that it is perceived as more favorably (Eby et al., 2000).
22
change. When top management is not seen as backing the change effort, it is unlikely that a change effort will be successful among the recipients. Besides, Carter, Ulrich and Goldsmith (2005) describe in their book that the top leaders at the organization must not only budget for the change initiative, they must also strongly believe in the initiative and model this behavior throughout the organization. Support from top management has been identified by 88 percent of the contributors (37 authors of case studies) of their book as a critical step in creating readiness for change.
In addition, Devos et al. (2007) stress that change recipients will not take change efforts serious, if top management does not actively support the change process.
Supervisor support.
Besides top management, the immediate supervisor also has to support the change. However, in this study, the change recipients are middle managers and; consequently, the supervisors themselves. The supervisor support literature, as described above, is intended for the workforce as change recipients and not for the middle managers as change recipients. As result, the variable ‘supervisor support’ is not applicable in this study.
For the reason the middle manager is the change recipient, there is more emphasis on top management support and on local change agents during change instead on the direct supervisors. Therefore, change agent support will be described.
Change agent support.
The local change agents have to genuinely support the change as top management does, due to the fact top management appointed these local change agents (Armenakis et al., 2007b). Local change agents can be seen as coaches for change or as change project managers (Armenakis et al., 1993; Bouckenooghe, 2008). These change agents need the ability to lead the change recipients to the change process (Bouckenooghe et al., 2009). Change scholars indicated that change agent support positively influences recipients’ readiness for change (Armenakis et al., 1999; Holt et al., 2007). Therefore:
Hypothesis 2a (H2a): A change process that is characterized by more change agent support will
influence readiness for change more positively than a change process that is characterized by less change agent support.
Hypothesis 2b (H2b): A change process that is characterized by more top management support will
23 2.4 Mediator: Personal Valence
Cobb, Wooten, and Folger (1995) emphasize that, during an organizational change, members of the change target will assess the distribution of positive and negative results, the change’s fairness, and the way in which individuals are treated. According to Armenakis & Harris (2002), members of the change target are interested in what is in there for them personally. Do the recipients perceive the change as personally beneficial or not? (Holt et al., 2007). Kim, Hornung and Rousseau (2011) express it as recipient’s anticipation of positive and/or negative outcomes, which is a well-established motivating force in human behavior.
Organizational change researchers widely acknowledge the importance of positive expectations regarding change outcomes in defining the level of recipient support (e.g. Bartunek et al., 2006; Piderit, 2000). For instance, Armenakis et al. (2007a) recently executed a qualitative review of organizational change publications which reported that nineteen out of forty-five studies included a variable related to the benefits employees anticipated from the change. They define this variable as personal valence, which originated in Vroom’s (1964) work on motivation, and refers to: “the
attractiveness (from the change recipient’s perspective) associated with the perceived outcome of the change” (Armenakis et al., 2007a, p. 488). In order to explain, Vroom (1964) stated that individuals
are likely to engage in behavior they anticipate to lead to an expected or a desired result, which in turn, is linked to positive second-order consequences that they personally value and attempt to attain (Kim et al., 2011).
Moreover, Armenakis and colleagues (2007a) suggest that personal valence can have intrinsic (e.g., more job autonomy) as well as extrinsic components (e.g., financial rewards). However, they dropped the extrinsic component in subsequent versions of their scale, and only focused on intrinsic benefits. Intrinsic benefits/disadvantages are also at the core of Holt et al.’s (2007) construct of personal valence. Kim et al. (2011) state that evidence for the extrinsic benefits of change is less consistent than for the intrinsic benefits. In addition, Kim et al. (2011) describe that efforts to operationalize personal valence need to take into account the specific context characteristics of the particular change and avoid confounding anticipated benefits with one’s job or the organization in general.
Oreg et al. (2011, p. 493) stress that personal valence (“the extent to which the change is perceived as
personally beneficial or harmful”) is a key determinant of whether change recipients will accept or
24
Rousseau, 2007; Johnson et al., 1996). In other words, personal valence was found to be a more proximal determinant of change recipients’ reactions than other antecedents, such as change process antecedents (Oreg et al., 2011).For instance, Johnson et al. (1996) describe that the supportive communication of change agents negatively influences the perceived harm of the recipient. Consequently, the change agents can attempt to control the communication during the change in order to create less uncertainty and more perceived benefit among the recipients.
