Thesis
University of Amsterdam
The relationship between ambivalence and commitment to change and behavioral support for change moderated by leader-member exchange (LMX)
Lennart Dijkstra (10884084) University of Amsterdam
Executive Programme in Management Studies Specialization: Leadership and Management Academic year: 2016-2017
Supervisor: dr. M. Venus
2
Statement of Originality
This document is written by Lennart Dijkstra who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Table of Contents
1. Introduction ... 5
2. Literature review ... 8
2.1 Importance for organizations to change ... 8
2.2 Commitment to change ... 9
2.3 Behavioral support for change ... 11
2.4 Resistance to change ... 13 2.5 Ambivalence ... 14 2.6 Leader-Member Exchange (LMX) ... 18 2.7 Conceptual model ... 22 3. Method ... 23 3.1 Measures ... 24 4. Results ... 27 4.1 Exploratory research ... 30 5. Discussion ... 33 5.1 Theoretical implications ... 34
5.2 Limitations and future research ... 38
5.3 Practical implications ... 42
6. Conclusion ... 44
4
Abstract
This study examines the relationship between ambivalence and both commitment to change and behavioral support for change, and the influence leader-member exchange (LMX) has on this relationship. This study proposes six hypotheses about the variables ambivalence,
commitment to change, behavioral support for change and LMX. The results were measured with the use of two different questionnaires for both employees and managers. The results indicated that ambivalence significantly negatively relates to behavioral support for change, which indicates that ambivalence reduces employee behavioral support for change in the perception of the manager. Furthermore, implications for practitioners and future research are offered.
5
1. Introduction
Organizations struggle to cope with technological developments, changing consumer demands, economic and legislative changes, globalization, and the denationalization and deregulation of marketplaces, which have resulted in a highly complex environment (De Meuse, Marks & Dai, 2010; Gordon, Stewart, Sweo & Luker, 2000). Therefore, organizations have to constantly change in order to keep pace with competitors and to meet changing
market demands (Rafferty et al, 2012).
According to Oreg (2003) organizations that attempt to initiate change are often hindered by individuals or groups within the organization who resist the initiated change. In the research and practitioner literature on organizational change the term resistance to change is regularly used. This term often explains why organizational efforts to introduce large-scale changes, production methods, management practices, or compensation systems do not meet
expectations and do not succeed (Oreg, 2006). Many studies have been conducted emphasizing resistance to change. For instance, Dent and Goldberg (1999) stated that
resistance to change misrepresents what really happens in the change dynamic and suggested to abandon the term as organizational members resist negative consequences of change and not necessarily change itself. Furthermore, the majority of work has focused on the various forces that encourage employees to not support changes proposed by managers.
When focusing on resistance to change, Piderit (2000) suggests that it may involve a sense of
ambivalence whereby employees’ feelings, behaviors, and thoughts about the change do not
correspond with each other. Despite the fact ambivalence is often experienced as unpleasant it has several meaningful consequences for both individual’s experiences and their perceptions and decisions. People holding ambivalent attitudes tend to process information more
systematically (Sverdlik & Oreg, 2011) and could be effective change agents who provide a more balanced and realistic perspective of the opportunities and threats the specific change is
6
offering. By acknowledging ambivalence organizations could use it to improve itself and organizational decisions (Meyerson & Scully, 1995).
In previous research, Oreg and Sverdlik (2009) linked ambivalence to type of change, and argued that imposed change should yield greater ambivalence than voluntary change. Furthermore, Oreg (2006) analyzed the relationship between cognitive resistance and continuance commitment to change. They found substantiated evidence for a negative
relationship between the variables. Furthermore, to gain a more realistic understanding of how employees respond to organizational change the aim of Oreg and Sverdlik (2011) was to acknowledge, assess and systematically predict the existence of ambivalence. Not before, researchers studied ambivalence in relation to both commitment to change and behavioral support for change. This is unfortunate, because to gain understanding of how manager’s perceive ambivalence it is valuable to know what effect ambivalence which is often
experienced as unpleasant while it holds several meaningful consequences (Sverdlik & Oreg, 2011) has on commitment to change and behavioral support for change which are both important to make the change initiation successful (Herscovitsch & Meyer, 2002).
To further investigate ambivalence, this study will yield additional insights in this research area by examining the effect this phenomenon has on employees’ commitment to change and behavioral support for change and the moderating effect of leader-member exchange (LMX) on this relationship. LMX refers to the quality of the relationship between employees and managers (Eisenberger et al, 2010). It is anticipated that the direct relationship between ambivalence and both commitment to change and behavioral support for change is negative and that this relationship depends on the quality of the relationship between the manager and the employee. It is anticipated that in a high quality relationship ambivalence will be
7
quality relationship ambivalence will be more negatively related to both commitment to change and behavioral support for change.
Therefore, the results of this research will be useful for both academics and management professionals as this study will contribute to the existing knowledge in the area of
organizational change and gives additional insights into the described relationships. For
management professionals this study will give insights in the effect their relationship with the employee influences has on the relationship between employee ambivalence and both
commitment to change and behavioral support for change. This paper aims to prove that employee ambivalent attitudes can positively relate to employees’ commitment to change and behavioral support for change when the quality of the relationship between the employee and the manager is good in the perception of the manager, by answering the following question:
How does LMX influences the relationship between ambivalence and both commitment to change and behavioral support for change?
8
2.
Literature review
In an attempt to answer the stated research question the applicable variables have to be closely examined. The literature review starts with a description of the importance of organizational change, followed by a description of commitment to change and behavioral support for change. Subsequently, resistance to change and ambivalence are described, ending with the first hypothesis. Finally, the concept of leader-member exchange (LMX) and the effect this concept has on the relationship between ambivalence and both commitment to change and behavioral support for change is described, ending with the second and third hypothesis. This chapter ends with the conceptual model, in which the stated hypotheses are graphically illustrated.
2.1 Importance for organizations to change
First of all, it is important to get an understanding why it is important for organizations to change. Therefore, in this section the importance of change, and accordingly the importance of positive employee reactions to change will be discussed.
Technological developments, changing consumer demands, economic and legislative changes, globalization, and the denationalization and deregulation of marketplaces, have resulted in a highly complex environment. In general, organizations have difficulties to cope with this highly complex environment. (De Meuse, Marks & Dai, 2010; Gordon, Stewart, Sweo & Luker, 2000). Therefore, organizations have to constantly change in order to keep pace with competitors and to meet changing market demands (Rafferty et al, 2012). According to Entin et al (2003), change has become of eminent value for the existence of the organization.
