Two Sides of the Same Coin: The Role of Change Agents in
Recipient Resistance to Change
Karina Klaus S1940201
University of Groningen Faculty of Economics and Business
MSc BA Change Management
Supervisor: Dr. J.F.J. Vos Co-Assessor: H.P. van Peet
2 Abstract
Even though it is obvious that there are two parties involved in the change process, namely the change agent and the change recipient, most research on resistance to change has taken the perspective of only one party, and therefore neglecting the other and also possible differences in perceptions that might exist due to different roles and interests. In a sample of 100 change projects, each consisting of groups of one change agent and a number of corresponding change recipients, we found that there are differences in perceptions of resistance and also in how leadership behavior affects resistance. Agents perceived more resistance than recipients did. Additionally, agents did not think they had an influence on recipient resistance while recipients perceived agents engaging in Shaping leader behavior to decrease resistance. The scope of the change project was found to be a moderation for Shaping as well as Framing leader behaviors. Implications of the findings as well as suggestions for further research are discussed.
Keywords: resistance to change, change leadership, change agent, change recipient, scope of
3 TWO SIDES OF THE SAME COIN: THE ROLE OF CHANGE AGENTS IN RECIPIENT
RESISTANCE TO CHANGE
In today’s fast moving environment, change seems inevitable. Modern technology permits tremendous flows of information, allowing organizations to perform globally, around the clock. To be able to compete in these fast changing markets, organizations have to constantly adapt and, even better so, innovate – otherwise they risk losing competitiveness or company death (Burnes, 2014). In order to bring change, many organizations rely on change agents to communicate the necessary measures to the employees whose work lives will be affected (Ford & Ford, 1995, 2009). Change agents are the ones who sponsor and promote the change by taking the necessary measures to make the change happen, such as creating a vision of the desired outcome and
providing leadership to change recipients, the employees who have to implement the new ways of working, thereby adapting to the change (Ford, Ford, & D’Amelio, 2008; Higgs & Rowland, 2011; Kanter, Stein, & Jick, 1992). This distinction between two parties implies that both agents and recipients have different perspectives on the change process. A reason for this might be that, based on social-cognition perspective, the two parties are likely to have different frames of reference (Hauenstein & Foti, 1989), reflected by the different roles, responsibilities, and interests they have.
4 agent-centric view has become commonly accepted, which is also reflected by the fact that it is described in the majority of management textbooks (Dent & Goldberg, 1999). However, by focusing on one side of the coin only, by definition, the other side is being neglected
(Bouckenooghe, Devos, & Van den Broeck, 2009; Oreg, Vakola, & Armenakis, 2011). Thus, Ford et al. (2008) criticize the traditional view of recipient resistance to change. They argue that both sides of the coin, that is, the perspectives of both parties involved in the process, have to be taken into account in order to fully understand resistance to change. Their main criticism is that the traditional perspective sees resistance as something solely inherent to the recipients.
However, they argue that change agents can act in ways that contribute to resistance.
The manner in which change agents act during a change process might have an influence on recipient resistance to change as well. Even though this factor is relatively unexplored
(Klonek, Lehman-Willenbrock, & Kauffeld, 2014), research suggests that leader behavior at change implementation does affect recipient resistance. Examples of this are, among others, the way agents communicate the change (Higgs & Rowland, 2005, 2011), and whether they stick to agreements and preserve trust (Tomlinson, Dineen, & Lewicki, 2004).
Another interesting factor that might influence resistance is the scope of the change, which refers to the impact, the level of novelty the change brings (Greenwood & Hinings, 1996, Plowman et al., 2007). Naturally, the scope can be small and only little changes are made or large and things change drastically (Burke, 2002; Hage, 1999; Weick & Quinn, 1999). Changes with large scope imply a larger impact on employees’ work lives than changes with low scope
(Plowman et al., 2007). Del Val & Fuentes (2003) found that changes with large impact result in more recipient resistance than lower impact changes.
