p 0
Sense of urgency and Willingness to change
SENSE OF URGENCY AND
WILLINGNESS TO CHANGE
Master thesis, MSC BA, specialization Change Management University of Groningen, Faculty of Economics and Business
January 2011 ZWIER SMITH Student number: 1466038 H.J. Schimmelplein 16 3532 TE Utrecht Tel.: +31 (0)6-25212427 E-mail: zwier_smith@hotmail.com
First university supervisor: Dr. C. Reezigt
Second university supervisor: Drs. F.M. de Poel
p 1
SENSE OF URGENCY AND
WILLINGNESS TO CHANGE
ABSTRACT
Purpose – This research seeks to explore the effects of a sense of urgency, composed out of a need for
change (i.e. discrepancy) in conjunction with time pressure on an individuals’ willingness to change.
Design/methodology/approach – An experiment is designed to statistically examine the effects of a
sense of urgency on willingness to change, which is enriched by interviews of several participants to enhance validity. This is followed by 20 in-depth interviews in 7 different change projects to empirically substantiate the hypothesized relation.
Findings – Based on the experiment, the results show that a sense of urgency is positively related with
willingness to change, whereby discrepancy is the strongest predictor and time pressure is of minor importance. Outcomes of both interview efforts indicate that a sense of urgency is however not a condition sine qua non for willingness to change. Willingness to change is found to be mostly determined by personal valence and loyalty. Real time pressure for change barely exists.
Research limitations/implications – Accurately measuring a sense of urgency is difficult, in this
research, perceptions of respondents are measured based on an online experiment. The in-depth interviews were conducted in public sector change initiatives; therefore the outcomes of this study may not be applicable to private sector change initiatives.
Originality/value – This research adds knowledge to the research of organizational change and the role
of sense of urgency in willingness to change. A multilateral methodology by means of an experiment, validating interviews and in-depth interviews presents a unique triangulation effort for empirical research.
Key words – Willingness to change, sense of urgency, discrepancy, time pressure, public sector
p 2
TABLE OF CONTENTS
1. INTRODUCTION 3
2. THEORETICAL FRAMEWORK 4
2.1 Resistance, readiness and willingness to change 4
2.2 Sense of urgency as antecedent condition 6
3. METHODOLOGY 9
3.1 Research design and data collection (A) 9
3.2 Data collection by triangulation interviews (B) 10
3.3 Interview method and data collection (C) 10
3.4 Measures in the experiment 11
3.5 Data analysis 13
4. RESULTS 15
4.1 Results of the experiment (A) 15
4.1.1 Correlation analysis results 16
4.1.2 Regression analysis 17
4.1.3 Comparison of correlation and regression coefficients 19
4.1.4 Conclusion statistical analyses 20
4.2 Results interview outcomes for triangulation purposes (B) 21
4.2.1 Time Pressure 21
4.2.2 Discrepancy and Willingness to change 22
4.2.3 Conclusion triangulation interviews 24
4.3 Results in-depth interviews in public sector change initiatives (C) 25
4.3.1 Overall quantitative interview outcomes 25
4.3.2 Comparing change manager and recipient outcomes 25
4.3.3 Conclusion quantitative outcomes 26
4.3.4 Willingness to change 27
4.3.5 Discrepancy, time pressure and a sense of urgency 28
4.3.6 Conclusion in-depth interviews 31
5. CONCLUSION AND DISCUSSION 32
5.1 Triangulation results for hypotheses 32
5.2 Theoretical implications 35
5.3 Limitations and recommendations for future research 35
REFERENCES 37
APPENDIX A: EXPERIMENT DESCRIPTION 44
APPENDIX B: TRIANGULATION INTERVIEW QUESTIONS 48
APPENDIX C1: EXPLORATORY FACTOR ANALYSIS 49
APPENDIX C2: FORCED FACTOR ANALYSIS 51
APPENDIX D: COMPARING CORRELATION COEFFICIENTS 52
APPENDIX E: COMPARING REGRESSION COEFFICIENTS 53
APPENDIX F: INTERVIEW QUESTIONS 54
p 3
1. INTRODUCTION
The body of literature dedicated to the understanding of organizational change is impressive (Elias, 2009). Organizational change has emerged as one of the most prevalent topics of management theory and practice (Klarner, Probst & Soparnot, 2008; Thurlow & Mills, 2009; Van de Ven & Poole, 1995, 2005) For over one-half century, researchers have attempted to provide insight into change dynamics and help organizations successfully implement change (Walker, Armenakis & Bernerth, 2007). Gradually, one axiom regarding change management arises on which researchers generally tend to agree. It seems commonly accepted that about 70 percent of change initiatives tend to fail (e.g. Allen, Jimmieson, Bordia & Imer, 2007; Beer & Nohria, 2000; Burnes, 2009; By, 2007; Devos, Buelens & Bouckenooghe, 2007; Kotter, 1995; Szabla, 2007; Ten Have, Ten Have, Janssen, 2009). Many academic researchers are inspired by this axiom to contribute in finding solutions, by identifying key variables or conditions for successfully bringing about change. According to one of those researchers for change to happen those involved must (1) feel a need or desire to change, (2) perceive that the change is for the better, and (3) have the capabilities to make the change come about (Caluwé, Que & Vermaak, 2004). This partially refers to Lewin’s change theory, when he describes that the present situation – the status quo – as being maintained by certain conditions or forces (Lewin, 1943). Enlarging on Lewin’s ideas, Burnes (2009) argues that for the introduction of change there has to be a “felt-need”, an ‘individual’s inner realization
that change is necessary’. It is crucial to create such sense of urgency in order get the required cooperation of employees and managers (Kanter, Stein & Jick, 1992; Kotter, 2008; 1995; Pfeifer, Schmitt & Voigt, 2005). Kotter (2008) unfortunately concludes that the single biggest error people make when they try to change is not creating a high enough sense of urgency among enough people to set the stage for making a challenging leap into some new direction.
However, while it might be vital for an organization to be “change ready” (to set the stage) before attempting to implement and manage any kind of change (Armenakis & Harris, 2009; Armenakis, Harris & Mossholder, 1993; By, 2007; Jones, Jimmieson & Griffiths, 2005; Kotter, 1996), one could doubt the importance of a sense of urgency in this. Results from a study by Bennebroek Gravenhorst, Werkman and Boonstra (2003) suggest that a sense of urgency is not a “conditio sine qua non” for change to happen. People can be willing to change because they see personal benefit in change, without having a sense of urgency (Werkman, Boonstra & Bennebroek Gravenhorst, 2001). Secondly, there is scant empirical research concerning sense of urgency as an antecedent variable affecting willingness to change, albeit extensive practitioner-oriented literature regarding the necessity of urgency awareness. Therefore, this research investigates the relation between a sense of urgency and willingness to change.
p 4
2. THEORETICAL FRAMEWORK
In this theoretical framework the most relevant and recent academic literature on willingness to change and sense of urgency is presented. First of all, a concise overview on resistance and readiness for change is presented. Furthermore, the concept of willingness to change in relation with resistance and readiness to change is elaborated. Moreover, the role of a sense of urgency as antecedent condition for willingness to change is discussed. Finally, it is argued that a sense of urgency consists of a perceived discrepancy and time pressure, which both are hypothesized to positively influence willingness to change as visually presented in a conceptual model.