The study of Holt et al. (2007) describe that participation positively influences personal valence; all participants reported higher mean scores on personal valence than non-participants. Armenakis et al. (2009), additionally, state that participation in change processes with change recipients enhances their valence by accepting them to participate in implementing difficulties they face.
Oreg et al. (2011) describe that the reasons for which the change process (a.o., communication, participation, and management support) may ultimately influence recipients’ reactions - as readiness for change - is because of the influence that process has on recipients’ personal valence. Therefore:
Hypothesis 3 (H3): A change process that is characterized by more positive personal valence will
influence readiness for change more positively than a change process that is characterized by more negative personal valence
Hypothesis 4a (H4a): Personal valence mediates the positive relationship between involvement in the
change process and readiness for change, such that involvement in the change process leads to a more positive personal valence and; consequently, leads to more readiness for change.
Hypothesis 4b (H4b): Personal valence mediates the positive relationship between change agent
support during a change process and readiness for change, such that top management support leads to a more positive personal valence and, consequently, leads to more readiness for change.
Hypothesis 4c (H4c): Personal valence mediates the positive relationship between top management
25 2.5 Conceptual Model
This study investigated seven hypotheses in order to provide an answer for the research question. Below, the conceptual model shows the hypothesized relationships between the change process antecedents, personal valence, and readiness for change. All relationships were hypothesized as positive.
TABLE 2
Summary of study’s hypotheses Hypotheses
H1 A change process that is characterized by more involvement of the change recipients will influence readiness for change more positively than a change process that is characterized by less involvement of the change recipients
H2a A change process that is characterized by more change agent support will influence readiness for change more positively than a change process that is characterized by less change agent support.
H2b A change process that is characterized by more top management support will influence readiness for change more positively than a change process that is characterized by less top management support. H3 A change process that is characterized by more positive personal valence will influence readiness for
change more positively than a change process that is characterized by more negative personal valence. H4a Personal valence mediates the positive relationship between involvement in the change process and
readiness for change, such that involvement in the change process leads to a more positive personal valence, and, consequently, leads to more readiness for change.
H4b Personal valence mediates the positive relationship between change agent support and readiness for change, such that top management support leads to a more positive personal valence and consequently, leads to more readiness for change.
H4c Personal valence mediates the positive relationship between top management support and readiness for change, such that change supervisor support leads to a more positive personal valence and consequently, leads to more readiness for change.
FIGURE 2
26
3. METHODS
This chapter describes the way data has been collected and analyzed. In accordance to the fourth basic assumption earlier made in the literature review of this study; this study was carried out using the variance strategy. Van de Ven and Poole (2005) state this strategy relies mainly on quantitative survey-based research designs that are grounded in the statistical general linear model. Therefore, quantitative survey-based research was used as method to investigate relationships, causal effects between the independent and dependent variables and; eventually, to test the hypotheses.
3.1 Data collection
In this study a self-administered internet survey was carried out via the online program “Qualtrics Survey Software”. An internet survey has advantages over other methods of delivery ( i.e. written survey, face-to-face or telephone interview surveys); such as automatic sequencing of the questions, automatic validation of answers (Thomas, 2004), and maintaining anonymity which in turn decreases the social desirability bias (Kantowitz & Elmes, 2005). This was important for KPN - organization under study - in the way that the answers of the respondents reflected their real attitude towards the change process. However, an internet survey has also its disadvantages. For instance, inexperienced internet users and the lack of an interviewer for clarification.
Besides the survey, several conversations were held with the program lead of the change project, and two semi-structured interviews were held with the change agent and a former recipient of the change project (See appendix G for transcriptions of the interviews with the change agent and former recipient). These interviews were used to understand the context of the situation in which the change took place.
The internet survey was carried out in the organization segment KPN Residential. This segment was involved in a large-scale change project WoW (Way-of-Working). This change project WoW is described in detail in the introduction. The change recipients were 53 Dutch chain- and line managers (middle managers) of the KPN Residential department, who participated in WoW wave 3. At the time of data collection these 53 change recipients were already introduced and involved in the change project for approximately four weeks.
