Organizational change: Ready, set, go?

(1)

Organizational change:

Ready, set, go?

Master thesis, Msc Business Administration Specialization: Change Management

University of Groningen Faculty of Economics and Business

December, 2011

R. de Jong University of Groningen Faculty of Economics and Business

Eendrachtskade Zuidzijde 194 9726 DD Groningen Tel: +31 (0) 641337793

E-mail: rianne.de.jong@live.nl

Supervisor/ unive rsity

J. Rupert / C. Reezigt

Supervisor/ field of study

R. Admiraal (Concern controller), Gemeente Súdwest Fryslân

Acknowledgme nt:

First and foremost, I would to thank Joyce Rupert for her guidance throughout this research project. She helped me were needed and always provided me with critical but honest

(2)

Rianne de Jong, December 2011

Organizational change:

Ready,set,go?

ABSTRACT

This research explores the extent to which the change message components (appropriateness, self- and computer efficacy, management support and personal valence) have an effect on the level of readiness for change and how organizational identification influences this

(3)

Rianne de Jong, December 2011 TABLE OF CONTENTS

1. INTRODUCTION ... 5

1.1 DESCRIPTION OF THE ORGANIZATION ... 5

1.2 CHANGE PROCESS ... 6

1.3 RESEARCH QUESTION ... 7

2. THEORY... 9

2.1 ORGANIZATIONAL CHANGE ... 9

2.2 READINESS FOR CHANGE ... 9

2.3 CHANGE MESSA GE ... 11

2.3.1 Perceived level of appropriateness ... 12

2.3.2. Perceived level of efficacy ... 13

2.3.3 Perceived level of management support ... 14

2.3.4 Perceived level of personal valence ... 14

2.4 ORGANIZATIONAL IDENTIFICATION ... 15

2.4.1 Moderation relationship appropriateness &readiness ... 17

2.4.2 Moderation relationship self~ and computer efficacy & readiness ... 18

2.4.3 Moderation relationship management support & readiness... 19

2.4.4 Moderation relationship personal valence & readiness ... 20

2.5 CONCEPTUAL MODEL... 22

3. RESEARCH METHOD ... 23

3.1 SAMPLE AND PROCEDURE ... 23

3.2. MEASURES ... 24

3.2.1 Extreme values and outliers ... 24

3.2.2 Measurements (factor analysis & reliability)... 24

3.2.3 Dependent variable: Readiness for change... 25

3.2.4 Independent variables: Appropriateness, efficacy, management support and personal valence ... 25

3.2.5 Moderator variable: Organizational identification ... 27

3.2.6 Control variables ... 28

3.3 DATA ANALYSIS ... 28

3.3.1 Descriptive statistics and correlations ... 28

3.3.2 Regression analysis ... 28

4. RESULTS... 30

4.1 DESCRIPTIVE STATISTICS AND CORRELATIONS ... 30

4.2 REGRESSION ANALYSIS... 32

4.2.1 Regression analysis linear relationships ... 32

4.2.4 Multiple regression analysis moderation ... 33

4.2.2 Additional regression tests ... 34

(4)

5.1 LIMITATIONS AND FUTURE RESEARCH ... 39

6. CONCLUSION AND MANAGERIAL IMPLICATIONS... 40

7. REFERENCES ... 42

APPENDIX A: ORGANIZATIONAL CHART MUNICIPALITY OF SÚDWEST- FRYSLÂN ... 49

APPENDIX B: QUESTIONNAIRE ... 50

APPENDIX C: SAMPLE S TATISTICS... 58

APPENDIX D: FACTOR ANALYSIS ... 62

(5)

Rianne de Jong, December 2011 1. INTRODUCTION

Kavaliauskaite (2010) indicates that organizations in the public sector, including municipalities, often receive critique concerning their ineffectiveness. These organizations usually deal with many layers in the decision making process which prevents them from making speedy decisions. To increase this speediness, technology has become an important tool for governmental institutions (Gascó, 2003). This does not only account for IT systems that connect the government and municipalities to citizens, but also for IT systems that

support internal processes. Loukis and Tsouma (2000) refer to well known advantages of such internal IT systems including increases in productivity and decreases in costs. Based on these advantages, the goal of implementing IT systems is usually to increase efficiency and

effectiveness. This goal also applies to the municipality of Súdwest Fryslân (SWF). After having merged in January 2011, the municipality decided on the use of one overarching IT system for internal use; Corsa. This research focuses on the realization of the implementation of this system and specifically on the level readiness for change of the employees of the municipality of Súdwest Fryslân (SWF). The following sections include a description of the organization, the change process and explain the research question that guides this paper.

1.1 Description of the Organization

The municipality of SWF has only recently been established. On the first of January 2011, five local municipalities; the municipality of Sneek, Nijefurd, Bolsward, Wûnseradiel and Wymbritseradiel, merged into the current municipality. The decision to merge was mainly taken in order to pursue a durable increase in governmental authority and to improve the (digital) service given by the municipality to the residents.

Geographically speaking, the municipality of SWF is the largest municipality in the Netherlands. Six cities and 63 villages are located in the area. The municipality is the second biggest municipality of the province of Friesland in regard to the number of residents (82.000) and amount of jobs (30.000). Currently, the municipal organization consists of six

(6)

people are employed by the municipality of SWF. An overview of the organizational chart can be found in appendix A.

1.2 Change Process

Quite recently, as mentioned, the municipality of SWF officially merged on the first of January 2011. Since the offices of the municipality are still located all around Friesland, it became very clear that the „old‟ way of handling documents (analogue) needed to be replaced by a digital system. Therefore, Corsa was chosen to be one of the leading IT-systems in the newly established municipality. To clarify, the implementation of Corsa is a significant consequence of the merger.

Four processes should be supported by Corsa; Incoming & outgoing mail, archive registration and the process of municipal decision making („Bestuurlijke besluitvorming‟). This latter process encapsulates all proposals that are produced by officials within a municipality. Starting with the productio n of a proposal by an official, a proposal goes through several stages of the decision making process and eventually is (if it is approved by several stakeholders) proposed to the Mayor and Executive Board (College van

Burgemeesters & Wethouders). By using Corsa, all documents that were previously handled analogue, are now documented in Corsa. Subsequently, many other processes that are linked to the document registration can also be handled digitally.