2.1 Resistance, readiness and willingness to change
Creating a basis that supports change is a key issue in managing and planning change (Bouckenooghe, Devos & Van Den Broeck, 2009). The successfulness of change efforts is not only due to their substantive nature but also to the processes followed or actions undertaken during their implementation (Armenakis et al., 1993; Neves, 2009). Therefore, managers should focus on preparing for change, by overcoming resistance, involving employees and transforming them into change agents, while using a change message that facilitates the adoption of behaviors required for the success of a change initiative (Neves, 2009). In this section, I chronologically describe the most relevant theoretical concepts regarding preparations for change, i.e. resistance, readiness and willingness to change.
Resistance. Kurt Lewin (1947) introduced the term resistance to change in his Field Theory and
later work on Group Dynamics (Bennebroek Gravenhorst, 2003; Metselaar & Cozijnsen 2010). According to Lewin (1943) the status quo is maintained when forces driving change are in balance with the forces restraining change. These restraining forces laid the foundation for the conceptualization of resistance to change. Lewin (1951: 240) believed that a field, “a totality of coexisting facts which are conceived of as mutually interdependent”, is in a continuous state of adaptation (Burnes, 2009; Madsen, John & Miller, 2006). Lewin therefore used the term ‘quasi-equilibrium’ to illustrate that there might be some constancy in the behaviour and processes of a group, but these tend to continuously fluctuate owing to the changes in the forces or circumstances that influences the group (Burnes, 2009; Metselaar & Cozijnsen, 2010). Subsequently, Coch and French (1948) researched the topic of reducing resistance to change by demonstrating the value of organization member participation in change efforts, therefore they were the first to make the connection between resistance and change in organizations.
Kotter and Schlesinger (1979) do not define resistance to change, but identify four main reasons for resistance: ‘parochial self-interest’, misunderstanding and lack of trust, different assessments, and low tolerance for change. They describe six approaches for dealing with these reasons accordingly. For them, resistance is an inevitable reaction to change initiatives owing to the incapability of employees to understand the necessity of change for the good of the organization (Bennebroek Gravenhorst, 2003).
p 5
resistance to change manifests in grievances, turnover, low efficiency, restriction of output, and aggressions against management (Coch & French, 1948).
Readiness. Readiness to change originally stems from the opposite concept, the negative reactions of employees in terms of resistance to change (Armenakis & Harris, 2009; Ten Have, Ten Have, Janssen, 2009). The concept of readiness to change was first mentioned in 1957 by Jacobson on a symposium for social psychologists (Bouckenooghe, 2009; Holt, Armenakis, Feild & Harris, 2007). According to Bouckenooghe (2009) this polarization (between resistance and readiness) typifies the positive-negative attitude distinction and even reflects the positive-negative mindset in academic research (Cameron, 2008). For instance, Ford, Ford and D’Amelio (2008) argue that the reactions of those receiving change are not necessarily ‘dysfunctional obstacles’ and possibly serve as an asset or resource with strengthening value.
Although readiness for change was introduced in 1957, a theoretically sound conceptualization of the idea was only provided in 1993 by Armenakis et al. They stress the similarity with Lewin’s (1951) concept of unfreezing, by arguing that readiness to change is reflected in ‘organizational members’ 1)
beliefs, 2) attitudes, and 3) intentions regarding the extent to which changes are needed and the organization’s capacity to successfully make those changes’. In short, change readiness can be defined as a cognitive precursor to the behaviors of either resistance to, or support for, a change effort (Armenakis et al., 1993).
By reflecting on their work on readiness to change, as defined earlier, Armenakis and Harris (2009) distinguish five key components on the concept: discrepancy, appropriateness, efficacy, principal support and valence. According to them, discrepancy refers to the belief that a change is needed, while appropriateness reflects the belief that the proposed change to address the discrepancy is the correct one given a particular situation. Efficacy has to do with the belief that the change recipient and the organization are capable of successfully implementing the change. The fourth component of principal support deals with the commitment of formal leaders (‘vertical change agents’) and opinion leaders (‘horizontal change agents’). Finally, valence reflects the belief that change is beneficial to the change recipient. Armenakis and Harris (2009) state that these five components together determine the level of readiness to change.
Willingness. Willingness to change is closely related to the concept of readiness, but explicitly
describes one’s actual intention to cooperate on a particular change initiative based on one’s beliefs (Metselaar & Cozijnsen, 2010), while leaving the issue of capability to change out. To put it differently, one can be willing to change without being ready in terms of the five beliefs of Armenakis and Harris (2009). However, one cannot be ready for change without the actual intention to make the change happen. For an organization, the ability to change depends heavily on the motivation to change of its individual employees (Armenakis et al., 1993; Bernerth et al., 2007; Lundberg, 2004).
Metselaar (1997: 34) defines willingness to change, based on Ajzen’s (1991) model of planned behaviour, as a “positive behavioural intention towards the implementation of modifications in an
p 6
behavioural intention. Intentions are assumed to capture the motivational factors that influence behaviour (Ajzen, 1991). Furthermore, he asserts that intentions are indications of how hard people are willing to try, assuming that the stronger the intention to engage in a behaviour, the more likely its performance should be.
For measuring the willingness of change several instruments exists (Holt et al., 2007). They identify more than 40 instruments that qualitatively or quantitatively measure readiness and/or willingness to change. Measuring willingness to change should be done on an individual level, because individual attributes significantly influence the willingness to change (Holt et al, 2007; Metselaar & Cozijnsen, 2010). This implies that the level of willingness to change varies per individual. Besides,
“organizations only act and change through their members and even collective activities that occur in organizations are the result of some amalgamation of the activities of individuals” (Lundberg, 2004: 247).
In this research it is assumed that readiness for change is a more comprehensive concept than a positive behavioural intention alone. By concentrating on willingness to change the scope of this research is narrowed down by avoiding investigating a ‘trash can’ (De Caluwé & Vermaak, 2006) or an unruly concept (Dent & Powley, 2001) of either resistance or readiness to change. The concept of a sense of urgency as an antecedent variable for willingness to change will be described in the following paragraph.