5 resistance has mostly taken a unilateral approach. We acknowledge the change process as non-linear and complex and thus being likely to be affected by a great variety of factors. The
contextual complexity of change implementations can only be acknowledged by recognizing that both change agent and recipient have their own perspectives of the same change process. By taking a bilateral view we strive to address this gap and get a clearer view on resistance and the possible contribution of leader behavior. To the best of our knowledge, this is one of the first attempts to take a bilateral perspective on leader behaviors and their effects on change resistance. Because of this exploratory character, we are going to answer open questions. The overarching research question that leads our reasoning is the following:
In what way are perceptions of the change agents and change recipients different regarding the agents’ behavior and how is this related to recipient resistance towards the change?
Concerning practical contributions, we expect to be able to make inferences for the practice of change management. Specifically, we aim at change managers that are directly involved in the change process and thereby communicating with the recipients.
THEORETICAL BACKGROUND
In the following we are going to review the relevant literature with regard to the concepts as stated in the introduction. Based on this literature we will form more specific research
questions. A conceptual model of our research will be provided at the end of this section. Resistance to Change
One of the key topics in the change literature is resistance to change, which can be
6 namely, that resistance is inherent to the change recipient. Here, resistance is seen as negative behavior intended to ruin or at least alter the change process. As such, resistance is seen as problem that has to be minimized or eliminated, as it otherwise might lead to failure of the change process (Armenakis et al., 1993; Maurer, 1997; Michael, By, and Burnes, 2013; Nadler, 1993). It is believed that recipients are not motivated to change unless they have compelling reasons to do so (Bouckenooghe, 2009). Reasons that make them resist are, among others, the human desire to maintain the status quo (Lewin, 1952; Conner, 1998), dislike of the change content itself, fear of uncertainty, anxiety of possible consequences, and belief that the change is not necessary (Palmer et al., 2009). Other studies focus on what change leaders can do in order to reduce resistance, having in mind that resistance can obscure the change process (Del Val & Fuentes, 2003). An example for this is that agents should enable recipients to actively take part in the change process. This way, they see it as their own project, as part of their own responsibility, and as a consequence are less likely to resist (Boohene & Williams, 2012; Higgs & Rowland, 2005, 2011). In this view, it is the burden of the change agent to overcome resistance. Agents are seen as “undeserving victims of the irrational and dysfunctional responses of change recipients” (Ford et al, 2008: 362). However, Ford et al. (2008) criticize this agent-centric view of resistance. They argue that change agents might have a much more active role in resistance than commonly acknowledged and that it is not reasonable to assume that recipients resist just for the sake of resisting, regardless of context and agent behavior.
The more modern view of resistance takes into account this criticism and considers both the agent and the recipient. Adding to the traditional view of resistance, the more modern view holds that leaders can actively contribute to resistance through their own behavior, such as
7 negative and destructive but has positive sides as well, such as being a useful source of
information for change leaders to develop more successful change approaches (Ford et al., 2008; Ford & Ford, 2009; Piderit, 2000; Waddell & Sohal, 1998). Additionally, Ford et al. (2008) suggest that resistance often results in conversations about the topic that is being resisted.
Interpreted positively, this would mean that if employees talk about the change, the conversations keep the topic in existence and thus allow more people to participate. A second positive aspect of resistance is, according to Ford et al. (2008), that resisting change provokes thinking and
consideration as well as an inner conflict in recipients. They argue that thoughtful acceptance results in more commitment than unthoughtful acceptance, especially after resolving an inner conflict. Moreover, this consideration can result in valuable suggestions for further improvement of the change.
In pursuit of taking the modern perspective further, we are interested in the different perspectives that agent and recipient might have during the change process. As the different roles, responsibilities, and interests of the two parties imply, we would like to investigate whether agents and recipients differ in their perspectives on the level of resistance. Therefore, we pose the following research question:
Q1: To what extent do agents and recipients have different perceptions regarding the level of resistance?
Change Agent Behavior
8 accountable and using an individual, leader-centric focus, 2) Framing, which is characterized by designing and managing the change journey and communicating guiding principles, and 3)
Creating capacity, which is characterized by creating capabilities, communicating, and making
connections. Analyses of these behaviors show that leader-centric Shaping behaviors have a negative impact on recipient behavior (Avolio, Walumba, & Weber, 2009; Higgs & Rowland, 2005; Judge, & Piccolo, 2004). Additionally, Higgs and Rowland (2005)1 found that Framing and Creating were positively related to recipient behavior. These findings are also supported by other researchers (e.g. Kee & Setzer, 2006; Kramer, 2005; Senge, 1997), who advocate the need to move away from a leader-centric approach as complexity increases. For example, Kee and Setzer (2006) argue that the top-down implementation of leader-centric approaches can cause resistance.