Moreover, D’Aveni (1994), state that the ability to cope with this often dramatically changing environment and to meet the changing demands and expectations is a key determinant of competitive advantage and organizational survival. In order to cope with these developments, organizations need to be flexible, adaptive, and being capable of rapidly changing themselves
9
(Gordon et al, 2000) which results in a situation in which organizations at least every four to five years make moderate to major changes (Lewis, 2000). According to Entin et al (2003) it is generally accepted that organizations that are capable of adapting to internal and external changes have a much better chance to be successful. Indeed, it has been demonstrated that organizations that perform well are adept in recognizing changes in their environment and altering their strategies to accommodate these changes without comprising performance.
Accordingly, positive employee reactions to change are important for successful organizational change. When discussing change, positive attitudes towards change are
associated with successful change implementation and thus improved outcomes (Oreg, 2006). Furthermore, Stensaker and Meyer (2011) state that exposure to organizational change can generate positive reactions to change. An example of a positive attitude toward change is commitment to change. Research has shown that employees’ commitment to change positively influences their behaviors supporting change. Organizations are frequently unsuccessful in motivating appropriate levels of employee commitment to change and therefore this phenomenon will be discussed in the next section (Hill et al, 2012).
2.2 Commitment to change
According to Meyer and Herscovitch (2001), commitment to change can be defined as a mind-set that binds an individual to a course of action considered necessary for the successful implementation of a change initiation. The mind-set that binds an individual to this course of action can reflect three things. First, affective commitment to change (ACC), which is an employee desire to provide support for the change based on a belief in its fundamental benefits. Second, continuance commitment to change (CCC), which is employee recognition that there are costs associated when not providing support for change. Third, normative
10
the change. Respectively, employees can feel bound to support a change because they want to, have to, and/or ought to. This study will focus on affective commitment to change.
Furthermore, gaining employees’ commitment to change is predominantly important during radical change, due to the involvement of a fundamental, qualitative change in the
organization’s philosophy and strategic orientation (Greenwood & Hinings, 1996, Nadler & Tushman, 1995). Without employee support to change throughout the organization, radical change efforts are not likely to succeed (Herscovitsch & Meyer, 2002, Romanelli & Tushman, 1994). In conclusion, the fact that many radical change implementations fail to succeed
underscore the fact that organizations are frequently unsuccessful in committing employees to change at a required level (Hill et al, 2012).
Prior research to commitment to change conducted by Meyer and Herscovitch (2002) identified that ACC and NCC influences employee behaviors in support of change. They found substantiated evidence that ACC and NCC promotes important employee behaviors that actively support change, including cooperating with the spirit of the change, making personal sacrifices to support the change, and promoting it to others inside and outside the organization (Herold et al, 2007). Moreover, Herold et al (2007) state that besides commitment to change reflects positive attitudes toward the change, it also reflects alignment with the change, intentions to support it, and willingness to make the change implementation successful. This view of a positive, proactive behavioral intent toward initiated change makes commitment different from other attitudinal constructs that focus on negative attitudes, such as resistance to change (Kotter & Schlesinger, 1979; Piderit, 2000), or positive dispositions toward a change, such as readiness for change (Armenakis et al, 1993) or openness to change (Wanberg & Banas, 2000).
11
Besides commitment to change another positive attitude towards change is behavioral support for change. As stated before, commitment to change is a mind-set that binds an individual to a course of action considered necessary for the successful implementation of a change
initiation. When behaviorally supporting the change an employee expresses support for the change with his/her behavior (Meyer et al, 2007). Despite the fact that commitment to change has been shown to be an important determinant of behavioral support for change
(Herscovitsch & Meyer, 2002) both variables will not be investigated in relation to each other. In the next section I will elaborate on behavioral support for change.
2.3 Behavioral support for change
According to Meyer et al (2007) behavioral support for change consists of compliance and two forms of discretionary support which are cooperation and championing. Compliance taps employees’ willingness to do what is required of them by the organization in the
implementation of the change. Furthermore, cooperation assesses employees ‘acceptance of the spirit’ of the change and willingness to do something extra to make it work. Finally, championing addresses employees’ willingness to embrace the change and to ‘sell’ it to others.
According to Oreg (2003) organizations value employees who are willing and able to initiate and respond positively to change. However, organizations that attempt to initiate change are often hindered by individuals or groups within the organization who resist the initiated change. According to Piderit (2000) resistance to change has a negative connotation and can be conceptualized in three different emphases: as a cognitive state, as an emotional state and as a behavior. These three emphases in the conceptualization of resistance to change can be framed in an integrative approach by using the concept of attitude from social psychology. Piderit (2000) labels these three dimensions of attitudes as cognitive, emotional and intentional, which is a conceptualization also known as the tripartite view of attitudes.
12
Using the multidimensional definition to describe employees’ attitudes toward proposed changes is beneficial because conceptualizing each dimension as a separate continuum makes different reactions along the different dimensions a possibility. Therefore, this definition recognizes the possibility of ambivalent attitudes, where two alternative perspectives are both strongly experienced in the opposite direction (Piderit, 2000).
According to Piderit (2000), in this conceptualization of resistance to change, the cognitive dimension of an attitude refers to an individual’s beliefs about the proposed change. These beliefs are positive or negative. Second, the emotional dimension of an attitude refers to an individual’s feelings in response to the proposed change. This dimension can be defined as feelings, moods, emotions, and sympathetic nervous-system activity that people have
experienced in relation to the proposed change and subsequently associate with it. Lastly, the intentional dimension of an attitude refers to an individual’s plan or determination to take some action, rather than try to achieve some goal.
While focusing on commitment to change and behavioral support for change it is due to the fact that organizations are often hindered by individuals or groups within the organization who resist the initiated change (Oreg,2003) and the possibility of ambivalent attitudes across dimensions (Piderit,2000) important to get an understanding of employee resistance to change and predominantly the ambivalent attitudes. For instance, Oreg and Sverdlik (2011) state that if manager’s don’t see ambivalence as a possibility, it will be often misinterpreted and
perceived as indifferent, while it actually holds several positive attributes to contribute to the change situation. This is the main reason to further investigate ambivalence.
Furthermore, this makes it interesting how manager’s perceive employee commitment to change and behavioral support for change when they show ambivalent attitudes. In the next sections I will elaborate on resistance to change and ambivalence.
13 2.4 Resistance to change
Dent and Goldberg (1999) state that resistance to change is a pervasive mental model in terms of how recipients react to change. Resistance to change is rooted in employees, and managers must do the right things to overcome this resistance, because it interferes with successful change implementation. The belief that people resist change causes unproductive actions within organizations. Employees do not resist change, they resist the unknown, being dictated to, or resist management ideas that do not seem feasible for them. Employees understanding of the new vision and the motivation to make it happen are often present, however there are obstacles that prevent execution. These obstacles are often found in the organization’s structure and not in the individual.