Firstly, the most obvious aspect of this change relates to working methods and concerns the transfer from paperwork (analogue) to digital documents. This especially

accounts for the process of municipal decision making since the documents will be transferred through the decision making chain digitally & automatically when a document has been approved. Ultimately, the document arrives at the Mayor and Executive Board. Compared to the „old‟ situation, the transparency within the organization increases since employees can easily see „where‟ the document is and „who‟ is working on it.

Also, documents (for example incoming and outgoing mail) are now being linked together in a so called „case‟ where they previously were registrated as a separate document. In this way, all aspects that concern the same person and/or event are placed together which makes it easier for officials to find documents regarding the same case.

Several other processes that are supported by Corsa include automatic agenda generation for meetings of the major and the executive board, internal memo‟s, notes of telephone

(7)

The use of Corsa requires people to change their way of working significantly. Firstly, an important element of the use of Corsa is that all documents have to be documented and saved within the system which was previously not done on a consistent basis. This documentation includes (perceived) extra work for employees and makes them more dependent upon the system. Due to this dependency, employees will have to acquire (new) computer skills and knowledge about processes. Also, employees receive a large part of their „tasks‟ related to a certain document digitally where previously, especially in the smaller municipalities (that existed before the merger), many analogue documents were used and documents were

discussed face to face. Employees have to check their workload in Corsa, accept a certain task and consequently, finish that task to eventually, digitally, forward the revised (or new)

document to either their supervisor or other employees. Due to this process, much face-to-face contact is lost. Besides, Corsa gives team managers options to check the workload of employees (within their department) which co uld increase the amount of control within a department. On the hand, this could have a positive effect when managers use this option to better deploy resources among their employees. This could for example result in a decrease of the workload of one employee and adding tasks to the workload of another employee who is less occupied. On the other hand, negative effec ts could arise since control from the upper levels in the organization increases and people feel as if they are being checked and have less freedom in when to perform their tasks.

This change effort does not include all employees of the municipality of SWF, since Corsa will not be used by the operational organizational units such as the fire department and some organization units of Villages &cities(for example maintenance). This research will focus on the organizational units that will use Corsa including; Business operations, Development, Backoffice, the corporate staff and some parts of Public and Realization. In total, around 300 people will work with Corsa in the near future.

1.3 Research Question

(8)

causal factor of unsuccessful change is the fact that employees feel the organization is not ready for change and as a consequence, do not accept the change. The key mechanism to achieve readiness for change is considered to be the change message (Armenakis et al. 1993) As such, the management has requested to investigate the change message and how

this message influences the readiness for change of the employees of the municipality of SWF who will be using Corsa.

Aforementioned, implementing Corsa is a significant consequence of the merger between the five municipalities. Interestingly, Bartels, Douwes, De Jong, & Pruyn (2006) found that organizational identification is an important concept in and after merger processes since employees could fell less identified with the newly established organization as a result of the merger. The level of organizational identification has been found to be connected to the will to strive for organization goals (Elsbach and Glynn, 1996). In this particular research, implementing Corsa is an important organizational goal and therefore, assumed is that organizational identification could play a significant role in this change effort. Besides, a preliminary enquiry showed that employees still refer to their old organization a lot which again could imply that the relationship between the change message and the level of change readiness is influenced by organizational identification.

The above information results in the following research question;

(9)

Rianne de Jong, December 2011 2. THEORY

In order to be able to provide theoretical insight in the concept of change readiness, a short introduction in the concept of organizational change needs to be provided. Thereafter, the concept of change readiness will be described which is followed by an explanation of the antecedents of change readiness and the possible influence of organizational identification on the relationship between these antecedents and change readiness.

2.1 Organizational Change

Several authors have provided theoretical contributions to improve the understanding of organizational change, which can be defined as “a situation that interrupts the normal patterns of an organization” (Choi & Ruona, 2011: 51). Beer and Nohria (2000) for example, who distinguish between theory E and theory O, respectively hard and soft approaches to change. Weick & Quinn (1999) also categorize change, specifically discerning episodic change from continuous change. The former type of change, episodic change, is considered to be intentional and infrequent. This type of change is very much connected to the most quoted model of change produced by Lewin (1951). He is famous for his three stage model which describes the sequence of unfreezing, changing and refreezing. The first phase requires a destabilization of the equilibrium in order to prepare for change and make sure that the old behavior is „unlearnt‟ and can be replaced by new behavior. The second stage concerns the actual moving from the current state to the desired state whereas the third stage seeks to “stabilise the group at a new quasi-stationary equilibrium in order to ensure that the new behaviors are relatively safe from regression” (Burnes, 2009: 339). Specifically the first phase of the model of Lewin (1951), unfreezing, has been given considerable attention with regard to the concept of change readiness. Arme nakis et al. (1993) as well as Eby, Adams, Russell, & Gaby (2000) and, quite recently, Choi & Ruona (2011) claim that change readiness is grounded in the unfreezing stage of the model of Lewin (1951). Among other authors, all these authors argue directly or indirectly that change readiness is pivotal to produce

successful change.

2.2 Readiness for Change

Several definitions of readiness for change can be found in the literature.

(10)

to which changes are needed and the organization‟s capacity to successfully make those changes”. They add that change readiness is the cognitive aspect of change, determining the behavior of employees, whether that includes supporting or resisting the change effort. Eby et al. (2000) and Jansen (2000) utilize similar definitions, both

emphasizing perceptions about the organization being ready for change and able to execute the change, where Jansen (2000) specifically accentuates the necessary belief of individuals that the change is needed (similar to Armenakis et al. (1993)). Jones, Jimmieson & Griffiths (2005) underscore another aspect; the belief that employees have about whether the change benefits themselves and the organization as a whole. Recently, Holt, Armenakis, Field & Harris (2007) described change readiness as “collectively reflecting the extent to which an individual or individuals are cognitively and emotionally inclined to accept, embrace, and adopt a particular plan to purposefully alter the status quo”. The most striking similarity in all these definitions is the fact that change readiness is about the belief of an individual, not the organization as a whole. Eby et al. (2000) clarify this by explaining that every employee has its own interpretation of the objective reality. They refer to Spreitzer (1996) who already argued that employees can have very different perceptions even though they experience similar objective realities and the same organizational context. This would especially apply for a research context in which the organizational context just has undergone significant changes, such a merger, which is applicable in this specific research. A second similarity can be found in the aspects of the belief of the employee; employees have to feel they need the change, they need to believe that they and the organization as a whole are capable to make the change happen and finally, added by Jones et al. (2005), they must feel the change is

beneficial for them as well as for the organization.