Next to their leadership behavior, agents can contribute to recipient resistance in other ways. During a change process, both the recipient and the agent engage in sense-making, a process in which they try to seek and interpret information about the consequences, ascribe meaning to them, and decide how to react to the process (Thomas, Clark, & Gioia, 1993). Ford et al. (2008) state two important ways in which change agents can actually contribute to resistance. First, by means of expectation effects, such as self-fulfilling prophecies. Here, change agents who expect to find resistance are likely to behave as if they were already facing it. By acting in ways that are meant to minimize the seemingly existing resistance, they are likely to cause the very behavior that they were trying to reduce (Dent & Goldberg, 1999). Second, change agents may use resistance as self-serving account. That is, they might blame difficulties in the change process
1In a later study, Higgs and Rowland (2011) further defined the roles of a change leader. They distinguished among
9 on resisting recipients rather than looking for problems in their own behavior in order to make themselves look better. This way agents might conceal ineffective behavior, such as the failure to call people to action or to legitimize the change (Ford et al. 2008).
Adopting the three broad sets of leader behavior described by Higgs and Rowland (2005), this study examines the relationships among those behaviors and recipient resistance. Thus, our second question is:
Q2: To what extent do agents and recipients have different perceptions of the leader behavior and its contributions to resistance to change?
Scope of Change
Another interesting factor that might influence resistance is the scope of a change. The scope of a change concerns its magnitude and impact, meaning the level of novelty the change brings and how much the work lives of the recipients have to change accordingly. The scope can range from low to high, with a high scope indicating a higher impact on employees’ work lives than changes with low scope (Plowman et al., 2007). We believe that the scope of a change project is likely to influence both agents and recipients. Agents are bound to be influenced in how they will try to implement the change and thus which leader behavior to adopt (Burnes, 2014). Del Val & Fuentes (2003) found that changes with large impact result in more recipient resistance than lower impact changes. Against this background, our third question is:
Q3: How does the scope of a change project influence the relationship between leadership behaviors and recipient resistance to change, for both agents and recipients?
10 METHOD
Participants and Procedure
Participants were 100 agents with 309 corresponding recipients. In a field study, we collected data on 100 groups of agent-recipient combinations from 95 change projects in 75 organizations. Selection criteria were at least one corresponding recipient per change agent and that the agent-recipient groups had undergone a change within the past three years or were currently in the process of changing. Per agent, most recipient groups consisted of 3-4 recipients (68% of the sample) with a minimum of one and a maximum of ten recipients.
The group of change agents (N = 100) consisted of 70% males and 30% females, with a mean age of 43 (SD = 10). The majority of the participants had the Dutch nationality (92%), 7% were German and one agent was Surinamese. Most agents completed higher vocational training or university (80%), others had received lower vocational training (13%), or other (2%). The
Q2
Q3
Figure 1: Research framework.
Note: Framework will be tested for agent and recipient separately (bilateral approach)
Leader Behavior Shaping, Framing, Creating
Context Scope
11 agents in our sample had an average experience of 7 years (SD = 6.3) in their current position and an overall work experience of 18.5 years (SD = 10.4).
The group of change recipients (N = 309) consisted of 56% males and 44% females, with a mean age of 39 (SD = 12.1). The majority of the recipients had the Dutch nationality (93%), the other 7% were German. The educational level ranged from higher vocational training or
university (57.5%) to lower vocational training (33.4%), or other (7.1%). The recipients in our sample had an average experience of 7 years (SD = 6.8) in their current position and an overall work experience of 17 years (SD = 11.2).
In addition to an existing database2, respondents were approached within the author’s personal network, using a convenience sampling approach. The main criterion was that
respondents were part of a work team which had undergone a change project within the past three years, or which was still in the process of changing. The cut was made at a time span of three years to ensure that the respondents were still able to accurately recall their experiences. As we wanted to include context in the data, we had no criteria regarding topic of the change, industry, company size, type of organization, or organizational level.