27
Eventually, the web-based survey including a letter of introduction was sent via the mailbox of the change program management of KPN WoW to all participating recipients. This survey and letter of introduction are provided in appendix A. The survey was sent among all 53 change recipients of 3rd WoW, including 37 (70%) men and 16 (30%) women. Besides the 53 change recipients, the survey was also carried out among 42 additional managers. Namely, the change recipients of WoW wave 1 and 2. A readiness assessment was done on all survey participants (wave 1 – 3) and comparisons were made between WoW waves by using descriptive statistics. This assessment showed an increase in recipients’ readiness for change from wave 1 to 3. The change process antecedents- and personal valence perceptions were also more positive among wave 3 managers than among wave 1 and 2 managers. Since this comparison of the waves was not the focus of this study, the whole assessment can be found in appendix H.
For this study; eventually, 39 participants (excluding 4 incorrect responses) filled in the survey correctly. These 39 participants were all participants of WoW wave 3, since this indeed this study’s focus. This constituted a response rate of 74%. This response rate was reviewed as high, since - according to Anseel, Lievens, Schollaert & Choragwicka (2010) - the average response rate was 47.1% of 357 studies which all mailed surveys to managerial respondents.
The high response rate of this study was achieved by sending reminder e-mails to the change recipients via the program management mailbox (after one week and after 3 weeks), writing a message in the newsletter of the change program, and finally by giving a presentation for the change agents to remind them about the urgency of the response height (due to the small population of change recipients) and in order to encourage them to remind their coachees (i.e. change recipients). These documents are shown in appendix B.
The 39 participants included 25 men (64%) and 14 women (36%). This gender percentage ratio of the participants came close to the gender percentage ratio of all the 53 change recipients (target population).
The age of all the participants ranged from 21 to 60 years. More precisely, 5 participants were 21 to 30 years (13%), 17 participants were 31 to 40 years (44%), 13 participants were 41 to 50 years (33%), and 4 participants were 51 to 60 years (10%). A rough estimate of the mean age was 39,6 years. Table 3 displays this information on the age of the participants.
TABLE 3 Age of the participants
Age Response % percentage
21 to 30 years 5 13%
31 to 40 years 17 44%
41 to 50 years 13 33%
51 to 60 years 4 10%
28
The participants’ tenure ranged from 0 to 16 years or more. 3 participants shorter than 1 year (8%), 7 participants 1 to 5 years (18%), 12 participants had a tenure of 6 to 10 years (31%), 9 participants had a tenure of 11 to 15 years (23%), and 8 participants 16 years or more (21%). A rough estimate of the mean tenure was 9,7 years. Table 4 displays this information on the tenure of the participants.
TABLE 4
Tenure of the participants
Tenure Response % percentage
Shorter than 1 year 3 8%
1 to 5 years 7 18%
6 to 10 years 12 31%
11 to 15 years 9 23%
16 years or more 8 21%
Total 39 100%
After assessing the representativeness of the sample, the 39 survey participants were seen as representative for all 53 change recipients of wave 3 of the change project WoW. This was assumed for the following reasons: (1) The response rate was very high, (2) the gender percentage ratios of target population and sample size did not differ much; only +/- six percent difference, (3) the sample means of age and tenure were comparable with the real situation; however, these response classes were not that detailed. Therefore, age and tenure was analyzed with caution. The department ratio was not further analyzed since it could be harming the anonymity of departments; and consequently, the managers themselves.
3.2 Measures
Multi-item measures were used to ensure adequate measurement of each variable. All scales where borrowed from pre-existing measures, some were adjusted for this study. Even though some of the scales were adjusted, efforts were made to select scales that had been widely used in related areas of research and have considerable empirical data to indicate that they were reliable and valid.
Table 5 shows an overview of the existing and validated measures used per variable. This will be followed by a short description of each measurement.
TABLE 5
Measurements per variable
Variables Measure
29
Dependent variables: Readiness for Change dimensions.
The readiness for change variables were gauged by scales of Bouckenooghe et al. (2009) and included three dimensions: The emotional dimension (EmoRead), the intentional dimension (IntRead), and the cognitive dimension (CogRead), each consist of three items and demonstrated good internal consistency in Bouckenooghe et al. (2009)’s study. The 6 items of the emotional and intentional dimension were focused on a specific change and were positively formulated statements. The 3 items of the cognitive dimension were focused on changes in general and were negatively formulated statements.