(11)

Besides the vital importance of change readiness in general change efforts, its

significance for IT projects has also been highlighted in literature. Markus (2004) found that lack of change readiness is one of the major risks in IT related organizational change. Kwahk & Kim (2008) argue that resistance towards IT-driven change is one of most important reasons for failure in IT projects and add that readiness for change is a way to reduce this resistance.

After highlighting the importance of readiness for change, the next step is to clarify how to accomplish readiness for change. Armenakis et al. (1993) believe that the change message delivered by the management of an organization (to the employees) is the key mechanism in achieving readiness for change. The next section will discuss this change message in detail.

2.3 Change Message

Self & Schraeder (2009) clarify that the change message should be communicated in such a way that it reduces the uncertainty that could arise due to the introduction of a change. They explain that the change message should be used to address five questions of employees; “(1) Is the change necessary?, (2) Is the change being introduced the right change to make?, (3) Are key organizational members supportive of the change?, (4) Do I or we (the

organizational members) have the ability to successfully implement the change? and (5) What is in it for me if we change?”(Self & Schraeder, 2009:172). The answers to these questions can be derived from the components of the change message (in order); Discrepancy,

appropriateness, management support, self-efficacy and personal valence.

Initially, Armenakis et al. (1993) only p ropose two components of the change

message. They stated that the change message should include: “( a) The need for change, that is, the discrepancy between the desired end state (which must be appropriate for the

organization) and the present state; and, (b) the individual and collective efficacy (i.e. perceived ability to change) of parties affected by the change effort”(Armenakis et al., 1993:684). In 1999, Armenakis, Harris &Field present the three other components;

appropriateness, management support and personal valence. Many authors still present these five components (E.g. Self & Schraeder, 2009). However, Holt et al. (2007) developed an instrument to measure change readiness, and proved that the measurement scale of

antecedents for change readiness actually consists of four factors.

In their research of 2007, an analysis of the literature, interviews (qualitatively analyzed) and open-ended questionnaires (distributed among public and private sector

(12)

support and personal valence (Holt et al., 2007). Strikingly, these factors already differ from the five factors that were distinguished by Armenakis et al. in 1999. Scale items were developed for the five factors and subsequently, content adequacy tests and factor analyses were executed (Holt et al., 2007). The content adequacy tests showed that people tend to relate personal valence to organizational valence and also view the concepts of discrepancy and organizational valence as being very similar to each other. These findings were reinforced by the factor analysis and therefore, some items of personal valence and all items of

discrepancy and organizational valence were combined in one factor which was labeled appropriateness. All together, the research of Holt et al. (2007) results in 4 antecedents of change readiness (the change message components); Appropriateness, efficacy, management support and personal valence. The sample of Holt et al. (2007) consisted of 900 managers in both the public and private sector which, in addition to their thorough tests, provides strong evidence for these 4 antecedents.

Therefore, this research will also focus on 4 change message components. The following sections will elaborate on each of the change message components.

2.3.1 Perceived level of appropriateness

The first component of the change message is appropriateness. This component, according to Neves (2009:217); “focuses on the fact whether the specific change is adequate for the organization and if individuals agree with it. It emphasizes organizational values and if organizations change in accordance with such values”. The management of the organization should communicate that a specific change effort is suitable for the context in which the company operates (Self & Schraeder, 2009). Individuals might still not dedicate themselves to the change if they believe the „solution‟ is not appropriate for the current problems (Bernerth, 2004).

Based on the findings of Holt et al. (2007), appropriateness also reflects the need for change. If employees believe that the current situation is effective and efficient and does not need to change, it is difficult for them commit to the change (Bernerth, 2004). Employees need to experience that there is a “clear and present danger; a tangible and immediate problem that must be confronted if the organization is to remain economically viable ” (Beer et al., 1990:55). This statement already implies that employees need to view the change effort as being beneficial to the organization as a whole.

(13)

are going to feel much better about changes being implemented when they feel those changes are appropriate” (Cole et al., 2006:354). Subsequently, Neves (2009) found that

appropriateness of change is significantly, positively related to what he refers to as „affective commitment to change‟. As such, a higher level of perceived appropriateness should result in a higher level of readiness for change. This results in the following hypothesis.

Hypothesis 1: A high level of perceived appropriateness will be associated with a higher level of change readiness of employees within an organization

2.3.2. Perceived level of efficacy

Armenakis & Harris (2009: 129) define efficacy as “The belief that the change recipient and the organization can successfully implement a change”. Cunningham, Woodward, Shannon, MacIntosh, Lendrum, Rosenbloom, & Brown (2002:378) more

specifically talk about self-efficacy, the second component of the change message, which they define as; “the perceived ability to cope with change”. Armenakis, Harris & Mossholder (1993) indicated that a higher level of perceived efficacy should result in a higher level of readiness for change. Cunningham et al. (2002) found that self-efficacy indeed is positively related to readiness for organizational change. In their study, employees proactively

approached problems during their work and were, therefore, more confident about their ability to cope with change (Cunningham et al. 2002). Besides general efficacy, the term computer self-efficacy is applicable in the case of IT implementations. Shivers-Blackwell &Charles (2006) have tested the relationship between computer self –efficacy and readiness toward change. Their research did not show a significant positive relationship between these two concepts but did however, show a positive relationship between computer self- efficacy and attitude towards the use of an IT system (TAM model, Davis, 1989). Since this research was performed among students and their findings do imply that there is some sort of connection between computer self- efficacy and change readiness, it seems to be interesting to find out whether this research provides more insight in that relationship. As such, the following hypothesis was designed;

Hypothesis 2: A high level of perceived self~ and computer efficacy will be associated with a higher level of change readiness of employees within an

(14)

2.3.3 Perceived level of management support

Holt et al. (2007) indicate management support as the third component of the change message, one of the four most influential factors (components of the change message) to create readiness for change. Management support is defined in terms of “demonstration of support for the change by managers” (Kavaliauskaite, 2010:82). Cinite, Duxbury & Higgins (2009) focused their research on organizational readiness within the public sector. They also found, as well as Holt et al. (2007), a significant positive relationship between high

commitment of senior management to the change and readiness for change. Subsequently, their research also showed that employees perce ive a higher readiness for change when they receive a high level of support from their immediate manager. As Self & Schreader (2009) also clarify, when a change is announced by senior management, employees tend to ask for explanations and seek support with their immediate manager. They refer to Covin & Killman (1990) who argued that a negative perception of the change can be caused by a lack of visible support or inconsistent behaviors from managers. Whereas the opposite is also true, visible support and appropriate behavior created a positive perception. In this latter case, managers act as role models for employees illustrating responsibility and credibility with respect to a change effort (Covin & Killman, 1990).