Piderit (2000) elaborates on this subject by stating that despite the fact resistance to change is widely studied it has its limitations. The limitations of the concept can be framed in
philosophical and practical terms. Philosophical for instance, critical theorists argue that the interests of managers should not be more important than the interests of employees when studying organizational change. In practical terms for instance, practical scholars and
scholarly practitioners argue that the concept might have outlived its usefulness. According to Ford et al (2008), the predominant perspective from which resistance to change is primarily investigated is decidedly one sided in favor of the change agent and their sponsors. Previous studies investigating change appear to take the perspective, or bias, that during change initiatives it is presumed that change agents are doing the right things while recipients of change act as resistors, throwing up unreasonable obstacles or barriers towards change. Accordingly, change agents are portrayed as victims of the irrational and not cooperative response of change recipients. In this “change agent-centric” view it is presumed that resistance to change is an accurate report by an unbiased observer, which is in this case the change agent, of an objective reality, which is the recipient’s resistance.
14
Therefore, Ford et al (2008) disagree with Dent and Goldberg (1999) that resistance is rooted in recipients. Actually, change agents themselves are barriers for successful change. They are not victims of the irrational and dysfunctional responses of change recipients and not unbiased observers who deliver an accurate report of an objective reality. Moreover, Ford et al (2008) state that resistance is negatively interpreted by change agents because they want to see resistance in recipient’s behaviors and communications and this acts as a self-fulfilling prophecy. The possibility they contribute to the occurrence of recipient’s resistant behavior and communication, due to their own actions or inactions, is not taken in consideration by earlier views on resistance. Furthermore, Ford et al (2008) contribute the resistance of recipients to trust violations, communication breakdowns and by resisting resistance of change agents. In fact, Ford et al (2008) interpret resistance as a resource. They state that it has an existence value because people are talking about the change, an engagement value because people who show negative reactions are engaged and a strengthening value because of the value of conflict in which both parties can come out stronger.
2.5 Ambivalence
Sverdlik and Oreg (2011), suggest that employees’ reactions to change can be more complex than has been considered in previous research. As stated before, if ambivalence is not
considered as a possibility, this often leads to the misinterpretation of employees’ which can be considered as indifferent. For instance, according to Larson and Tompkins, (2005) ambivalence was shown to contribute significantly to employees’ resistance to change. Individuals perceived as indifferent could have actually hold strong, yet conflicting views about the change. Sverdlik and Oreg (2011) define ambivalence as an attitude comprising both positive and negative reactions along the different dimensions. Ambivalence will increase if these reactions are getting stronger.
15
Despite the fact ambivalence is often experienced as unpleasant it has several meaningful consequences for both individual’s experiences and their perceptions and decisions. People holding ambivalent attitudes tend to process information more systematically, in a more controlled and reflective, rather than automatic manner and tend to provide descriptions that are more balanced and accurate than those made by individuals not experiencing ambivalence. Furthermore, according to Sverdlik and Oreg (2011), ambivalent employees need particular support and guidance during the change initiation, to alleviate the discomfort ambivalence creates. Next to that, ambivalent employees can provide a valuable perspective about the change that could assist manager’s in both design and implementation. In detail, manager´s not seeing ambivalence as a possibility, while it holds several positive attributes, leads to misinterpretation of ambivalence.
Though this study is focusing on ambivalence across dimensions, ambivalence can also occur within a dimension. Firstly, an employee exhibiting emotional ambivalence could respond with fear and excitement towards a proposed change. Secondly, with regard to the cognitive dimension it is possible that an employee simultaneously believes that the proposed change is needed in his or her organization to survive in the future, only need to be further researched. Lastly, an employee exhibiting intentional ambivalence might anonymously plan to be against a proposed change while in public supporting it. This could be due to uncertainty how top-management respond to criticism addressed to the change initiative (Piderit, 2000).
Ambivalence across dimensions is for instance the case when an individual’s cognitive response to a proposed change does not match with his or her emotional response to the proposal (Piderit, 2000). An example of ambivalence across dimensions is when an employee shows positive reactions along the cognitive dimension, because he/she is aware of the fact that this change is essential for the organization to survive for instance bad financial times. On
16
the other hand this employee is showing negative reactions along the emotional dimension, because he/she is afraid that the change will result in a layoff.
According to Oreg (2006) the view of Piderit (2000) is likely to capture the complexity of resistance phenomenon and may provide a better understanding of the relationship between resistance and its antecedents and consequences. In this view some sources of resistance have a strong impact on employees’ emotions, others may influence their behaviors, and yet others may influence what employees think about the initiated change. Furthermore, Meyerson and Scully (1995) state that employees who experience ambivalence as in identifying conflicting thoughts across dimensions within the organization are perceived as effective change agents because they provide a more balanced and realistic perspective of the opportunities and threats the specific change is offering. Therefore, ambivalent employees act as constructive critics entertaining the possibility of both maintaining the status quo and implementing change in the organization. Moreover, the potential advantages of encouraging employee ambivalence towards proposed organizational change are addressed by Piderit (2000), who stated that ambivalent reactions towards change need to be acknowledged. If doing so, organizations could use it to improve itself and their decisions. Next to that, manager´s may find it easier to bring forth new ideas and solutions for dealing with the situation that activated the change. As stated before ambivalent employees have a balanced and realistic perspective of the opportunities and threats the specific change is offering. Therefore, they could be a valuable resource for manager´s to generate ideas and solutions from.
In detail, as described above, according to multiple authors ambivalence holds several positive attributions which can contribute to the change. However, when managers don’t see ambivalence as a possibility, they could misinterpret employee reactions and consider them as indifferent. In contrast to this latter, manager’s interpret employees who are committed to change to have a mind-set that binds them to their organizational course of action considered
17
necessary for the successful implementation of a change initiation and not to be indifferent (Meyer and Herscovitsch, 2001).
Furthermore, managers perceive employees who behaviorally support the change as follows: first, employees do what is required of them by the organization in the implementation of the change. Second, they show the willingness to do something extra to make it work and finally they embrace the change and ‘sell’ it to others (Meyer et al, 2007). This perception does not match with manager´s interpretation of employee ambivalence which they often misinterpret and perceive as indifferent. Consequently, manager’s misinterpretation of employee
ambivalence (not seeing it as a positive reaction to change) is opposite to what they perceive as behavior which support the change.