After agreeing on taking part in the study, the respondents were sent a link to the respective version of the questionnaire (either the agent or the recipient version) and a personal identification code via e-mail. The code allowed us to create agent-recipient pairs for further analysis. For the questionnaire the online survey software ‘Qualtrics’ was used. This software allowed respondents to fill in the questionnaire at a time that suited their schedules. When participants did not fill out the survey within a week, they were sent a friendly reminder.
12 Data Analysis
After performing preliminary analyses concerning missing values and assumptions, the scales of the variables used were subjected to factor analyses in order to verify whether the expected factor structure emerged. Next, the first question was answered by comparing the means of both groups with the help of an independent t-test. Question two was answered with a
hierarchical linear regression. The dependent variable (DV) was Resistance and the three leader behaviors Shaping, Framing, and Creating were the independent variables (IV). Scope, Age, and
Sex were used as control variables. In order to answer question three, the interactions of Scope
and the three different leader behaviors were entered in a third step of the hierarchical regression. Since we were interested in the different perspectives, we conducted the analyses for both the agent and recipient sample separately and then compared the results with each other. Because the agents were asked to rate recipient resistance for the group of recipients as a whole and the recipients rated their own resistance, we aggregated the recipient scores per project to the group level.
Measures
In order to examine whether the constructs we measured were distinct, we conducted a factor analysis with all survey items. The results of this analysis can be found in Table 1. Table 1
Factor loadings of all survey items
Item: Agent version / Recipient version Factor loadings
1 2 3 4 5
Recipient Resistance During the change…
Res_1…the employees (I) sought ways to obstruct the change. -.03 -.09 .81 -.1 .05 Res_2…the employees mutually (I) complained (with others) about the
change.
-.00 .07 .88 -.05 .09 Res_3…the employees (I) expressed their (my) complaints regarding the
change towards management.
13 Leader Behavior: Shaping
During the change, I (the change agent) …
Shap_1…regularly used my (his/her) experience to shape the implementation of the change.
.11 .02 -.04 -.69 -.05 Shap_2…showed my (his/her) personal involvement to motivate people
to accept the change.
.07 .05 .04 -.69 -.18
Shap_3…pointed the employees at their responsibilities regarding their role in the change.
.23 .03 -.01 -.69 .13 Shap_4…regularly tried to bring up my (his/her) views about the change. -.06 .08 -.02 -.82 -.03 Shap_5…implemented the change based on my (his/her) previous
experience with other changes.
-.01 -.05 .02 -.84 -.02
Leader Behavior: Framing
During the change, I (the change agent) …
Fram_1…kept the change focused on the organizational goals. .53 .16 -.1 -.09 -.17 Fram_2…did not make things look better than they were. I (He/She) stuck
with the reality.
.8 .01 .01 -.02 .06
Fram_3…did not shy away from difficulties. .76 .04 .00 -.06 -.05
Fram_4…knew how to keep the employees engaged in change even in difficult times.
.59 -.06 -.07 -.19 -.19
Fram_5…did not compromise from quality. .65 .00 .01 .05 -.05
Fram_6…discussed beliefs and assumptions when needed. .56 .03 .00 -.15 -.17 Fram_7…communicated in a straightforward manner with the employees
and did not hold anything back.
.81 .01 .01 .01 .14
Fram_8…knew how to create confidence and trust in difficult times. .67 -.06 -.13 -.01 -.24 Fram_9…showed confidence that I (he/she) would bring this change to a
successful conclusion.
.78 .04 -.05 -.03 .05 Fram_10…set clear rules and boundaries so that the employees knew
where they stood.
.81 -.1 .06 -.03 .09 Fram_11…confidently carried out my (his/her) vision, even when it was
tough to do so.
.59 .07 .03 -.14 -.11
Leader Behavior: Creating
During the change, I (the change agent)…
Crea_1…ensured a safe environment for others by making myself (him-/herself) vulnerable.
.32 .02 .06 -.14 -.52 Crea_2…made difficult topics discussable by trying out and living
through these difficulties.
.50 -.01 .05 -.09 -.44 Crea_3…ensured that there was room for the employees to think
differently.
.3 .11 .01 -.19 -.52 Crea_4…took the employees out of their daily routine to allow them to
think differently.
14 Crea_5…spent time with the employees to come up with creative
solutions.