In addition, research specifically focused on information systems (IS) also shows that management support is a critical success factor within IT implementations ( Young & Jordan, 2008; Akkermans & Van Helden, 2002). Based on the information above, the following hypothesis is designed.

Hypothesis 3: A high level of perceived immediate and senior management support will be associated with a higher level of change readiness of employees within an organization

2.3.4 Perceived level of personal valence

(15)

the change will bring for them. As opposed to readiness, Cinite et al. (2009) also

conceptualized the concept of unreadiness for change. Their findings show that perceptions of not being ready for change are mainly due to two forms of organization actions. One of these organizational actions is poor communication of the change. In practice, this is related to not providing a vision for the change, not clearly explaining the reasons for change and not clearly indicating the expected outcomes and benefits for the change. Actually, all t hese points can be related to the change message components. The latter however, a clear

indication of the expected benefits and outcomes, can be directly related to personal valence. The research of Cinite et al. (2009) shows that not providing this information about expected benefits could potentially result in unreadiness for change. This information implies that if personal valence is adequately addressed in the change message this could result in higher levels of change readiness. Based on this implication, the following hypothesis was designed;

Hypothesis 4: A high level of perceived personal valence will be associated with a higher level of change readiness

Besides the change message components, organizational commitment has often been discussed as an important factor in organizational change, especially in relation to acceptance of change (Darwish, 2000; Cordery, Sevastos, Mueller & Parker,1993). Organizational commitment can be defined as “the relative strength of an individual‟s identificatio n with and involvement in a particular organization” (Steers, 1977: 46). This definition already implies that organizational identification is a part of organizational commitment, as is confirmed by Cook & Wall (1980). They identified involvement, loyalty and identification as the three primary components of organizational commitment and associated identification with pride and feeling connected to an organization (Cook & Wall, 1980). Bartels et al. (2006) found that organizational identification is crucial during merger processes. Since this research includes a change project within a few months after a merger, it seems particularly important to investigate this concept more thoroughly. Therefore, the following section will review organization identification.

2.4 Organizational Identification

(16)

The concept is grounded in social identity theory, which focuses on people deriving their identity from being part of a certain group (Umphress, Bingham & Mitchell, 2010).

Subsequently, organizational identification has often been linked with affective commitment, one of the forms of organizational commitment as been described by Allen and Meyer (1990). They define affective commitment as “emotional attachment to, identification with, and involvement in the organization” (Allen & Meyer, 1990:1). However, as Asforth et al. (2008) clarify, the two concepts differ in the sense that commitment is about being satisfied or happy with the organization whereas organizational identification is linked to perception of the self in relation to the organization.

Also, commitment may be easier to transfer to another organization that aims for a similar positive attitude whereas identification is specific to an organization. To clarify, easier transferability of commitment can be explained by the fact that commitment is more rational where identification has many non-rational components. (Asforth et al., 2008).

Besides being specific to one organization, it is important to identify the consequences of organizational identification. Van Dick (2001) found that a high level of organizational identification has a positive effect on the intention to stay within the organization and organization citizenship behavior. This latter concept refers to so called extra role behavior, which aims to strengthen the effectiveness of an organization (Van Dick, Grojean, Christ, & Wieseke, 2006). It can be defined as “individual behavior that is discretionary, not directly or explicitly recognized by the formal reward system, and in the aggregate promotes the efficient and effective functioning of the organization” (Organ, Podsakoff, & MacKenzie, 2006: 3). Ellemers, de Gilder & Haslam (2004) refer to another important consequence; identification with the organization tends to result in intrinsic motivation as employees internalize collective goals. Dutton (1994) goes even further by arguing that employees link their own survival to the survival of the entire organization. As such, employees who strongly identify with their organization will pursue efforts that benefit the organization (Ashforth & Anand, 2003).

It should be noted that, as Albert & Whetten (1985) claim, organizational

identification is often considered to be a latent concept. They argue that identification with organizations only becomes salient in times of substantial change, for instance during

restructuring processes and mergers. In research about this latter concept, authors specifically distinguish between pre- merger and post-merger identification (Bartels, Pruyn, de Jong, 2009; Gleibs, Mummendey, Noack, 2008; van Dick, Wagner, Lemmer, 2004). Among other

(17)

with the „old‟ organization) strongly influences post-merger identification. Earlier, Bartels, Douwes, de Jong and Pruyn (2006) already indicated that there is a positive relationship between pre- and post merger identification when employees perceive the merged organization as a partial continuation of the „old‟ organization. Reaso n for this is that

organizational identification can be (partly) transferred. On the other hand, when a merger is perceived as (partly) discontinuous and therefore seen as a threat to identity, there is no or a negative relationship between pre- and post merger identification.

Interestingly, Gleibs et al. (2008) measured post-merger six months after a merger and their results showed that at that time, respondents still felt more identified with their old

organization and therefore were refusing to identify with the new organization. Based on these findings, employees would be less likely to display the consequences of organizational identification described above including; organizational citizenship behavior, internalizing collective goals, linking their own survival to that of the organization and pursuing efforts that benefit the organization. That is, it can be assumed that employees will be less likely to

support change efforts linked to a merger.

Based on this information, it can be concluded that there are distinct differences between employees identifying with the so called „old‟ organization and the „new‟

organization in mergers and that the former could influence the latter. The findings of Albert & Whetten (1985) and Bartels et al. (2006) emphasize that organizational identification is a highly important element during change efforts, and it is assumed that the concept of pre- and post merger identification is likely to play a significant role in the relationship between each of the change message components and the level of change readiness. Therefore, this concept is included in this specific research.

2.4.1 Moderation relationship appropriateness &readiness

Employees who feel that a specific change effort is needed, beneficial for the organization and appropriate for the situation could be more ready for change (H1 ).

Also, if employees identify highly with the new organization they will internalize collective goals (Ellemers, de Gilder & Haslam, 2004) and as such, see the goals of the „new‟ organization (a specific change) as their own goals. In this situation, this would result in a high level of organizational identification with the new organization. This high level of identification with the new organization could amplify the relationship between the perceived level of appropriateness and the level of change readiness. On the other hand, when

(18)

their own goals (high level of organizational identification with the old organization), this might weaken the effect on the relationship between the perceived level of appropriateness and the level of change readiness. In this case they might not see a collective goal in favor of the new organization, such as successfully executing a change effort, as their own goal. The information above results in the following hypotheses;

Hypothesis 5a: The relationship between the perceived level of appropriateness and the level of change readiness will be positively moderated by a high level of

organizational identification with the new organization. In this respect, the positive relationship between the perceived level of appropriateness and the level of change readiness will be reinforced when organizational identification with the new organization is high.