In detail, manager’s interpretation of employee ambivalence which they could misinterpret and consider as indifferent does not match with their perception of people who are committed to change and behaviorally support the change. This could actually lead to employee behavior which is not committed to and supporting the change. The manager will for instance, publicly disapprove the employees behavior or ignore them. Consequently, the employee finds himself excluded from the change process which will lead to employee behavior which is not
committed to the change and behaviorally supporting the change.
Based on the negative perception the manager has of employee ambivalence, the perceived mismatch with positive reactions and the negative employee behavior this interpretation could cause I have reason to believe that ambivalent employee reactions to change significantly reduces employee commitment to change and behavioral support for change in the perception of the manager. Therefore, I hypothesize:
H1a: Ambivalence is negatively related to commitment to change.
18 2.6 Leader-Member Exchange (LMX)
In the previous section I have discussed the negative and positive sides of ambivalence. As stated before, ambivalence is generally negatively interpreted by manager’s while it actually holds several positive attributes to contribute to the change. I argue that the way manager’s interpret ambivalence depends on how they value their own relationship with the employee.
Therefore, I will elaborate on the effect the quality of the relationship between the manager and the employee has on the relationship between ambivalence and employee commitment to change and behavioral support for change perceived by the manager. I use leader-member exchange (LMX) as a moderator and predict that when LMX is high the relationship between ambivalence and commitment to change and behavioral support for change is positive, whereas the relationship between ambivalence and commitment to change and behavioral support for change is negative when LMX is low. When the relationship between manager and employee is good in the perception of the manager, ambivalence will be interpreted as a sign of engagement and maybe as useful feedback. In this situation it will have a positive relation with commitment to change and behavioral support for change. This is supported by Ford et al (2008) who state that negative reactions can be positively related to engagement. Ambivalence is one possible form of engagement with change and because it is thoughtful it reflects a high level of commitment. If employee show thoughtful attitudes like ambivalence they are likely to generate scrutiny and well-considered counterarguments which may be interpreted by the manager as highly committed and motivated over the duration of change when LMX is high.
Furthermore, due to the suggestion that manager’s interpret ambivalence positively when LMX is high, I believe the manager cares about the opinion of the employee and gives them the confidence to express him/her self. By noticing that the manager cares about their opinion, the employee will consider themselves as taken seriously and will have the feeling that their
19
opinion counts. As a result they will show more commitment and behavioral support. Consequentially manager’s label ambivalent attitudes as a sign of engagement.
On the contrary, when LMX is low the relationship between ambivalence and commitment to change and behavioral support for change will be more negative. When the relationship between the employee and manager is bad in the perception of the manager, ambivalence will be interpreted as indifferent which increases the negative relationship with commitment to change and behavioral support for change. I believe that when the employee and the manager have a bad relationship the manager does not give the employee the confidence to express him/her self and in such a relationship they do not care about the opinion of the employee and consequential label ambivalent attitudes as resistance to change. As a result, the employee will show less commitment and support for change.
To explain the interpretations of managers under different levels of LMX I will go into detail on the article of Furst and Cable (2008). They state that literature regarding change
management has explored various strategies that managers use to reduce employees resistance to change (e.g., Kotter & Schlesinger, 1979; Nutt, 1986). Unfortunately, it is still
undetermined how these strategies affect employee resistance to change. The fact that
employees may respond differently to managerial behaviors depending on how they interpret the intention of their manager is a possible explanation for the conflicting findings in the change literature. Furst and Cable (2008) refer to the attribution theory of Heider (1958) which states that observers try to make sense of an actor’s behavior by searching for a cause for the behavior. In an attempt to find a causal explanation, individuals rely on an array of signals to determine the underlying motives of the actor. Furthermore, these signals are also used to assess whether a behavior is caused by dispositional factors (i.e. caused by the actor) or situational factors (i.e. caused by context or situation).
20
Comparing an actor’s present behavior with past behavior is a useful way in which
individuals search for meaning. Actions which can be considered to be consistent with past behaviors typically are assigned a dispositional cause. Next to that, actions which can be considered to be inconsistent with past behaviors are assigned a situational cause. This is exemplified by the fact that studies show that individuals tend to accept feedback when it comes from a source they like. This is because they attribute good intentions to the source. However, when negative feedback comes from a source they don’t like, individuals tend to reject that feedback due to the fact they do not trust the source’s intentions (Fedor,1991).
The attribution theory suggests that an employees’ reaction to managerial influence attempts could be a reflection of the quality of the interpersonal relationship the employee and the manager have (Ferris & Judge, 1991; Porras & Robertson, 1992). In fact, LMX suggests that by an ongoing series of interpersonal exchanges employees develop unique relationships with their managers (Dansereau, Graen, & Haga, 1975). The expected behaviors of both parties are shaped by this relationship. Interactions in high-quality LMX relationships are based on loyalty, emotional support, mutual trust and liking (Dienesch & Liden, 1986; Graen & Scandura, 1987). On the other hand, interactions in low-quality LMX relationships are transactional and impersonal, grounded in contractual exchanges between the employee and the manager (Furst & Cable, 2008).
As stated before, Meyerson and Scully (1995) suggests that employees who experience ambivalence as in identifying conflicting thoughts across dimensions are perceived as
effective change agents due to the fact they provide a more balanced and realistic perspective of the opportunities and threats the specific change is offering. Therefore, ambivalent
employees act as constructive critics entertaining the possibility of both maintaining the status quo and implementing change in the organization. Moreover, Ford et al (2008) state that
21
recipient’s resistance reactions can be interpreted as valuable for the existence and recognition by them for change and their engagement, which can be considered beneficial for change.
I have reason to believe that the statements of Meyerson and Scully (1995) and Ford et al (2008) only have a positive effect on commitment to change and behavioral support for change when LMX is high. Namely, interactions in high-quality LMX relationships are characterized by loyalty, emotional support, mutual trust and liking. Ambivalent reactions to change will be appreciated due to this good understanding between the manager and the member, because the manager takes employee ambivalence seriously when LMX is high. As stated before, the employee notice this and consider themselves as taken seriously which leads to employee commitment and behavioral support for change.
On the other hand, these statements will not apply when LMX is low. Low LMX interactions between employee and manager are transactional and impersonal, grounded in contractual exchanges (Furst & Cable, 2008) and not based on an emotional relationship. Therefore, this could lead to negative managerial interpretation of employee ambivalence. I therefore hypothesize:
H2 a: It is expected that when LMX is high the relationship between ambivalence and
commitment to change is positive.