2.2 Sense of urgency as antecedent condition
A shared sense of urgency is one of the driving forces of change initiatives (Delden, 2010; Kanter et al., 1992; Kotter, 2008). Therefore, the role of a sense of urgency as an antecedent condition for willingness to change is elaborated. For the sake of clarity, antecedents are those factors which influence the positive behavioural intention of employees to participate in change.
A sense of urgency can be defined as ‘a distinctive attitude and gut-level feeling that lead people to
grab opportunities and avoid hazards, to make something important happen today, and constantly shed low-priority activities to move faster and smarter, now’ (Kotter, 2008). In addition, the urgency of a change effort can be defined as the amount of time available before changes must be implemented (Armenakis et al., 1993). A common goal of establishing a sense of urgency is to make the status quo seem more dangerous than launching into the unknown (Beer & Walton, 1990). Whether one calls this effort to motivate change ‘dissatisfaction’ (Beer & Walton, 1990) or ‘felt-need’ (e.g. Burnes, 2009; Lewin, 1951) it is a necessity to create a broad sense of urgency (Caluwé, Que & Vermaak 2004; Kanter et al., 1992). De Bruin (2009) asserts that change only evolves when there is a sense of urgency, because people do not change unless they really have to. In addition, the greatest obstacle to creating a sense for the urgency of the situation is the overbearingness and complacency of organizations (Kotter, 2008; Pfeifer, Schmitt & Voigt, 2005). Moreover, in academic literature it is suggested that a sense of urgency is needed to get the required cooperation (i.e. willingness to change) of employees and managers (De Bruin, 2009; Kanter, Stein & Jick, 1992; Kotter, 2008; 1995; Pfeifer, Schmitt & Voigt, 2005). Therefore, the main hypothesis in this research is that a sense of urgency positively influences one’s willingness to change.
p 7
Furthermore, to unravel the concept of a sense of urgency it encompasses 1) an identified significant discrepancy between the current state and the desired state of organization (Que & Vermaak, 2004), caused by either internal pressures (e.g. dissatisfaction) or by external pressures (e.g. major opportunities/threats) (Kotter, 2008) and 2) scarcity of time available for a change effort (Armenakis et al., 1993).
Discrepancy or organizational disconfirmation, whatever its source, functions as a primary driving force in changing individuals and therefore organizations (Schein, 1996). According to Warren (2004) widespread dissatisfaction is needed to bring about change. Likewise, revealing discrepancy throughout the organization should enforce a sense of urgency (Kotter, 2008). Moreover, it is not hard to convince an organization at the edge of bankruptcy that change is necessary (Kanter et al., 1992). Kanter et al. (1992) contend that when a need for change is not that obvious it is important to sensitize an organization to pressures for change. Furthermore, employees will generally experience deep levels of hurt before they will seriously undertake meaningful change (Cummings & Worley, 2005).
Discrepancy refers to the belief that a change is needed; by identification of a significant gap between the current and desired state of an organization. This discrepancy should be consistent with relevant contextual factors and can be used to justify the need for organizational change (Armenakis et al., 1993). Furthermore, it could be useful to diffuse dissatisfaction, as mentioned before, throughout the organization, to make discrepancies self-evident. Holt et al.’s (2007) analysis of academic literature indicated that discrepancy was one of the most influential factors in determining willingness to change. It is therefore hypothesized that a high level of perceived discrepancy will lead to a high willingness to change.
Hypothesis 1a: Discrepancy is positively related to willingness to change
Time pressure. As argued by Armenakis and Harris (2009), a willingness creation program may
not have the luxury of time due to the urgency of the need for change. Of course, time is an irreversible variable, but effectively timing change initiatives enhance the change success (Eisenhardt & Brown, 1998; Huy, 2001). There is scant empirical research concerning time pressure as an antecedent variable affecting willingness to change. Nodqvist, Hovmark and Zika-Viktorsson (2004) researched the effect of time pressure on the social processes in project teams. Their results show that time pressure was negatively related to both estimated goal fulfillment and job satisfaction. Absence of time pressure, however, can lead to ineffectiveness or indifference of project groups (Gevers, Eerde van & Rutte, 2001). Deadlines regulate and help to structure work through the breakdown of projects in interim goals given a certain time frame (Nordqvist et al., 2004) which motivates a project team to start working on a task. A lack of a sense of urgency to change now, has the potential to be the single reason why individuals do not change (Burrowes & Needs, 2009). They argue that individuals may recognize a need for change, as well as its importance and benefits, but may not feel that they need to instigate change at a particular point in time. Hence, I hypothesize that time pressure, as a second element of a sense of urgency, will positively affect one’s willingness to change.
p 8
Conceptual model. In sum, this research assesses the effects of a sense of urgency on willingness to change. Based on a literature study the concept of a sense of urgency is decomposed in two underlying variables: discrepancy and time pressure. Both these variables and the construct of a sense of urgency are expected to positively relate to willingness to change. The hypothesized relations are virtually depicted in a conceptual model (see figure 1).
FIGURE 1
Conceptual model of hypothesized relation
p 9
3. METHODOLOGY
This section presents a methodology, based on an A) online experiment, B) subsequent validating triangulation interviews and C) in-depth interviews in public sector change initiatives, for assessing the theoretically constructed hypotheses. Firstly, all three ways of data collection are described. Secondly, the measures for interpreting data, obtained from the experiment, are described by means of a reliability test and a factor analysis. Finally, the statistical data analyses techniques are presented.
3.1 Research design and data collection (A)
An experiment was designed and used to test the relationship between perception on discrepancy and time pressure on the one hand and willingness to change on the other. Experiments involve interventions by the researcher beyond that required for measurement (Cooper & Schindler, 2003). In essence, throughout the presented cases in the experiment the relevant variables (perceived discrepancy and time pressure) are manipulated to observe how it affects the variable of interest (willingness to change). The experiment consists of four cases which were presented to the respondents in an online questionnaire. The cases were nonfiction albeit little adjustments and were derived from actual change projects in the public sector accompanied by the consulting firm called Andersson Elffers Felix (AEF) in Utrecht (the Netherlands). Unit of analysis for this experiment is the individual consultant at AEF. By answering the questions the respondents reveal their perception on discrepancy, time pressure and willingness to change. The consultants had two weeks to participate in the online experiment. Participating in the experiment was voluntarily and a guarantee of confidentiality was given. In total 41 persons participated in the online experiment.