.19 -.03 -.04 -.28 -.53
Scope of the change
(The same items for both agent and recipient)
Scope_1 This change affected the work of the employees very much. .16 .79 .03 .05 .18 Scope_2 This change affected the responsibilities between the employees
very much.
.05 .79 .04 -.06 .3 Scope_3 This change affected the partnerships between the employees
very much.
-.17 .70 .03 -.16 -.10 Scope_4 Through this change, the work conditions of the employees
changed very much.
-.05 .68 .07 .18 -.27
Scope_5 This change is pioneering. .03 .64 -.11 -.11 -.1
Note: Extraction method: Principal component analysis. Rotation: Oblimin with Kaiser Normalization
All constructs loaded on different factors, which showed that the constructs were distinct. Items Crea_1 and Crea_2 (in Table 1: items 20 and 21) were an exception, however, as they loaded high on both, components 1 and 5, referring to Framing and Creating leader behavior, respectively. The literature describes the behavior that is measured by those items (‘Container’) as being part of both, Framing and Creating leader behavior (Higgs & Rowland, 2011). Because the items loaded high on both constructs and the resulting pattern was still reflecting what is described in the literature, we decided to keep both items.
15 the five behavioral items because we considered the other two not perfectly suited for measuring resistant behavior. Because Oreg (2006) used a unilateral perspective, the original questions were posed from the point of view of the recipient. Therefore, we translated the same questions to the agent view, so that we could take a bilateral perspective. The items used can be found in Table 1. All answers were given on a Likert type scale ranging from 1 (“strongly disagree”) to 7
(“strongly agree”).
The three items were all significantly correlated, ranging from r = .418 to r = .623. The determinant of the correlation matrix was .407, which was higher than the necessary value of .00001, which led us to exclude the problem of multicollinearity (Field, 2009). The values of the anti-image correlation matrix were all above the minimum of .5 (Field, 2009), with the lowest value being .609. Bartlett’s test of sphericity was significant χ² (3) = 365.475; p = .000, and in combination with the KMO value of .656, the sampling adequacy was verified (Field, 2009). After verifying that factor analysis was appropriate, we ran a principal component analysis. This analysis extracted one component, which explained 69.371% of the variance. This was supported by the Scree Plot. As only one factor was extracted, we did not add a rotation method in a second step. The results of the factor analysis can be found in Table 2. The subsequent reliability
16 Table 2
Factor loadings resistance scale
Item Factor loadings
1
1. Res_1 .789
2. Res_2 .890
3. Res_3 .817
Total % variance explained 69.37%
Cronbach’s α .773
Leader behavior scale. Initially, this scale was developed based on Higgs and Rowland (2011), who found five distinct leader behaviors. As an extension of their earlier work, Higgs and Rowland (2011) refined the three leadership styles Shaping, Framing, and Creating which they described in their 2005 paper. Specifically, they further distinguished Framing and Creating into the new sub-components Attractor, which is characterized by pulling people towards
17 Table 3
Pattern matrix leader behavior scale
Item Factor loadings
1 2 3 Shaping 1. Shap_1 .087 .661 -.106 2. Shap_2 -.023 .638 -.296 3. Shap_3 .242 .702 .095 4. Shap_4 -.053 .829 -.038 5. Shap_5 .009 .834 .035 Framing 6. Fram_1 .494 .172 -.143 7. Fram_2 .757 .057 .014 8. Fram_3 .696 .067 -.119 9. Fram_4 .506 .117 -.304 10. Fram_5 .609 -.084 -.114 11. Fram_6 .474 .121 -.264 12. Fram_7 .821 .052 .148 13. Fram_8 .605 -.062 -.328 14. Fram_9 .788 .068 .049 15. Fram_10 .816 .007 .054 16. Fram_11 .541 .140 -.152 Creating 17. Crea_1 .103 .062 -.734 18. Crea_2 -.085 .163 -.726 19. Crea_3 .001 .110 -.756 20. Crea_4 .148 -.001 -.691 21. Crea_5 .334 -.036 -.615
Note: Extraction method: Principal component analysis. Rotation: Oblimin with Kaiser Normalization
The final scale consisted of a total of 21 items reflecting the three original leadership styles as described by Higgs and Rowland (2005): Shaping (five items, example: “I often
18 the change agent.”), Creating (five items, example: “I organized discussions with the employees to come up with different solutions.”). The example questions given reflect the agent version of the questionnaire. All items used can be found in Table 1. All answers were given on a Likert type scale ranging from 1 (“strongly disagree”) to 7 (“strongly agree”).