H2 b: It is expected that when LMX is low the relationship between ambivalence and
commitment to change will be more negative.
H3 a: It is expected that when LMX is high the relationship between ambivalence and
behavioral support for change is positive
H3 b: It is expected that when LMX is low the relationship between ambivalence and
22 2.7 Conceptual model
Figure 1 demonstrates the conceptual model of the relationship between ambivalence and commitment to change and behavioral support for change. Furthermore, the moderating effect of leader-member exchange (LMX) on the relationship between ambivalence and
commitment to change and behavioral support for change is depicted. Ambivalence is measured using employees self-report. Furthermore, commitment to change, behavioral support for change and leader-member exchange (LMX) are measured from the perspective of the manager.
Figure 1: Conceptual model
- + + - + Ambivalence Leader-member exchange (LMX)
Behavioral support for change
23
3. Method
Questionnaires were used to collect information about the influence of employee ambivalent attitudes on commitment to change and behavioral support for change. Both managers and their according employee where asked to fill out a questionnaire to measure the constructs which are studied. The dyadic relationship between the employees and their direct manager was assessed using two different questionnaires for both employees and managers. The employee questionnaire included measures of ambivalence. Furthermore the manager
questionnaire included measures of leader-member exchange (LMX), commitment to change and behavioral support for change. To make it clear which questionnaires formed a pair they were coded.
The questionnaires were physically distributed to different organizations who planned to initiate change. All organizations are established and operating in The Netherlands. Data was collected from the beginning of October till the end of October 2016. The distribution of the questionnaire was done personally by the author and by people in my network. Due to the fact the distribution was done personally the response rate is rather high, 83%. This also increased the response time which was on average one week for one organization. Because
questionnaires were formulated in Dutch it was a requirement that all participants could read and speak Dutch. Furthermore, all participants were assured that their individual answers would be kept confidential and would not be shared within the organization they work for. Therefore, the questionnaires were distributed via envelopes, so that participants could return their completed questionnaires in an envelope to guarantee anonymity.
In total, 91 pairs of questionnaires were collected from 91 employees and 14 supervisors of 9 different organizations. Organizations that participated in this study operated in different branches, including automotive, childcare, optician, seeds industry and flavor industry. The age of the employee questionnaire ranged from 19 till 64 . Of this group 52.7% were male and
24
47.3% were female. The education level was divided as 56.1% MBO, 40.6% HBO and 3.3% University degree. Employees had an average tenure of 5.1 years.
Of the supervisors, 54.9% were male and 45.1% were female, where 78% had daily contact with their employees. The education level was divided as 24.2% MBO, 70.3% HBO and 5.5% University degree. The age of the supervisors ranged from 23 till 55 and they had an average tenure of 8.9 years in a leading position. Type of changed varied from another and included, business expansion, business relocation, reorganization, loss of a major customer and a change in the management structure.
3.1 Measures
The employee questionnaire entailed the following content: Ambivalence, which consists of affective ambivalence (5 items) and cognitive ambivalence (5 items), leader-member exchange (8 items) and finally some demographics which asked for the age, gender,
educational level, tenure, and industry. Furthermore, the manager questionnaire entailed the following content: Leader-member exchange, (8 items), commitment to change (6 items) and behavioral support for change (6 items) and the same demographical questions as in the employee questionnaire. The questions were measured with a 5-point Likert scale (1 meaning totally disagree and 5 meaning completely agree).
Commitment to change.
Commitment to change was measured with the 6 affective items of the 18-item scale from Herscovitsch and Meyer (2002). Originally these questions are asked from the perspective of the employee. Due to the fact this study focuses on commitment to change of the employee from the perspective of the manager questions were asked from the perspective of the
manager. Sample items include “The employee believes in the value of this change” and “The employee believes this change serves an important purpose”. Cronbach’s Alpha was .819.
25
Behavioral support for change.
Behavioral support for change was measured with six items. Three items were adapted from Herold et al (2007) that represent behavioral support for change with the organization as target. Sample items include, “The employee does whatever he/she can to help this change be successful” and “The employee is fully supportive of this change”. The other three items were adapted from Venus et al (2013). Venus et al (2013) developed three additional items to ensure the measure captured the three key activities of behavioral support for change: participation, facilitation and contribution (Kim, Hornung & Rousseau, 2011). Sample items include, “The employee is willing to do more to implement the plans than is required of me” and “The employee is willing to promote the vision of change with enthusiasm”. Cronbach’s Alpha was .949.
Ambivalence.
For calculating ambivalence, I use Thompson et al.’s (1995) operationalization’s of ambivalence. They suggested that beyond simultaneously experiencing both positive and negative attitudes, ambivalence will be particularly strong when these positive and negative attitudes are strong. Therefore, they proposed the following equation:
Ambivalence = (Positive + Negative / 2) – (Positive – Negative)
This equation takes into consideration both the balance between positive and negative reactions and the intensity of these reactions. In this equation, the intensity of the attitudes is considered by calculating the mean attitude of the positive and negative strength, and the balance between them calculated through the absolute value of their difference. Consequently, the stronger positive and negative attitudes are in magnitude, the greater will the ambivalence be.
26
To calculate the positive and negative attitudes indicated in Thompson et al.’s (1995) equation, I used 5 affective items and 5 cognitive items from Oreg’s (2006) change-attitude scale. The affective items were negatively worded (e.g., “I am afraid of the change”) and the cognitive items were positively worded (e.g., “ I believe that the change would benefit the organization”). High scores signify high employee ambivalence. Cronbach’s Alpha was .912 for the negative items and .883 for the positive items.
Leader-member exchange (LMX).
Leader-member exchange (LMX) was measured with seven-item LMX-7 scale (Graen et al., 1982; see Dienesch & Liden (1986) and Graen & Scandura (1987) for reviews). Originally these questions are asked from the perspective of the employee. In this study leader-member exchange (LMX) questions was asked from the perspective of the manager Sample items for the manager include, “I have an effective working relationship with my employee” and “I recognize the potential of my employee”. Cronbach’s Alpha was .929.
27
4. Results
Means, standard deviations, inter-correlations and internal consistency for all measured variables are provided in table 1. The associations between the variables were tested with a Pearson correlation test. Table 1 shows several significant correlations. Ambivalence and behavioral support for change are negatively correlated with each other (r = -.346, p <0.01). As expected, ambivalence was associated with lower behavioral support for change.