The experiment can be best described as a one-group pre-test-post-test design (Cooper & Schindler, 2003). The effect of the experiment is measured by comparing the levels of perceived discrepancy and time pressure with perceived level of willingness to change across the four experimental cases. The first presented case can be used as control measurement (pre-test), since both independent and explanatory variables are kept moderate.
p 10 3.2 Data collection by triangulation interviews (B)
A triangulation effort by means of interviews with respondents of the online experiment has been conducted to enhance validity of the statistical findings. Validity has to do with the quality of collected data, particularly to the extent of which the data is a valid representation of the observed phenomenon in reality (Baarda, de Goede & Teunissen, 2005). They argue that when the results of the triangulation effort converge with the obtained quantitative data by the experiment it is a strong indicator of validity.
The interviews were based on the conceptual model derived from the previously presented literature study focusing on the factors discrepancy, time pressure, and urgency to change. In particular, assumptions underlying this research are tested and questions were asked to explore the (unexpected) outcomes of the online experiment
For the interviews nine representatives of the sample (consultants of AEF) are selected based on their willingness to participate, availability and experience with change processes in the public sector. The interviews are lightly structured starting with a short elaboration of the quantitative outcomes. The interviews lasted about one and a half hour. The interview questions are depicted in Appendix B.
Data analysis. To analyze the outcomes of the interviews I identified relevant issues regarding the hypothesized factors that foster or inhibit willingness to change. By comparison the most important or often mentioned issues will be described in the interview results.
3.3 Interview method and data collection (C)
Multiple in-depth interviews have been conducted to get more grip on the concept of sense of urgency and willingness to change in actual change processes in the public sector. The aim for this third research effort is to empirically enrich the outcomes of the experiment and its subsequent triangulation interview results. With in-depth interviews the researcher explores some topics in precision, so that personal and often emotional issues are addressed (Baarda, de Goede & Teunissen, 2005).
In total 20 interviewees in seven change projects in the public sector are selected for in-depth interviews. These projects were selected based on: 1) significance of change projects in public sector (based on amount of people involved and time span of change effort), 2) maturity of change projects (to examine the effects of sense of urgency over the course of a change process) and 3) diversity of change projects (to achieve a certain level of generalizability). Andersson Elffers Felix, a Dutch public sector consultancy firm, made it possible for the researcher to contact these projects with a participation request. In order to get a balanced view of the change projects it was tried to interview at least two different persons per project: a strategist (principal) and/or an implementer (change manager) and a change recipient; this is based on a distinction by Kanter et al. (1992). As change cascades down through the organization, it is believed that, at each level, change is perceived differently (Bouckenooghe, 2009). Besides, interviewees were selected on: direct involvement, time spent on change process and intensity in which they experienced the change. Hereby it is possible to get a multilateral in-depth view of the change projects at hand concerning the level of sensed urgency and willingness to change. The interview questions reveal a retrospective perspective at change projects.
p 11
strategy and scope of these processes. Based on this information, interview questions were adapted to the individual respondent and specific context of the change project to enhance the fruitfulness of the conversation.
Three people were interviewed for pilot testing. Based on the pilot test, some interview questions were deleted or adjusted to reduce negative side-effects. In addition, a check for social desirability was done as well as determining the time needed to complete an interview. After the pilot, all other interviewees were approached and interview questions were sent upfront. After the interviews, interview reports were produced and send to the interviewees for verification in order to raise the level of reliability of the research.
The interviews were semi-structured starting with a short elaboration of the research goals. Thereafter, the interviews followed a chronological timeframe discussing important events from the start to the end of the change initiative. For some questions the interviewees had to indicate their response on a 5-point Likert scale (ranging from 1= very low to 5= very high) to increase comparability between interviewees. On average the interviews lasted about one and a half hour. The questions used are depicted in Appendix F.
In total, seven cases were used ranging from reorganizations in municipal departments (3), mergers of housing corporations (3) and a collaborative network concerning public safety (1). A concise description of each change project concerning goals, scope, timeframe and key events can be found in Appendix G.
Data analysis. First of all, the outcomes of the interviews are analyzed based on the quantified responses in two ways: per change project and per role (change manager or recipient). Subsequently, the interesting phenomena identified by this first analysis, concerning the hypothesized factors that foster or inhibit willingness to change, are explored with qualitative responses. This is done by grouping similar opinions regarding discrepancy, time pressure and willingness to change. By comparison the most important or often mentioned issues are described in the interview results.
3.4 Measures in the experiment
p 12
To test for internal consistency of measures a Cronbach’s alpha-test is conducted. The experiment consists of three variables and nine items in total. Since four different cases are used, the Cronbach’s Alpha is also computed for each case and variable differently. The .70 value was used as a cut-off score (Malhotra, 2007). Therefore, an Alpha equal or higher than .70 is considered sufficient internal consistent to be reliable. Lower alpha scores indicate low inter-correlations between items meaning that the items do not measure the same construct.
Discrepancy. A sense of urgency is operationalized by measuring perceived level of discrepancy and perceived level of time pressure. Discrepancy refers to the belief that a change is needed; by identification of a significant gap between the current and desired state of an organization (Armenakis et al., 1993). This factor was measured by using three items, with a Cronbach’s Alpha overall score of .91 and per case respectively .86, .77, .91, and .84. The scores exceed the minimum criteria score of .70 and are therefore accepted. Minimal improvements could be made per case when deleting a single item, but no item consistently decreased Alpha’s scores across the cases.
Time pressure is defined as the insistence of the need for change (Armenakis & Harris, 2009). This factor was also measured using three items, with a highly reliable Cronbach’s Alpha overall score of .93 and sufficient Alpha scores of respectively .82, .79, .95, and .88 per case. In the first two cases, one item, concerning insisting pressures for change improved the Alpha score to respectively .91 and .87 if deleted. For the third case, the item concerning rapidity of implementation improved the alpha to .96 if deleted. For the fourth case, no item could improve the score on Cronbach’s Alpha when deleted.
Sense of urgency. To test reliability for the construct of sense of urgency, items of both discrepancy and time pressure are combined. Overall, this scores a high Alpha of .93 and respectively, .86, .78, .95, and .80 per case. Overall and for each case individually, these scores indicate interrelatedness between the items that compose the construct of a sense of urgency. There was no opportunity to improve reliability by deleting items.
Willingness to change is defined as ‘a positive behavioural intention towards the implementation
of modifications in an organization’s structure, or work and administrative processes, resulting in efforts from the organization member’s side to support or enhance the change process’ (Metselaar, 1997). Three items were used to measure this factor, with an overall Alpha score of .97 and respectively, .96, .91, .97, and .97 per case. This indicates highly reliable inter-correlations throughout the measures across cases. Except for case 1, these scores could even be improved by deleting the item concerning convincing colleagues about the need for change. Table 1 summarizes all reliability scores for measures used in this experiment.