Hypothesis 5b: The relationship between the perceived level of appropriateness and the level of change readiness will be negatively moderated by a high level of

organizational identification with the old organization. In this respect, the positive relationship between the perceived level of appropriateness and the level of change readiness will be weakened when organizational identification with the old

organization is high.

2.4.2 Moderation relationship self~ and computer efficacy & readiness

The second hypothesis states that a high level of self~ and computer efficacy should result in a higher level of change readiness (H2 ). In addition, a high level of organizational identification with the new organization co uld strengthen the interaction between these two components. The reasoning for this is that employees engage in extra role behavior, meaning that they display behavior that goes beyond their task to support a change effort

(organizational goal), due to a high level of identification with the new organization.

If employees feel they are able to execute the change and are convinced of their own abilities (efficacy) and subsequently feel highly identified with the new organization, their change readiness will be even higher. In this situation, a high level of organization identification with the new organization positively moderates the relationship between the level of self~ and computer efficacy and the level of change readiness.

(19)

organizational identification with the old organization indicates that employees are willing to pursue extra role behavior for the old organization. Consequently, they might be less open to engage in extra role behavior for this change since this change effort is executed in the new organization. That is, if employees are confident in their abilities (efficacy) the strength of the relationship between self- and computer efficacy and readiness for change might be weakened by a strong identification level with the old organization.

Hypothesis 6b: The relationship between the perceived level of self~ and computer efficacy and the level of change readiness will be negatively moderated by a high level of organizational identification with the old organization. In this respect, the positive relationship between the perceived level of self~ and computer efficacy and the level of change readiness will be weakened when organizational identification with the old organization is high.

Hypothesis 6a: The relationship between the perceived level of self~ and computer efficacy and the level of change readiness will be positively moderated by a high level of organizational identification with the new organization. In this respect, the positive relationship between the perceived level of self~ and computer efficacy and the level of change readiness will be reinforced when organizational identification with the new organization is high.

2.4.3 Moderation relationship management support & readiness

It is hypothesized that a high level of perceived immediate and senior management support results in a higher level of change readiness (H3). When employees would feel highly identified with the new organization they pursue efforts that benefit the new organization (Ashforth & Anand, 2003). Successfully executing a change process would also fall within the category of actions that benefit the new organization.

(20)

new organization, assumed is that they might not be open to pursue efforts within this

organization (negative moderation). Therefore, the relationship between the perceived level of immediate and senior support and the level of change readiness for this specific change effort might decrease in strength (negative moderation) by a high level of organizational

identification with the old organization.

Hypothesis 7a: The relationship between the perceived level of immediate and senior management support and the level of change readiness will be positively moderated by a high level of organizational identification with the new organization. In this respect, the positive relationship between the perceived level of immediate and senior

management support and the level of change readiness will be reinforced when organizational identification with the new organization is high.

Hypothesis 7b: The relationship between the perceived level of immediate and senior management support and the level of change readiness will be negatively moderated by a high level of organizational identification with the old organization. In this respect, the positive relationship between the perceived level of immediate and senior management support and the level of change readiness will be weakened when organizational identification with the old organization is high.

2.4.4 Moderation relationship personal valence & readiness

Finally, it was hypothesized that the level of personal valence should result in a higher level of change readiness (H4). That is, when employees see the benefits for themselves and for the organization, they should be more open towards change (readiness).

(21)

because employees do not link their survival to the new organization (but to the old

organization) and consequently, might be less supportive of the change effort. To conclude, a high level of organizational identification with the o ld organization might result in a weaker interaction between personal valence and the level of change readiness.

Hypothesis 8a: The relationship between the perceived level of personal and

organizational valence and the level of change readiness will be positively moderated by a high level of organizational identification with the new organization. In this respect, the positive relationship between the perceived level of personal and organizational valence and the level of change readiness will be reinforced when organizational identification with the new organization is high.

Hypothesis 8b: The relationship between the perceived level of personal and

(22)

Rianne de Jong, December 2011 2.5 Conceptual Model

The proposed hypotheses in the theory section above result in the following conceptual model (figure 1).

FIGURE 1 Conceptual Model

The following section describes the methods that are used to test the hypotheses represented in the model above.

Perceived

appropriateness(H1)

Perceived self ~and computer efficacy(H2)

Perceived management support (H3)

Perceived personal valence (H4)

Readiness for

organizational change

Organizational identification(H5-H8):

(23)

Rianne de Jong, December 2011 3. RESEARCH METHOD

3.1 Sample and Procedure

To test the hypotheses presented in the chapter above, a questionnaire was distributed in the municipality of SWF (see Appendix B). As mentioned, around 300 people will work with Corsa in the near future.

Firstly, this questionnaire was distributed digitally. To accomplish this, questions were translated in Dutch and with the help of Qualtrics survey software, the questionnaire was created online. Before actually sending the questionnaire, a small pilot of four persons was set up to test whether the questions were clear, to see how much time respondents would need to fill in the questionnaire and to determine whether the software was working appropriately. As a result of the pilot, some minor adjustments were made to the formulation of the questions. Thereafter, the questionnaire was send to 200 people by mail. A week before the deadline, a reminder was sent to all these employees. This method resulted in 56 completed

questionnaires, a response rate of 28%. Secondly, to increase the number of respondents, around 100 people on the location in Sneek and Ijlst were also approached directly using questionnaires on paper. This second method turned out to be very successful and another 56 questionnaires were gathered. The 56 % response rate was significantly higher with this second method. Thus, in total, 112 questionnaires out of 300 questionnaires were gathered with an average response rate of 37,3%.

From the 112 respondents, 42,9 % are male and 57,1% is female. Their age ranges from 21 to 65 but the largest part (87,5%) falls in the category of 31 till 60 years old. All respondents are Dutch. Almost half of the respondents (49,11 %) has completed a higher vocational education (HBO), whereas 24,11% has completed intermediate vocational training (MBO) and 17,86 % of the respondents has graduated at a university (WO). Currently, 48,21 % of the respondents works as a policy advisor (beleidsadviseur) or policy officer

(24)

Rianne de Jong, December 2011 3.2. Measures

The questionnaire originally consisted of 43 items, which represented the six variables in the conceptual model (page 22). To execute analyses on all scales, SPSS (version 16.0) was used. Before these analyses were performed, negatively formulated items were recoded (item 6,11,12,15,16,29,30,33,34 &35) and categories were created (for the question about the field of study of the respondents). This section provides insight in the extreme values and outliers, the items that were used for each variable, the factor analysis and the Cronbach‟s Alpha values for each scale (factor analysis is presented in appendix D).