Furthermore, ambivalence and leader-member exchange (LMX) are negatively correlated with each other (r = -.299, p <0.01). The table also shows a positive correlation between leader-member exchange (LMX) and commitment to change (r = .560, p <0.01) and behavioral support for change (r = .758, p <0.01). Furthermore, despite the fact that it is not the aim of this study, data revealed that commitment to change and behavioral support are positively correlated with each other (r = .712, p <0.01). This suggests, as anticipated that commitment to change and behavioral support for change overlap with each other and also differ from each other.
Table 1: Means, standard deviations and intercorrelations between the scales
Variables M SD 1 2 3 4
1. Ambivalence 1.3209 1.41260 (.912/.883) 2. Commitment to change 35.824 0.59700 -.197 (.819) 3. Behavioral support for
change 33.663 0.87567 -.346** .712** (.949) 4. Leader-member exchange (LMX) 35.322 0.84972 -.299** .560** .758** (.929) Note. N=91
*. Correlation is significant at 0.05 level (2-tailed) **. Correlation is significant at 0.01 level (2-tailed)
Further analysis of the data was conducted by using linear regression to test the direct
relations between ambivalence and commitment to change and behavioral support for change and moderated regression analyses to test whether the moderator, LMX had a significant effect on the direct relations between ambivalence and both commitment to change and behavioral support for change. To perform the moderated regression analyses all independent
28
variables were centered around zero by subtracting their mean. All results are summarized in table 2, 3, 4 and 5.
Hypothesis 1a predicted that ambivalence would negatively relate to commitment to change, however this relationship (as evident in table 2) is not significant (p > 0.05). Hypothesis 1b predicted that ambivalence would negatively relate to behavioral support. This model was (as evident in table 2) statistically significant F (1, 89) = 12.08; p < 0.01 and explained 12% of variance in behavioral support for change. Furthermore, ambivalence records a Beta value of b= -.34. In other words, if ambivalence increases with one, employees behavioral support for change will decrease with .34.
Table 2: Linear regression (Commitment to change and Behavioral support for change)
Commitment to change
R R2 R2
Change
B SE ST.Beta t
Ambivalence .197 .039 .028 -.083 .044 -.197 -1.892
Behavioral support for change R R2 R2 Change B SE ST.Beta t Ambivalence .346 .120** .110 -.214 .062 -.346** -3.477 **. p<0.01
Though it is not the aim of this study, a closer inspection of the effects indicates that the relationship between leader-member exchange (LMX) and commitment to change (as evident in table 3) is significant, F (2, 88) = 20.21 p < 0.01 and explained 31.5% of variance. LMX records a Beta value of b= .55. In other words, if LMX increases with one, commitment to change will increase with .55. Furthermore, the effects indicate that the relationship between LMX and behavioral support for change (as evident in table 3) is significant, F (2, 88) = 63.21 p < 0.01 and explained 59% of variance. LMX records a Beta value of b= .71. In other words, if LMX increases with one, behavioral support for change will increase with .71.
29
Table 3: Linear regression (Commitment to change and Behavioral support for change) Commitment to change R R2 R2 Change B SE ST.Beta t .561 .315** .299 Ambivalence -.013 .255 -.032 -.344 LMX .387 .065 .551** 5.956
Behavioral support for change R R2 R2 Change B SE ST.Beta t .768 .590** .580 Ambivalence -.081 .044 -.131 -1.827 LMX .740 .074 .719** 10.039 **. p<0.01
Hypothesis 2a predicted that when LMX is high the relationship between ambivalence and commitment to change is positive and hypothesis 2b predicted that when LMX is low the relationship between ambivalence and commitment will be more negative. The regression coefficient for XM is (as evident in table 4) c3= -.025 and is statistically not significant, p>0.05. Thus, the effect of ambivalence on commitment to change does not depend on LMX.
Hypothesis 3a predicted that when LMX is high the relationship between ambivalence and behavioral support for change is positive and hypothesis 3b predicted that when LMX is low the relationship between behavioral support for change will be more negative. The regression coefficient for XM is (as evident in table 4) c3= .014 and is statistically not significant
p>0.05. Thus, the effect of ambivalence on behavioral support for change does not depend on LMX.
30
Table 4 : Moderation effect (Commitment to change and Behavioral support for change)
Commitment to change Behavioral support for change Coefficient SE P Coefficient SE t P Intercept I1 3.5373 .055 64.909 .000 3.371 .063 53.871 .000 Ambivalence (X) C1 -.011 .039 -.288 .774 .082 .045 -1.835 .070 LMX (M) C2 .383 .066 5.828 .000 .743 .075 9.950 .000 LMX*Ambivalence (XM) C3 -.025 .046 -5.49 .584 .014 .052 .269 .789 R2=.317 p<0.001 R2=.590 p<0.001 F(3.87)=13.473 F(3.87)=41.270 4.1 Exploratory research
In this study commitment to change and behavioral support for change were tested from the perspective of the manager. From that perspective ambivalence has a significant negative effect on behavioral support for change, however does not have a significant effect on
commitment to change. Therefore, I also tested commitment to change and behavioral support for change from the perspective of the employee to get an understanding how ambivalent employees perceive their own commitment to change and behavioral support for change.
With regard to the relationship between ambivalence and commitment to change this model was (as evident in table 5) statistically significant F (1, 89) = 15.80; p < 0.01 and explained 37% of variance in commitment to change. Furthermore, ambivalence records a Beta value of b= -.61. In other words, if ambivalence increases with one, their behavioral support for change will decrease with .61.
Furthermore, with regard to the relationship between ambivalence and behavioral support for change this model was (as evident in table 5) statistically significant F (1, 89) = 11.26; p < 0.01 and explained 27% of variance in behavioral support for change. Furthermore,
ambivalence records a Beta value of B= -.51. In other words, if ambivalence increases with one, their behavioral support for change will decrease with .51.
31
Table 5: Linear regression (Commitment to change and Behavioral support for change) Commitment to change
R R2 R2
Change
B SE ST.Beta t
Ambivalence .616 .379** .372 -.297 .040 -.616** -7.375
Behavioral support for change R R2 R2 Change B SE ST.Beta t Ambivalence .519 .270** .262 -.250 .044 -.519** -5.733 **. p<0.01
Furthermore, in this study leader-member exchange (LMX) was tested from the perspective of the manager. From that perspective LMX has no effect on the relationship between
ambivalence and both commitment to change and behavioral support. Therefore, I also tested LMX from the perspective of the employee. As evident in table 6 the regression coefficient for XM is c3= -.027 and is statistically not significant p>0.05. Thus, despite the fact that LMX is reported from the perspective of the employee the effect of ambivalence on commitment to change does not depend on it.
Furthermore, as evident in table 6 the regression coefficient for XM is c3= .103 and is statistically not significant p>0.05. Thus, despite the fact that LMX is reported from the perspective of the employee the effect of ambivalence on behavioral support for change does not depend on it.