TABLE 1 Cronbach’s Alpha scores
Variables Items αααα Overall αααα Case 1 αααα Case 2 αααα Case 3 αααα Case 4
1. Sense of urgency 6 .93 .86 .78 .95 .80
1a. Discrepancy 3 .91 .86 .77 .91 .84
1b. Time Pressure 3 .93 .82 .79 .95 .88
p 13
Validity. To test the validity of the constructs used in this experiment across the four cases, two factor analyses has been conducted (see Appendix C). A factor analysis is used to explore which item loads on which factor or variable in order to adequately minimize the number of items by combining them into the only relevant distinct factors or variables. Therefore, a factor analysis can test whether the items indeed measure the theoretical constructs as intended. An exploratory factor analysis per case could not confirm the theoretical constructs for all cases. For case 1 to 3 the analyses revealed two factors (based on Eigenvalues of 1 or higher), instead of three as was hypothesized. Case 4 revealed three factors whereby all items correctly loaded on the appropriate factor. The result of the exploratory factor analyses implies questionable validity of the experiment for some cases. Two causes may affect this suboptimal validity: the rather small sample size in this experiment and the relatively low number of items used in this study which foster inter-correlations (Malhotra, 2007). Therefore, the results of the factor analysis might be unreliable and still legitimize the use of the predefined theoretical constructs to base the findings of this research on, notwithstanding implications for validity. Subsequently, when a forced factor analysis was conducted, forcing the analysis to identify three factors per case (see Appendix C2), all items loaded on the appropriate factor except for case 1. For case 1 another forced factor analysis was performed whereby item 4 was removed due to its deviant nature. Now all items load appropriately and the theoretical constructs are confirmed.
3.5 Data analysis
Data derived from the online experiment are statistically analyzed using IBM SPSS 19. Parametric tests as correlation and regression analyses can be used to test the hypotheses. There are four basic assumptions that must be met for the parametric tests to be accurate: 1) interval data, 2) normally distributed data, 3) homogeneity of variance, and 4) independence (Field, 2000). Firstly, Likert-scales actually obtain ordinal data, but it is commonly accepted that these data may interpreted as interval data (Cooper & Schindler, 2003). Secondly, to test whether data are normally distributed and variance is homogeneous, a Kolmogorov-Smirnov test is used. This test compares the set of scores in the sample to a normally distributed set of scores with the same mean and standard deviation (Huizingh, 2006). If the test is non-significant (p>0,05) the distribution in question is not significantly different from a normal distribution (i.e. it is probably normally distributed). The results of the Kolmogorov-Smirnov test are displayed in table 2.
TABLE 2
Results of a Kolmogorov-Smirnov test
Variables
Case 1 Case 2 Case 3 Case 4
Statistic Sig. Statistic Sig. Statistic Sig. Statistic Sig.
Discrepancy .12 .57* .21 .054* .24 .02 .11 .70*
Time pressure .09 .86* .16 .27* .21 .057* .20 .09*
Willingness to change .20 .07* .30 .00 .26 .01 .13 .52*
p 14
Table 2 demonstrates that the variable willingness to change in case 2 and 3 is not normally distributed, as well as discrepancy in case 3. Based on the Kolmogorov-Smirnov test, all other sets of scores across the cases are normally distributed, which legitimizes the use of parametric tests. For the non-significantly normally distributed variables nonparametric tests are used as a means of testing the hypotheses of interest. A Spearman’s correlation coefficient will be used as a non-parametric statistic for the data that violated distributional assumptions. Finally, regarding the prescription of independence, it can be assumed that the respondents are selected independently.
Comparing coefficients. To statistically compare the correlation coefficients within and between the four experimental cases, respectively a Steiger Z and Pearson-Filon test (based on the difference of Fischer r-to-Z transformed coefficients) will be used. The Steiger Z test is applicable for making statistical comparisons between correlation coefficients (discrepancy versus time pressure) measured on the same individuals (Steiger, 1980). The Pearson-Filon test is appropriate for testing between the experimental cases, since it involves the comparison of two correlations on the same sample of subjects, in which the correlations are non-overlapping in the sense of having a variable in common (Raghunathan, Rosenthal & Rubin, 1996).
p 15
4. RESULTS
In the results section the outcomes of the A) experiment, B) validating interviews for triangulation purposes and C) in-depth interviews in public sector change initiatives are presented. Firstly, descriptive statistics of the A) online experiment and the subsequent results of a correlation, regression and comparison analysis are presented in order to test the hypotheses statistically. Furthermore, the outcomes of the B) interviews for triangulation purposes are elaborated. Moreover, the outcomes of the C) in-depth interviews are presented in two ways: 1) overall quantitative responses split out by the interviewee’s role and 2) a qualitative exploration of these findings. Lastly, conclusions are drawn based on these analyses.
4.1 Results of the experiment (A)
In total 56 consultants were approached to participate in the online experiment, 41 of them actually participated, representing a response rate of 73%. A first check for missing values showed that all respondents correctly filled in all relevant questions. Of these 41 respondents, 22 were women and 19 were men, with a mean age of 35 years ranging from 24 to 61 years old. On average the consultants have 7,5 years experience in consultancy business. Figure 2 displays an error bar chart indicating all mean scores on discrepancy, time pressure and willingness to change across the experimental cases with a 95% confidence interval.
FIGURE 2
Mean scores of variables per case
p 16
Discrepancy scores are the highest in case 3, with a mean of 6,34 and lowest in case 4 with a mean of 3,57. Discrepancy scores in case 1 and 2 are comparable, while case 1 is more dispersed with a lower mean of 5,42 against a mean of 5,94 in case 2. Analyzing the scores on perception of time pressure it is obvious that again case 3 has the highest mean (6,12) with relatively low dispersion if compared with case 1 and 2. Scores of time pressure in these cases are similar with respectively a mean of 3,62 and 3,67. In case 4 respondents indicate the lowest perception of time pressure with a mean of 2,15. Analyzing the dependent variable, willingness to change, it appears that case 3 again reveals the highest score with a mean of 6,28. Again case 4 has the lowest mean score of 3,33 while widely dispersed. In case 2 respondents are predominantly willing to change with on average a score of 5,79. In case 1 respondents are less willing to change with a mean of 5,33. Hence, the scores correspond to the researcher’s intention.
4.1.1 Correlation analysis results
Table 3 presents the results of the correlation scores between sense of urgency, discrepancy, time pressure and willingness per case. For all cases a sense of urgency significantly correlates (one-tailed with a confidence level of 99%) with willingness to change. These results support the main hypothesis 1 showing diverse significant correlation scores ranging from .46 to .79. Furthermore, except for case 2, all variables positively correlate significantly (one-tailed with a confidence level of at least 95%) with each other. Therefore, hypotheses 1a and 1b are supported as well. Discrepancy shows the strongest correlations with willingness throughout the cases (respectively .79, .63, .61, .58). Time pressure also positively correlates with willingness (except for case 2), although to a lesser degree (respectively .61, .59, .45). Unexpectedly, discrepancy and time pressure also appear to be significantly correlated throughout the cases (respectively .58, .40, .79, .29). In sum, the results of the correlation analysis points in the same direction as the results of the factor analysis, which also revealed inter-correlations between variables.