3.2.1 Extreme values and outliers

Since extreme values in the data set might result in drawing incorrect conclusions, the data was analyzed using box plots. The box plots showed a few outliers which do not cause any problems. However, one extreme value (respondent 54) was identified within the box plot of readiness for change. In order to prevent drawing incorrect conclusions, the decision was made to exclude the answers of respondent 54 from the data set.

3.2.2 Measurements (factor analysis & reliability)

The factor analysis is presented in appendix D. In general, all items loaded quite well on the different factors and as such, the items indeed measure the variables presented in the theory section. In the section below, each heading concerns a variable. Within each heading, the (double) loadings within the factor analysis are explained and furthermore, it is explained why some items are included in the scale whereas one item was removed. Table 1 below represents the number of items for each scale and the Cronbach´s alpha values.

TABLE 1 Reliability Analysis

Variable Amount of Items Cronbach‟s Alpha Value

Readiness for change 4 0.80

Appropriateness 9 0.87

Self- and computer efficacy 7 0.87

Management support 11 0.92

Personal valence 3 0.79

Organizational identification with the new organization

5 0.79

Organizational identification with the old organization

3 0.78

(25)

3.2.3 Dependent variable: Readiness for change

The items to measure readiness for change were derived from the research of

Metselaar (2007). The items include statements such as; “I am willing to overcome possible resistance towards the change” and “I am willing to make time to implement the change” and are tested using a seven-point Likert scale. The factor analysis showed that the first item of change readiness loaded on a different factor. Quite likely, this is caused by the formulation (wording) of item one which resembles the independent factor appropriateness. Also, the double loading for item two can be explained using a similar reasoning. The content

(wording) of item two is linked to appropriateness which is likely to cause the item to load on two factors. Removing item one would result in only a slight increase in reliability whereas the removal of item two would result is a significant decrease in reliability. Besides, the content of the questions seems to be of particular importance for the level of change

readiness. For theoretical reasons, the choice has been made to enclose all items in the scale for readiness for change. As such, the scale for readiness for change consists of four items with a Cronbach‟s Alpha value of 0.80.

3.2.4 Independent variables: Appropriateness, efficacy, management support and personal valence

As mentioned before, the four components of the change message were derived from the research of Holt et al. (2007). All items of the independent variables were tested using a seven-point Likert scale.

Appropriateness

(26)

appropriateness. In total, the scale of appropriateness consisted of nine items with a Cronbach´s alpha value of 0,87.

Efficacy

All items of the efficacy scale have been derived from the research of Holt et al. (2007), translated in Dutch and adjusted to this research within the municipality of SWF. Two items have been added by me to represent computer-efficacy. These items are formulated in a similar way as the self-efficacy items but instead of „skills‟ these items referred to „computer skills‟, and instead of „tasks‟ they stated „computer-related tasks‟ (Question 15 and 18). Examples of items are; “I do not anticipate any problems adjusting to the work I will have when Corsa is implemented” and “I have the (computer) skills that are needed to make this change work”. Item one and four (Question 14 and 17) showed double loadings due to unclear reasons (see appendix D). However, removing these items would results in a slight decrease in reliability. The research of Holt et al. (2007) does include these items in the scale for efficacy and therefore, these items were also included in the scale for efficacy. The reliability analysis did result in the omission of item eight (Question 21) since the Cronbach Alpha value for self- & computer efficacy could be increased from 0,86 to 0,87 and more importantly, by removing this item, several loadings on factors were improved. As such, the Cronbach‟s alpha value for this scale is calculated at 0,87 with a total of seven items.

Management support

Similar to the other independent variables, the items for management support are derived from the research of Holt et al. (2007). All items are presented twice in the

questionnaire, one item relates to the direct manager of the employees whereas the other item is linked to the top management (directie). This choice was made based on the research of Cinite, Duxbury & Higgins (2009) mentioned earlier, who distinguished top management support from immediate manager support. For example, question 22 states; “The top

management has encouraged me to approach the implementation of Corsa in a positive way” and question 23 states; “My direct manager has encouraged me to approach the

(27)

items in this scale was 11 and internal consistency was high with a Cronbach‟s alpha value of 0,92.

Personal valence

Also the items for the final change message compone nt, personal valence, have been derived from the research of Holt et al. (2007). Besides translation, no adjustments have been made to these items. The items include statements such as; “I am worried I will lose some of my status in the organization when Corsa is implemented” and “My future in this job will be limited because of the implementation of Corsa”. With a total of three items, the reliability analysis showed a Cronbach‟s alpha value of 0,799.

3.2.5 Moderator variable: Organizational identification

Riketta (2005) argues that the Mael scale is most appropriate to measure the concept of organizational identification and therefore, items of the scale Mael, & Tetrick (1992) were used to test this concept. Additionally, I adjusted three items of their scale to represent the pre-merger identification levels since the distinction between pre- and post merger

identification is proven to be of considerable importance (Bartels, Pruyn, de Jong, 2009; Gleibs, Mummendey, Noack, 2008; van Dick, Wagner, Lemmer, 2004). To clarify, question 36 states; “When somebody criticizes the municipality of SWF, it feels like a personal insult” (item related to the „new‟ organization) and question 38 states; “When somebody criticizes the municipality where I previously worked, it feels like a personal insult” (item related to the „old‟ organization). As such, the moderator organizational identification has two types of questions; „organizational identification with the new organization‟ and „organizational identification with the new organization‟. A factor analysis forced into seven factors

(Appendix D) confirms the two separate aspects of organizational identification; identification with the new and with the old organization. Therefore, two separate scales could be used. Within the scale for organization identification with the new organization, item 5 (Question 40) showed a double loading. Why this double loading occurred is not clear. However, removal of item 5 (Question 40) would have resulted in a significant decrease in reliability and therefore, the choice was made to include this item in the scale for organizational

(28)

3.2.6 Control variables

Several control variables were included in the questionnaire; age, gender, nationa lity, education, current function and whether respondents attended an introductory session about Corsa. Since all respondents were Dutch the control variable nationality was not used in further analysis. Furthermore, whether respondents attended an introductory session about Corsa was not associated with any other variable and therefore, this variable was not presented in the correlation analysis and controlled for in the regression analysis. Three dummy variables were created for the control variable; current function. Only two out of three dummy variables were associated with other variables and therefore, the dummy variables were combined into one dummy variable (0= Ondersteunend/ uitvoerend/ facilitair, 1= Beleidsmedewerker/ Beleidsadviseur). Succeeding the measurements section, the following section describes the data analyses that were performed to test the hypotheses.