Table 6 : Moderation effect (Commitment to change and Behavioral support for change)
Commitment to change Behavioral support for change Coefficient SE t P Coefficient SE t P Intercept I1 3.594 .064 56.510 .000 3.412 .085 40.076 .000 Ambivalence (X) C1 -.004 .046 -.087 .931 -.091 .061 -1.479 .143 LMX (M) C2 .357 .090 3.964 .000 .572 .121 4.740 .000 LMX*Ambivalence (XM) C3 -.027 .064 .416 .678 .103 .085 1.206 .231 R2=.200 p<0.001 R2=.334 p<0.001 F(3.87)=7.247 F(3.87)=14.522
32
In conclusion, ambivalence did correlate with behavioral support for change and the linear regression showed a significant negative relationship between the both variables. Therefore, hypothesis 1b was not rejected. This relationship was also significant when behavioral support for change was measured from the perspective of the employee. This means that ambivalent employee reactions to change significantly reduces employee commitment to change and behavioral support for change in the perception of the manager. Next to that, ambivalence and commitment to change did not correlate and consequently did not significantly related to each other. However, this relationship was significant when commitment to change was measured from the perspective of the employee. Furthermore, none of the hypothesized moderation models were supported. The moderation models were also not supported when LMX was tested from the perspective of the employee instead of from the perspective of the leader. Though, it is not the aim of this study, LMX related significant with both commitment to change and behavioral support for change.
33
5. Discussion
Below, the most evident findings of this study and the implications of these findings with respect to the literature and management practice are discussed. Furthermore, the limitations with regard to the conclusions of this study are outlined. Finally, suggestions for further research are formulated.
This study contributed to recent research by examining the effect of employee ambivalent attitudes on both commitment to change and behavioral support for change and how leader-member exchange (LMX) influences this effect. It was anticipated that the direct relationship between ambivalence and both commitment to change and behavioral support for change would be negative and that this relationship depends on the quality of the relationship between the manager and the employee. This means that in a high quality relationship
ambivalence relates positively to both commitment to change and behavioral to change and in a low quality relationship ambivalence relates more negatively to both commitment to change and behavioral support for change. Although the argumentation presented when developing the hypotheses is supported sufficient in current change management literature, the statistical analysis did not show the sufficient support for all hypotheses. When focusing on the direct relationships between ambivalence and both commitment to change and behavioral support for change, results showed that ambivalence is not significantly negatively related to commitment to change (meaning H1a). However, results showed that ambivalence is significantly negatively related to behavioral support for change (meaning H1B). Furthermore, one clear pattern can be observed from the results. That is, that in all
relationships that have been tested, no significant results were found for the moderation effect of leader-member exchange (LMX) on both commitment to change and behavioral support for change (meaning H2a, H2b , H3a and H3b).
34 5.1 Theoretical implications
The fact that this study proves that ambivalent employee attitudes to change significantly reduces their behavioral support for change in the perception of the manager contributes to the existing literature on resistance. Although previous studies have argued that ambivalence toward change is an important phenomenon (e.g. Piderit, 2000, Sverdlik & Oreg, 2011) this study is the first in which ambivalence is related to both commitment to change and
behavioral support for change. We are aware of the fact that in this field of research earlier Oreg (2006) found substantiated evidence for the negative relationship between cognitive resistance and continuance commitment. This means that respondents who reported having negative cognitive evaluations regarding the change when it was introduced for the first time were also less likely to remain in the organization. Although ambivalence was not assessed in this study, the items used to measure employees’ reactions to change can be used to assess a form of ambivalence (Sverdlik & Oreg, 2011).
The significant negative relationship between ambivalence and behavioral support for change found in this study means that managers perceive ambivalent employees as not behaviorally supportive to the change. Therefore, ambivalent employees are considered not to show the willingness to do what is required of them by the organization, the willingness to do
something extra to make the change work and finally to ‘sell’ the change to others (Meyer et al, 2007). In line with the theory this suggests that manager’s misinterpret ambivalence and perceive it is as indifferent (Oreg & Sverdlik, 2011). The suggestion that manager’s
misinterpret ambivalence and could perceive it as indifferent could actually lead to employee behavior which is not supporting the change. For instance, by publicly disapprove the
employees behavior or ignore them, the employee finds him/herself excluded from the change process which will lead to employee behavior which is not behaviorally supporting the
35
To get more knowledge about the relationship between ambivalence and both commitment to change and behavioral support for change I also tested both commitment to change and behavioral support for change from the perspective of the employee. Equal to the results from the perspective of the manager, ambivalence is significantly negatively related to behavioral support for change. In detail, from both the self-reporting and managerial perspective
ambivalence has a negative effect on behavioral support for change. In contrast to the results from the perspective of the manager, ambivalence is significant negatively related to
commitment to change. The fact that ambivalence is negatively related to both commitment to change and behavioral support for change from the self-reporting perspective may result from the manager’s negative interpretation of ambivalence which leads to disapproving and
ignoring and consequently to employee behavior which is not committed to change and supporting the change. This study suggests that commitment to change and behavioral support for change are two independently occurring phenomena in the organization. From a
managerial perspective, ambivalence can decrease behavioral support for change, but commitment to change seems unaffected. From the self-reporting perspective ambivalence can decrease both commitment to change and behavioral support for change.
An explanation for the fact that ambivalent employee attitudes do not significantly reduces their commitment to change and does significantly reduces behavioral support for change in the perception of the manager, could be the fact that this study focusses on affective
commitment to change (ACC), which is an employee desire to provide support for the change based on a belief in its inherent benefits. Due to the fact that commitment is linked to
behavioral intentions (Swailes, 2004), and therefore a mind-set it is harder for a manager to value this variable in comparison to behavioral support for change, which is a observable behavior and therefore easier to value.
36
Furthermore, the quality of the relationship between the manager and the employee (LMX) does not influence the effect ambivalence has on both commitment to change and behavioral support for change. Initially, LMX was tested from the perspective of the manager. In the exploratory research LMX was tested from the perspective of the employee, yet even then there was no significant moderation effect. This means that although in the current literature several authors have described the positive attributes of ambivalence, LMX cannot influence the relationship between ambivalence and both commitment to change and behavioral support for change and therefore doesn’t change the negative perception of ambivalence. In the literature, ambivalent employees are perceived as effective change agents who provide a more balanced and realistic perspective of the opportunities and threats to the specific change is offering (Meyerson & Scully, 1995) and organizations could improve itself and its decisions by acknowledging ambivalent reactions towards change (Piderit, 2000). Moreover,
ambivalent employees need particular support and guidance during the change initiation, to alleviate the discomfort ambivalence creates (Sverdlik & Oreg, 2011). These positive attributes can only be acknowledged if manager’s see ambivalence as a possibility.