TABLE 3
Correlations of construct and variables per case
Case 1 2 2a 2b Case 2 2 2a 2b
1. Willingness .79** .79** .61** 1. Willingness .46** .63** .22
2. Sense of Urgency - - - 2. Sense of Urgency - - -
2a. Discrepancy - - .58** 2a. Discrepancy - - .40**
2b. Time Pressure - - - 2b. Time Pressure - - -
Case 3 2 2a 2b Case 4 2 2a 2b
1. Willingness .64** .61** .59** 1. Willingness .64** .58** .45**
2. Sense of Urgency - - - 2. Sense of Urgency - - -
2a. Discrepancy - - .79** 2a. Discrepancy - - .29*
2b. Time Pressure - - - 2b. Time Pressure - - -
p 17
Next to a correlation analysis per case, an overall correlation analysis is conducted whereby all scores on sense of urgency, discrepancy, time pressure and willingness are combined (see table 4).
TABLE 4
Overall correlations of construct and variables
Case 1 2 2a 2b
1. Willingness .80** .81** .67**
2. Sense of Urgency - - -
2a. Discrepancy - - .69**
2b. Time Pressure - - -
**. Correlation is significant at the 0.01 level (1-tailed). *. Correlation is significant at the 0.05 level (1-tailed).
Overall, a sense of urgency significantly correlates (one-tailed with a confidence level of 99%) with willingness to change with a correlation score of .80. This result is in line with the correlation analysis per case. Thereby, support is found for hypothesis 1. Likewise, support is found for hypothesis 1a as discrepancy also correlates significantly with willingness to change (.81, one-tailed with a confidence level of 99%). In addition, in the overall analysis time pressure does also significantly correlate with willingness to change (.67 with a 99% confidence level), thereby confirming hypothesis 1b. Finally, in line with the results of the correlation analysis per case, discrepancy and time pressure are significantly correlated (.69, with a 99% confidence level).
4.1.2 Regression analysis
p 18
the discrepancy scores dominate the multiple regression analysis in such way that the predictive power of time pressure is insignificant. Therefore, a single regression analysis has been carried out (see Table 5) which then reveals a regression model (with a satisfactory R² of .35) and a significant (p < 0,01) regression coefficient of .61. Overall, the predictive power of discrepancy is superior.
TABLE 5
Multiple and Single regression analysis of Discrepancy and Time pressure on Willingness Model 1: R²: .66, with F: 37.44, Sig.: .00 Model 1: R²: .40, with F: 12.82, Sig.: .00
Case 1 ß sig. Case 2 ß sig.
Discrepancy .73 .00 Discrepancy .69 .00
Time Pressure .23 .03 Time Pressure -.03 .40
Model 1: R²: .41, with F: 12.95, Sig.: .00 Model 1: R²: .41, with F: 13.18, Sig.: .00
Case 3 ß sig. Case 4 ß sig.
Discrepancy .47 .04 Discrepancy .56 .00
Time Pressure .30 .08 Time Pressure .49 .02
Model 2: R²: .35 with F: 20.95, Sig.: .00 Case 3 (single regr.) ß sig.
Time Pressure .61 .00
R²: determination coefficient. ß: Unstandardized regression coefficient. sig: significance (one-tailed)
In order to test hypothesis 1, the construct urgency is regressed on willingness to change. Table 6 shows the results of the regression analysis. For all cases the regression model is capable of significantly predicting willingness (based on the F-ratio’s of respectively: 62.96, 10.69, 26.34, 26.99, with p < 0,01). In case 1 the R² indicates that urgency is able to predict 62% of variance in willingness. For case 2 this value is lower (22%), and the regression model in case 3 and 4 account both for about 40% of the variance in willingness. All ß-values are significant ranging from .48 in case 2 to 1.07 in case 4. Based on these results, hypothesis 1 is supported
TABLE 6
Regression analysis of Urgency construct on Willingness
Model 1: R²: .62, with F: 62.96, Sig.: .00 Model 1: R²: .22, with F: 10,69, Sig.: .00
Case 1 ß sig. Case 2 ß sig.
Urgency .94 .00 Urgency .48 .00
Model 1: R²: .40, with F: 26.34, Sig.: .00 Model 1: R²: .41, with F: 26.99, Sig.: .00
Case 3 ß sig. Case 4 ß sig.
Urgency .75 .00 Urgency 1.07 .00
p 19
Similar to the treatment of the correlation analyses, an overall regression analysis is conducted whereby all scores across the four experimental cases on sense of urgency, discrepancy, time pressure and willingness to change are combined (see table 7).
TABLE 7
Overall regression analysis of variables and Urgency construct on Willingness Model 1: R²: .68, with F: 174.20, Sig.: .00 Model 1: R²: .64, with F: 284.25, Sig.: .00
Overall ß sig. Overall ß sig.
Discrepancy .74 .00 Urgency .87 .00
Time Pressure .19 .00
R²: determination coefficient. ß: Unstandardized regression coefficient. Sig: significance (one-tailed)
Both overall regression models are capable of predicting willingness to change according to the estimated strong F-ratio’s (respectively 174.20, 284.25, with p < 0,01). The regression model with discrepancy and time pressure reveals a higher R² (68%), when compared to the overall regression model of the construct of a sense of urgency (R²: 64%). When taking the individual contributions of discrepancy and time pressure into account, it appears that both variables are significant predictors (p < 0.01 one-tailed), while discrepancy has more predictive power (with ß= .74) than time pressure (with ß= .19). Lastly, the construct of a sense of urgency is a significant predictor (with ß= .87, p < 0.01 one-tailed) of willingness to change.
4.1.3 Comparison of correlation and regression coefficients
Both correlation and regression coefficients of hypotheses 1a and 1b within and between the experimental cases are statistically compared using various techniques (depicted in Appendix D and E) in line with the research design.