3.3 Data Analysis

The data was analyzed in two steps in order to test the hypotheses. These steps are described below.

3.3.1 Descriptive statistics and correlations

Firstly, the mean and standard deviation for each variable were calculated. Since the items of the questionnaire were tested using a seven- point Likert scale, these descriptive statistics help to create insight in the overall direction (either positive or nega tive) of the answers given by the respondents. Following the descriptive statistics, a correlation analysis was performed using the Spearman‟s Rho correlation coefficients since this test is appropriate for ordinal data.

3.3.2 Regression analysis

The second step concerns regression analyses. These analyses were used to test the relationships between the variables. According to de Heus, van der Leeden & Gazendam (1995) there are four assumptions that need to be controlled for before performing regression analysis;

(29)

scales of the variables (de Heus, van der Leeden & Gazendam, 1995). Therefore, this assumption can be verified for this specific research.

2. The second assumption concerns linearity of the variables, which was tested using residual plots. All plots showed a significant amount of linearity which verifies the second assumption.

3. The third assumption is related to randomness. All independent variables should be fixed (selected by the researcher) whereas the dependent variable should be random. Even though this assumption is mentioned, it is possible to conduct regression analysis without verifying this assumption. As de Heus, van der Leeden & Gazendam (1995) mention; this assumption is almost never verified when surveys are used, and as such, this also account for this specific research.

4. The final assumption is linked to a) independence, b) normality and c)

homoscedasticity of „error terms‟ (residues). The latter two concepts are tested using residual and q-q plots. The independency of the so called „error-terms‟ (residues) was accounted for by the procedure of data collection in this research. The residual plots did not show signs of violating the assumption about homoscedasticity. The q-q plots for all variables indicated normality and thus, the final assumption was also satisfied.

After the assumptions were verified, the actual regression analyses were performed. Firstly, a simple regression analysis was performed to create insight in the relationship

between an independent variable and the dependent variable. Control variables were included but did not show significant relationships with readiness for change. Thereafter, multiple regression was used to test the moderating influence of organizational identification with the new and old organization on the relationship between the independent and dependent

(30)

4. RESULTS

As mentioned earlier, the data was analyzed in two steps. Firstly, the descriptive statistics and the results of the correlation analysis are discussed. Thereafter, the outcomes of the regression tests are presented.

4.1 Descriptive Statistics and Correlations

The mean and standard deviation were calculated for each variable included in the conceptual model (see page 23). Following the descriptive statistics, a correlation analysis using the Spearman‟s Rho correlation coefficient was performed. The outcomes are presented in the table below.

TABLE 2

Descriptive Statistics & Correlations

Notes.** p <.01, * p< .05, † p<.10. ª Dummy coded variable (0= Ondersteunend/ uitvoerend/ facilitair, 1= Beleidsmedewerker/ Beleidsadviseur)

M SD 1 2 3 4 5 6 7 8 9 10 11

1. Gender 1.58 0.49 -

2. Age 3.7 2 0.98 -.26** -

3. Highest education (completed) 5.78 0.88 -.00 -.15 -

4. Current functionª 0.49 0.50 .01 -.13 .56** -

5. Personal Valence 5.80 0.80 .01 -.09 .16† .16† -

6. Management Support 4.44 1.10 -.01 .14 -.14 -.19* .18* -

7. Self- and computer efficacy 5.40 0.92 .12 -.19* -.05 -.07 .33** .18* -

8. Appropriateness 5.35 0.88 .20* -.06 -.05 -.09 .25** .31** .47** - 9.Organizational Identification with

the new organization

4.42 1.01 -.02 .02 -.14 -.00 .17† .19* .01 .13 -

10.Organizational Identification with the old organization

4.11 1.09 .03 .00 -.21* -.07 -.08 .10 .02 .00 .40** -

(31)

Since the items in the questionnaire were measured using a seven point Likert scale, the mean score for personal valence (M=5.80) can be considered quite high. These statistics indicate that people see the personal advantages of the change. Organizational identification with the old organization scores relatively low compared to the other variables (M=4.11).

As expected, based on the research of Holt et al. (2007), all relationships between the independent and dependent variables shows significant correlations (p<.01).

The highest correlation can be found between appropriateness and readiness for change (r = .60, p< .01), whereas the lowest correlation coefficient can be found for the relationship between management support and readiness for change (r=. 33, p<.01) and personal valence and readiness for change(r=. 33, p<.01). Also, all the relationships among the independent variables (appropriateness, self- and computer efficacy, management support and personal valence) are significant (p<.01 & p<.05). Appropriateness and self- and computer efficacy show the strongest correlation (r= .47, p<.01).

The correlations show several striking issues. Firstly, appropriateness and gender are significantly positively related (r=.20, p<.05). This indicates that there is a distinct difference between man and women in how they feel about this change effort in te rms of

appropriateness. Secondly, self- and computer efficacy and age are significantly negatively correlated (r=-.19, p<.05), which indicates that as the age of respondents increases, their level of self- and computer efficacy decreases. Thirdly, organizational identification with the new organization is significantly positively correlated with management support(r=.19, p<.05), whereas organizational identification with the old organization is significantly negatively correlated with highest education completed (r =- . 21, p<.05). Also, the correlations indicate that current function and management support are significantly negatively correlated (r =- . 19, p<.05). That is, when employees are in „higher‟ functions (beleidsadviseur/

beleidsmedewerker) this could be associated with perceiving less management support compared to employees in supporting functions.

Other notable issues include a marginal positive correlation between personal valence and highest education completed (r=.16, p<.10), which suggest that as the age of a respondent increases, their level of personal valence also increases. Also, personal valence and current function are marginally positively related (r=.16, p<.10). At the same time, personal valence is marginally positively correlated with organizational identification with the new

(32)

with the new organization is marginally positively correlated with readiness for change (r= .17, p <.10) which indicates that as one concept increases, the other also increases.