Due to the fact this study proves that ambivalence is significantly negatively related to behavioral support for change and this relationship does not depend on the quality of the relationship between the manager and the employee it suggests that managers overlook the potentially positive intentions that may motivate negative responses to change and therefore they don’t see ambivalence as a possibility (Piderit, 2000). Consequently, the positive
attributions of ambivalence to change are not acknowledged by the manager while it actually holds several positive attributes to contribute to the change. Therefore, based on this data I cannot predict that in for instance, a high quality relationship between the manager and the employee ambivalence will be perceived as a positive reaction to change by the manager. However, this does not mean that ambivalence can never be interpreted as a positive reaction
37
to change. Future researchers may use other moderators to prove that ambivalence can be interpreted as a positive reaction to change.
There appears to be at least two possible explanations for the fact that LMX does not influence the relationship between ambivalence and both commitment to change and behavioral support for change. First, according to Gerstner and Day (1997) LMX views leadership as a unique dynamic that occurs between the manager and the employee. The nature of this relationship creates an environment of reciprocity, more specifically between the manager and the employee, rather than between the employee and the organization. From the perspective of the employee, the change initiative would be recognized as coming from the organization, and not from the manager. Therefore, employee ambivalent attitudes is addressed to the organization. Any rationales for the change initiative would be viewed as needing to come from the top-management of the organization, and not from the manager.
While the manager’s influence cannot be denied, when an organizational change causes ambivalent employee attitudes the quality of the relationship between the manager and the employee does not influence the effect this behavior has on both commitment to change and behavioral support for change. Reactions to change, and in this case, ambivalent employee attitudes may greatly overshadow the quality of the relationship between the manager and the employee and therefore the managerial interpretation of ambivalence does not depend on the quality of this relationship (Armenakis et al, 2007). In detail, because ambivalence is
addressed to the organization and not to the manager it doesn’t matter if the relationship between the manager and the employee is good or bad. The quality of the relationship between the manager and the employee cannot influence their negative perception of ambivalent employee behavior.
38
A second possible explanation of why LMX did not influence the effect ambivalence has on both commitment to change and behavioral support for change is the limited variance in the LMX variable. Table 1 shows that the mean value of LMX was relatively high (M = 3.522) and the variance relatively low (σ = .849). From a statistical point of view there may not have been enough respondents with unfavorable LMX conditions to have effect on the relationship. Using this possible statistical explanation, I will not discount the importance of LMX in organizational change. In this study there may not have been enough variation in our respondents’ assessment of LMX to produce a moderator effect (Armenakis et al, 2007).
Though it was not the aim of this study, results show that LMX is significantly positively related to both commitment to change and behavioral support for change. LMX significantly augments employees commitment to change and behavioral support for change in the
perception of the manager. This means that the quality of the relationship between the manager and the employee has a direct effect on employee commitment to change and behavioral support for change in the perception of the manager. In detail, when the employee and the manager have a good relationship, the employee will be committed to change and behaviorally support the change in the perception of the manager. As LMX turned out to be a strong predictor of commitment to change and behavioral support for change, the research results that have been obtained can be interpreted as a confirmation of the attribution theory of Heider (1958) which suggests that an employees´ reaction to managerial influence attempts could be a reflection of the quality of the interpersonal relationship between the employee and the manager (Ferris & Judge, 1991; Porras & Robertson, 1992). The expected behaviors of both parties are shaped by this relationship (Furst & Cable (2008).
5.2 Limitations and future research
This study has next to its well-grounded theory on ambivalence and the relations with
39
is the exclusive reliance on cross-sectional data collection, which limits the ability to discuss causal relations. This cross-sectional study provides a clear ‘snapshot’ of the outcome and characteristics associated with it, at a specific point of time, namely before the change is initiated (Taris & Kompier, 2014). Therefore, based on findings I can employ a relatively passive approach to making causal inferences (Levin, 2006). Due to the fact that the
measurement is only based on cross-sectional statistical results via non open ended surveys, further research should conduct a longitudinal data collection which increases the ability to discuss causal relations. Furthermore, it could be interesting to conduct in-depth interviews to find underlying reasons and support for the results. This means that this study cannot further explain for instance the cause for the negative relation between ambivalence and behavioral support for change.
The second limitation of the study concerned the method. Data was partially (15%) derived from the organization I am working for which could be seen as limiting. However, I took precautions by emphasizing to the employees that I conducted the questionnaire as a student, and not as employer of the organization. Due to my employment my approach to the subject as a researcher could be seen as limiting.
The third limitation of this study is the survival bias. Due to the fact that in only 2 out of 9 (only 13.2% of the data) organizations who were studied the initiated change emerged from negative circumstances the study is mainly performed at organizations where the initiated change emerged from positive circumstances. Despite several attempts to perform the study at organizations which initiated change from negative circumstances, this proved unfeasible. I have experienced that these organizations are generally less willing to participate in such a study. I am convinced that when I could have studied more organizations who initiated change emerged from negative circumstances the ambivalence rate would be significantly
40
higher and more hypotheses would be supported. Therefore, future research should conduct the research within organizations where initiated change emerged from negative
circumstances.
The fourth and last limitation is the sensitivity of the topic. Because the questionnaires were distributed in person, the reasons for not cooperating to fill in the questionnaire was observed directly. There were two main reasons against filling in the questionnaire; the first was because of time constraints and the second because employees and also managers perceived some questions, mainly the LMX questions, as too sensitive. Furthermore, there is a chance that participants who did fill in the questions, however didn’t feel comfortable by doing so, have not answered the questions completely honestly. To reduce this, participants were informed before filling in the questionnaire that the data would only be revealed to the researcher. Furthermore, envelopes were distributed so that participants could return the questionnaire anonymously.
The research results successfully confirmed only one of the six hypotheses, and therefore shows a need for further research. Especially the mechanism that suggests that manager´s misinterpret employee ambivalence and see it as indifferent, which leads to employee behavior which is not behaviorally supporting the change needs more attention in future research. This study raises the question whether ambivalent employee attitudes are really disapproved and ignored by managers. If this could be confirmed, the next question is whether this instigates employee behavior which is not committed and supportive to change. Future researchers could elaborate on this mechanism by studying how manager´s actually respond to ambivalent employee attitudes and consequently how employees behaviorally react to this response.