Within cases. In the first place, a Steiger Z analysis to compare correlation coefficients within the experimental cases did not reveal any significant differences between discrepancy and time pressure. Although case 1 did show nearly significant difference between discrepancy and time pressure coefficients (Z= 1.949 and p= 0.051). Concerning the comparisons of regression coefficients within the experimental cases, it appeared that both case 1 and 2 entails significant difference between discrepancy and time pressure (respectively Z= 2.872 and p < 0.01, Z= 4.056 and p < 0.01). In these two cases I can conclude that discrepancy is a significantly stronger predictor of willingness to change than time pressure. Between cases. When using a Pearson-Filon analysis, based on Fischer r-to-Z transformed coefficients, to search for significant differences between the cases, it appears that the correlation coefficient of discrepancy in case 1 is almost significantly stronger than in case 4 (with ZPF= 1.893 and
p 20
In conclusion, discrepancy is a stronger predictor of willingness to change than time pressure in both case 1 and 2. Furthermore, when comparing correlation and regression coefficients between the experimental cases no significant difference was found between discrepancy coefficients. Time pressure in case 1 (and to a lesser degree for case 3) is significantly stronger correlated to willingness to change than in case 2. In addition, in case 4 time pressure is a significant stronger predictor of willingness than in case 2. However, this result was expected as the correlation analysis already revealed a deviant (insignificant) correlation score for time pressure in case 2. Overall, the comparison analysis did not reveal any consistent significant differences within and between the experimental cases. For instance, it was neither found that discrepancy, for all cases, is a significantly stronger predictor of willingness than time pressure (albeit signs in case 1 and 2), nor was one case superior over the others in the sense that both discrepancy and time pressure in that case were stronger predictors of (or stronger correlated to) willingness to change. Unfortunately, the comparison analyses do not yield any evident outcomes.
4.1.4 Conclusion statistical analyses
Based on the results of the statistical analyses conclusions can be made. Firstly, as hypothesized, perception of discrepancy positively affects willingness to change. Both overall and individual (i.e. per case) correlation and regression analysis support this finding. Secondly, based on both overall and individual correlation analysis time pressure is also positively related to willingness to change. The overall regression analysis proved support for this positive linear relationship; however the predictive power in case 2 was non-significant in line with the correlation analysis. In addition, based on the overall and individual correlation and regression analyses, the construct of perceived urgency also seems to positively affect willingness to change. Hence, the main hypothesis of this research is confirmed. Finally, when comparing both variables, it appears that discrepancy is the strongest predictor of willingness to change, although only significant difference between discrepancy and time pressure was found in two of the four experimental cases.
p 21
4.2 Results interview outcomes for triangulation purposes (B)
Interviews for triangulation purposes have been conducted to enhance validity of the statistical findings. Especially, the role of time pressure and willingness to change will be explored due to its subordinate role and insignificant correlation and regression score in one of the four cases. Nine representatives of the sample were interviewed, and their answers are analyzed by identifying relevant issues regarding urgency and willingness to change. Before describing the interviewee’s responses it is important to note that situational factors play an important role in assessing the willingness to change in public sector change initiatives as was indicated by all interviewees.
4.2.1 Time Pressure
When discussing the role of time pressure and willingness in change initiatives of the public sector the majority of the interviewees doubted that real time pressure in society exists. They can not recall any change project with such sincere pressures from societal demands or other threatening situations in need of an urgent response. Hence, they talk of a ‘self-created’ or ‘self-imposed’ time pressure based on a lengthy debated change plan with certain deadlines. One interviewee stated that ‘decision-making in the
public sector is rarely characterized by time pressure, only the result of that process that determines the period for effectuating a change initiative creates time pressure. Such a schedule for change is not bad per se, but it should give sufficient room for everyone to get involved and cooperate.’ Related to this statement another interviewee maintained that ‘…an adequate and properly prioritized time schedule is a
critical success factor for change in the public sector. In general, there are too many priorities of which only 20 percent gets executed, 30 percent will be outsourced and 50 percent is never looked at.’ In addition, inappropriate deadlines without addressing key conditions first works counterproductive.
Moreover, as argued by many interviewees, for public servants time pressure alone does not enhance willingness to change. Some of the interviewees functioned as Aldermen in large cities of the Netherlands for several years. They argued that public servants are extremely loyal to their directors or Alderman. In addition, ‘…the Mayor and Aldermen set the agenda for change initiatives and determine
the urgency and therefore time pressure for change in municipality. In fact, they decide what is going on in the municipality’, according to one of the interviewees. Public servants should comply with that agenda and focus on facilitating success of their Aldermen or directors. Otherwise, when concentrating on ‘urgent’ societal demands other than on the agenda, public servants are thought of as noisemakers and risking job loss. Therefore, ‘it is utmost important that an appropriate sense of urgency exists at the top of
the public organization’.
Besides this politically created time pressure, it is possible to intensify time pressure by promising change initiatives to the public or media with clear deadlines. This forces public employees to work hard towards these goals. In this case time pressure enhances necessity of change, while not paying attention to the willingness part. ‘A strong public director or Alderman knows whether his or her organization is
capable of bringing about change’, according to one interviewee.
Related to this, as interviewees discussed the non-existent relation between time pressure and willingness to change, they did contend that time pressure is needed on the individual level for personal productivity. ‘Unfortunately, my daily agenda is mostly determined by urgent matters, because time
p 22
change, but instead influences my personal effectiveness at a particular point in time’, according to one of the interviewees. Similarly, one interviewee stated: ‘it is unfortunate if time pressure overshadows the
substantive need for change. Time is of secondary importance, it could however function as catalyst’. As argued by another interviewee, ‘time pressure is always there but never decisive, you should avoid
situations with too much time pressure’.
Next to that, all interviewees argued that time pressure is in some way intertwined with the need for change, whereby time pressure is of minor importance. The effect of time pressure on willingness to change is somehow related to the effect of discrepancy on time pressure. Several interviewees maintained that time pressure either enforces the effect of discrepancy or diminishes it. This would imply a moderating influence of time pressure on the relation between discrepancy and willingness. Albeit, this moderating influence some other interviewees also suggested a direct influence of time pressure. One interviewee maintained that under certain conditions time pressure ‘…can have its own dimension when
pressures for change, external threads or internal dissatisfaction, become so sincere that a response is immediately needed. When this is clearly communicated and understood, time pressure then creates willingness to change’. Likewise, one interviewee suggested an inversely U-shaped (positively skewed) relationship to illustrate this direct influence of time pressure on willingness to change, stating: ‘without
time pressure for change there is no willingness to change. Too much time pressure for change makes people resistant.’
Finally, one interviewee questioned the methodology of the experiment; she argued that the items measuring willingness to change were noncommittal. She maintained that one could not make a proper consideration because willingness did not imply personal time-consuming consequences for a particular change effort to happen at the expense of daily (important) occupation. Therefore, despite low scores on time pressure, scores on willingness to change are rather high. In addition, some interviewees suggested that consultants are naturally more willing to change than public servants. As can be seen in Table 3, mean scores on time pressure in all cases are lower than mean scores on willingness to change.
To recapitulate, the interviewees doubt the actual existence of time pressure for change in the public sector since in fact it is self-imposed and initiated by directors in the public sector. Time pressure alone does not influence willingness to change per se. It depends on contingencies whether time pressure enforces the effect of need for change on willingness, or directly influences willingness. However, it is suggested that time pressure affects one’s personal effectiveness. The interviewee’s responses regarding the other variables of discrepancy and willingness to change are now briefly explored in the next section.