4.2 Regression Analysis

As mentioned in the methodology section, the regression analyses were performed in two steps. First, the simple linear and multiple regression analysis outcomes concerning the linear relationships are discussed where after the multiple regression analysis outcomes concerning the hypothesized moderations are presented. Finally, some additional analyses that were executed are presented.

4.2.1 Regression analysis linear relationships

Hypothesis one proposes that a high level of perceived appropriateness is associated with a higher level of change readiness of employees within an organization. This hypothesis can be confirmed, appropriateness and readiness for change are significantly positively related (ß= .63, p <.001, R² = .40). As such, when the perceived level of appropriateness of an employee is high their level of change readiness is also higher.

Hypothesis two was also supported by regression analysis. This hypothesis states that a high level of perceived self- and computer efficacy is associated with a higher level of change readiness of employees within an organization. Regression analysis confirmed that self- and computer efficacy are significantly positively related to change readiness (ß= .32 p <.001, R² = .10). That is, when the level of self- and computer efficacy of employees

increases, their level of change readiness does as well.

In hypothesis three, it was proposed that a high level of perceived immediate and senior management support would be associated with a higher level of change readiness of employees within an organization. The regression analysis showed that there is a significant positive relationship between management support and readiness for change (ß = .33 p <.001,

R² = .11). Therefore, this hypothesis was confirmed. This indicates that when an employee

has a high level of perceived management, they also feel more ready for change.

(33)

4.2.4 Multiple regression analysis moderation

Following the simple regression analysis, this section focuses on the outcomes of the multiple regression analyses that were performed to test the proposed moderating influence of organizational identification with the new and with the old organization (Hypothesis five until eight). Outcomes of this analysis can be found in appendix E.

Hypothesized was that a high level of organizational identification with the new organization would positively moderate the relationship between the perceived level of appropriateness and the level of change readiness (H5a). That is, when the level of organizational identification with the new organization is high, the positive relationship between the perceived level of appropriateness and the level of change readiness will be reinforced. Based on the outcomes of the regression analysis this hypothesis can be rejected (β = -.13, p= n.s, see appendix E, table 5). Also, no support is found for hypothesis 5b. This hypothesis states that the relationship between the perceived level of appropriateness and the level of change readiness will be negatively moderated by a high level of organizational identification with the old organization. In this respect, the positive relationship between the perceived level of appropriateness and the level of change readiness will be weakened when organizational identification with the old organization is high. However, regression analysis did not show a significant interaction (β = .10, p= n.s, see appendix E, table 6).

Hypothesis 6a proposed that the relationship between the perceived level of self~ and computer efficacy and the level of change readiness will be positively moderated by a high level of organizational identification with the new organization. As such, the positive relationship between the perceived level of self~ and computer efficacy and the level of change readiness will be reinforced when organizational identification with the old

organization is high. This hypothesis was rejected (β = .09, p= n.s, see appendix E, table 5). Hypothesis 6b was rejected as well. This hypothesis stated that the relationship between the perceived level of self~ and computer efficacy and the level of change readiness will be negatively moderated by a high level of organizational identification with the old

organization. So to say, a high level of organizational identification with the old organization will weaken the positive relationship between the perceived level of self~ and computer efficacy and the level of change readiness. This was not confirmed by the outcomes of regression analysis (β = -.01, p= n.s, see appendix E, table 6).

(34)

senior management support and the level of change readiness will be positively moderated by a high level of organizational identification with the new organization. In this respect, the positive relationship between the perceived level of immediate and senior management support and the level of change readiness will be reinforced when organizational

identification with the new organization is high. In hypothesis 7b it was proposed that the relationship between the perceived level of immediate and senior management support and the level of change readiness will be negatively moderated by a high level of organizational identification with the old organization. In this case, a high level of organizational

identification with the old organization would weaken the positive relationship between the perceived level of immediate and senior management support and the level of change readiness. Both hypothesis 7a (β = .12, p= n.s, see appendix E, table 5) and hypothesis 7b (β =.15, p= n.s, see appendix E, table 6) can‟t be confirmed.

Finally, hypothesis 8a en 8b were tested. Hypothesis 8a states that the relationship between the perceived level of personal and organizational valence and the level of change readiness will be positively moderated by a high level of organizational identification with the new organization. In this respect, the positive relationship between the perceived level of personal and organizational valence and the level of change readiness will be reinforced when organizational identification with the new organization is high. Hypothesis 8b proposed that a high level of organizational identification with the old organization would negatively

moderate the relationship between the perceived level of personal and organizational valence and the level of change readiness. That is, the positive relationship between the perceived level of personal and organizational valence and the level of change readiness will be weakened when organizational identification with the old organization is high. No support was found for both hypothesis 8a (β = -.07, p= n.s, see appendix E, table 5) and 8b (β = -.09, p= n.s, see appendix E, table 6). To conclude, the results do not show that organizational identification with either the new or the old organization has an interaction effect, as

hypothesized, on the relationships between the independent variables (appropriateness, self- and computer efficacy, management support and personal valance) and the dependent variable.

4.2.2 Additional regression tests

(35)

The correlation analysis showed that organizational identification and readiness for change were marginally positively correlated (r= .17, p <.10). Based on these outcomes, the relationship between organizational identification with the new organization and change readiness was also tested. Surprisingly, organizational identification with the new

organization is significantly positively related with readiness for change (ß= .26 p <.01, R² = .07). That is, a high level of organizational identification with the new organization is associated with a higher level of readiness for change. When controlling for gender, age, highest education completed and current function, the significant positive relationship becomes slight stronger (ß= .27 p <.01, R² = .08)

In addition to the simple regression ana lyses, multiple regression analysis was performed to analyze whether certain independent variables had a unique contribution to readiness for change. The results can be found in table 3 below.

TABLE 3

Multiple Regression Analysis

Notes. ***p<.001,**p<0.01,*p<.05. Dependent variable = Change readiness.

ª Dummy coded variable (0= Ondersteunend/ uitvoerend/ facilitair, 1= Beleidsmedewerker/ Beleidsadviseur).

This hierarchical multiple regression analysis in which I controlled for gender, age, highest education completed and current function shows that both appropriateness and personal valence make a unique contribution in predicting readiness for change. The regression coefficients for management support and self- and computer efficacy were not

Model 1 Model 2 B ß B ß Gender .14 .08 -.03 -.02 Age .05 .06 .08 .09 Highest education - .05 - .05 -.06 -.06 Current functionª - .02 - .01 .08 .04 Appropriateness .53 .55***

Self- and computer