Independent variables: Change process factors.
For the measurement of involvement in the change process (InvCP) and top management support (TopSup) the existing 9 items of Bouckenooghe et al. (2009) were used. A 6-item scale for involvement in the change process (including 2 items on participation and 4 items on quality of communication), and a 3-item scale for top management support.
For the measurement of change agent support (CASup) the 6-item scale of Bouckenooghe et al. (2009) - meant for measuring support by supervisors - was used. However, some adjustments were made since the measurement purpose of this study was the change agent instead of the supervisor; which was the purpose of Bouckenooghe et al. (2009). Therefore, 4 items were deleted because these items were not applicable and the two preserved items were adjusted (i.e. “The change agent coaches me very well about implementing change” and “the change agent pays sufficient attention to the personal consequences that the changes could have for me and my staff members”).
Mediator: Personal valence.
30
Finalization of survey.
The final survey included 26 statements and 5 control questions on gender, age, tenure, department and project wave. These 26 statements included 3 statements on participatory management (PartMan). However, this variable was not used for this study since PartMan is about the change context and not about the change process. As already mentioned, this final survey can be found in appendix A including the letter of introduction.
Likert-scale.
The participants were asked to indicate to what extent the statements were applicable to their situation on a seven-point Likert type scale (i.e. 1 = totally disagree and 7 = totally agree). Although both five-point and seven-five-point likert scales are supported because of their relatively equivalent outcomes (Colman, Norris & Preston; 1997), the seven-point Likert type scale was used; since several studies showed that the seven-point scale produces greater variability than the five-point scale (e.g. Rhodes, Matheson & Mark, 2010), and that the seven-point scale provides more sensitive indications of respondents’ feelings and has the highest overall reliability (Alwin, 1997).
Translation of survey.
The 26 statements – as originally developed by Bouckenooghe et al. (2009) and Holt et al. (2007) –
were translated into Dutch, since the organization that was subject to this study is Dutch. A common procedure for guarding against language bias in measurement scales is back translation (Bouckenooghe et al., 2009). The translation from English to Dutch was done in person and then a second person (Dutch-English interpreter) translated the Dutch translation back into English. Thereafter, the original and back translated statements were compared and no adjustments were made because the meaning was the same. In addition, the original Dutch version of OCQ of Bouckenooghe et al. (2009) was eventually and lately obtained via e-mail (Bouckenooghe et al. conducted their research both in English and Dutch; however, only the English version was published). After comparison, it appeared that the translated statements of OCQ and original Dutch statements of OCQ were equivalent. This could be seen as extra confirmation for sufficient translation.
Pretest survey.
31 3.3 Data analysis
The statistical program “SPSS version 20” was used for analyzing the data in order to test the hypotheses. The section below gives an overview of the data analysis.
Exploring data.
At first, the received data was checked for missing values and outliers. 3 cases were removed because these participants only completed five questions and 1 case was removed for the reason this participant answered all statements with the same option. A few extreme values were spotted by using boxplots (see appendix C); however, since these particular cases seemed to be entered correctly, these cases were not removed. Consequently, 39 cases were further analyzed in this study (43 minus 4 is 39 cases).
Secondly, the 6 negatively formulated items of the questionnaire were reverse-coded. This means that these item-scores were changed in order to point in the same direction as all other items (positively formulated). These 6 recoded items are indicated with an (R) and can be found in appendix A.
Factor analysis and reliability analysis.
An exploratory factor analysis (EFA) was carried out to test if the items were measuring the same factor. The conceptual model of this paper incorporated three separate categories aimed at measuring: (1) the process of change, (2) the personal valence, and (3) the readiness for change. The process items, personal valence items and readiness for change items were factor analyzed separately. In total 11 (change process factors), 3 (personal valence), and 9 (outcome; readiness for change) items were factor analyzed. As extra support; Bouckenooghe et al. (2009) also executed EFA’s separately for the different categories in their tool. Moreover, the EFA’s were purely for verifying the existing questionnaire scales (including some adjustments) as valid and reliable.
The procedure of the exploratory factor analysis is accurately described by Field (2009) and was used as guidance. The procedure contains four steps which will be shortly described:
Preliminary analysis.