4.2.2 Discrepancy and Willingness to change
Discrepancy. The interviewees confirm that a sensed discrepancy naturally does lead to an increase of willingness to change. Unfortunately, the majority of the interviewees indicate that in change processes the need for change is often communicated, but not understood. One interviewee emphasized the necessity to: ‘…work towards a common frame of reference with both clients and their employees. This
means that everyone involved speaks the same language and there is a shared understanding of the need for change as well as clarity of the process to make change happen’. Another interviewee stated that:
p 23
The cause for change and the sincerity of the discrepancy play an important role in determining one’s willingness to change. According to one of the interviewees: ‘every organization has its own “pain
threshold”, when this level is not yet reached employees grumble and whine and willingness to change is not yet there’. The majority of the interviewees argued that in times of negative economic or political developments, like the current hard cuts in the Dutch public sector, a favorable climate for change is created. Hence, ‘it is currently easy to convince employees of the need for change, since there is a broad
acceptance that cuts are needed to restore the damage of the financial crises’. Moreover, the interviewees agree that change initiatives arise out of either internal or external pressures, mostly powered by financial motives.
By one other experienced consultant it was argued that the need for change is closely related to the organization’s vision and mission. In addition, most of the interviewees made a distinction between inward and outward focused organization, stressing that for the latter the influence of discrepancy on willingness to change is probably stronger. Unfortunately, as one interviewee indicated: ‘the public sector
is mostly characterized by introverted organizations’.
Furthermore, a sense of urgency (based on discrepancy and time pressure) is not necessarily needed to create willingness to change as indicated by half of the interviewees. It is good to have but not necessarily a critical factor for creating willingness or the effectiveness of a change effort in general. One interviewee stated: ‘if a broadly sensed urgency is there it definitely helps, because it makes the process
go smoother with less hurdles to take. However, by clear communication and letting employees participate in determining the direction of change, with or without a need for change, does work as well.’
Willingness to change. Overall the interviewees agree that willingness to change is an essential
condition for effectively implementing change, it would otherwise be like: ‘…fighting a losing battle’, according to one interviewee. She furthermore maintained that: ‘willingness to change is a prerequisite
for effectively changing an organization; it can however steadily grow as a change project continues to evolve. It is important to just start with the support of the critical mass, the rest will follow eventually.’
Another interviewee indicated that: ‘one can not force employees to be willing to change, conflict
situations should be avoided at all cost’.
One consultant with extensive experience in advising large public sector reorganization efforts argued that: ‘…next to the importance of creating a widely understanding of the need for change, it
should be clear what the precise consequences are for those involved. Individually, willingness to change has to do with maintaining 1) position and status, 2) salary, 3) workplace and travel distance. If those issues are appropriately addressed then in general employees are willing to change.’ Most of the time in reorganization efforts, it is impossible to give certainty about these three personal issues in an early state the change effectuation. Therefore, employees need to be offered ‘…procedural certainties during
transition phase’ so that resistance temporarily fades.
Finally, according to most of the interviewees, willingness to change comes forth out of loyalty and equanimity of the public servants driven by political pressures. The former Aldermen in particular, praised the loyalty of a public servant, because: ‘…they are motivated to do their jobs well, or try to get
p 24
To summarize, the interviewees agree that discrepancy leads to willingness to change. The strength of this relationship is determined by the sincerity of the cause for change, but also by present conditions in society. A sense of urgency is, however, not a prerequisite for change, in contrast with a shared willingness to change. Personal issues play an important role in determining the level of willingness.
4.2.3 Conclusion triangulation interviews
The experiment revealed clear outcomes regarding discrepancy and its influence on willingness to change. However, results showed somewhat ambiguous outcomes for the relation between time pressure and willingness to change. The interviews for triangulation purposes revealed that indeed time pressure is of minor importance. Next to that, the interviewees doubt the actual existence of time pressure for change in the public sector. They talk about self-determined or self-imposed time pressure based on lengthy debated change plans. Based on the level of this self-imposed time pressure, it can affect willingness directly or influence the relation between discrepancy and willingness.
In line with the results of the experiment the interviewees confirm the existence of a direct positive relationship between discrepancy and willingness to change. Shared understanding of the need for change is important, next to a widely spread sense of urgency. It is, however, not an indispensable factor for willingness to change or effectively implementing change efforts, as long as an appropriate sense of urgency exists at the top of the organization. Next to that, public servants are in general extremely loyal to their employer which results in an unresisting attitude regarding proposed change initiatives. Nevertheless, the consequences of such initiative concerning personal issues as position and salary are important in inducing willingness to change.
p 25
4.3 Results in-depth interviews in public sector change initiatives (C)
This final part of the results section demonstrates the quantitative and qualitative outcomes of multiple in-depth interviews concerning sense of urgency and willingness to change in actual change processes in the Dutch public sector. Firstly, the overall quantitative responses are elaborated and displayed in two figures. In addition, the quantitative results are split out by the interviewee’s role during the change project, either change manager or recipient. Furthermore, outcomes of this first analysis are enriched with a qualitative exploration for explaining the most interesting findings.
4.3.1 Overall quantitative interview outcomes
In total 20 in-depth interviews were conducted with change managers and recipients in the public sector representing seven significant change initiatives. Figure 3 summarizes the mean scores of the responses on questions regarding the level of discrepancy, time pressure, sense of urgency and willingness to change with a 95% confidence interval based on a 5-point Likert scale. Despite low scores on discrepancy, time pressure and urgency, scores on willingness to change are superior with a mean of 4,55. In addition, scores on willingness to change are hardly dispersed compared to the other variables. Scores on time pressure are more dispersed with the lowest mean of 2,85. Similarly, scores on sense of urgency are equally dispersed with a higher mean score of 3,55. Finally, the degree to which interviewees had a feeling of discrepancy before a change initiative started scored on average 4,15.
FIGURE 3 Mean scores per variable
4.3.2 Comparing change manager and recipient outcomes
p 26
and extreme scores for the change recipients. Discrepancy mean scores for change managers and recipients are respectively 4,78 and 3,64. Scores for time pressure are more diverse when comparing change managers with change recipients (respectively 3,89 and 2,00). In addition, scores on sense of urgency are also rather dispersed (respectively 4,33 and 3,00). Finally, scores on willingness to change do not differ that much (respectively 4,78 and 4,36). Further statistical analysis for comparing means, or correlation and regression is not possible due to the small sample size.
FIGURE 4
Error bar chart for mean scores per variable and role
4.3.3 Conclusion quantitative outcomes