All 21 items were significantly correlated, ranging from r = .313 to r = .639. The values of the anti-image correlation matrix were all above the minimum of .5 (Field, 2009), with the lowest value being .928. Bartlett’s test of sphericity was significant χ² (210) = 5439.94; p = .000, and in combination with the KMO value of .959, the sampling adequacy was verified (Field, 2009). After verifying that factor analysis is appropriate, we ran a principal component analysis with Oblimin rotation method and a fixed number of three factors to extract. The three extracted factors cumulatively explained 62.54% of the variance. Internal consistency was good: Shaping behavior (Cronbach’s α = .87), Framing behavior (α = .93), and Creating (α = .87). The intra-class correlation coefficients were, respectively, for Shaping ICC(1) = .86 and ICC(2) = .87, for
Framing ICC(1) = .92 and ICC(2) = .93, and for Creating ICC(1) = .87 and ICC(2) = .87, which
indicates almost perfect agreement and therefore aggregation to the group level was appropriate (Bliese, 2000).
Scope of change scale. Scope was measured with six items of a scale developed by Vos and Brand (2012). Sample items are “This change affected the work of the employees very
much” and “Through this change, the work conditions of the employees changed very much” (see Table 1). All items were measured on a Likert-type scale ranging from 1 (“strongly disagree”) to 7 (“strongly agree”).
19 anti-image correlation matrix were all above the minimum of .5 (Field, 2009), with the lowest value being .655. Bartlett’s test of sphericity was significant χ² (10) = 588.66; p = .000, and in combination with the KMO value of .722, the sampling adequacy was verified (Field, 2009). After verifying that factor analysis was appropriate, we ran a principal component analysis. As this analysis only extracted one factor, we did not add a rotation method in a second step (Field, 2009). After the initial round we removed one item which did not meet the cut off of .4
(Matsunaga, 2010). The second round extracted one component, which explained 53.37% of the variance. This was supported by the Scree Plot. The results of the factor analysis can be found in Table 4. The subsequent reliability analysis resulted in a Cronbach’s α of .78 and an intra-class correlation coefficient of ICC(1) = .77 and ICC(2) = .78, which indicated substantial agreement and therefore aggregation to the group level was appropriate (Bliese, 2000).
Table 4
Factor loadings scope scale
Item Factor loadings
1 1. Scope_1 .789 2. Scope_2 .780 3. Scope_3 .726 4. Scope_4 .675 5. Scope_5 .675
Total % variance explained 53.37%
20 RESULTS
Correlations and Associations between the Variables
Table 5 depicts the means, standard deviations, and correlations between the variables for agents and recipients, respectively. The recipient sample was aggregated to the group level. This is because our questionnaire asked the agents to rate recipient resistance for the whole group while recipients rated their own resistance. Thus, for all analyses of the recipient sample we used the aggregated data.
In both samples the three leadership styles Shaping, Framing, and Creating were
significantly correlated. Interestingly, however, there were notable differences between agent and recipient perspectives regarding the relationship between the three leadership styles and recipient resistance. In the agent sample, none of the leadership behaviors was significantly correlated with resistance. That is, the agents in our sample did not report lower or higher levels of recipient resistance, regardless of their leader behavior style. This indicates that the agents did not perceive their leader behavior to have an influence on recipient resistance. In contrast, the recipient sample did indicate significant negative correlations between all three leader behavior styles and
recipient resistance. This pattern suggests that recipients think that all three leader behaviors contribute to lower resistance, while agents do not see a relationship between their leader behavior and recipient resistance.
The correlations between Scope and all three leader behaviors is significantly positive in the agent sample. That is, the agents in our sample reported that they displayed more of each of the three behaviors in change projects with larger scope. Again, in contrast, the recipient sample showed a different pattern. Here, Scope was not significantly related to any of the leader
behaviors. This indicates that recipients did not perceive a different in leader behaviors in
21 change projects with larger scope, while recipients do not see an association between these
variables, which is reflected in a non-significant correlation for the recipient sample. Another interesting pattern can be seen with the variable Age. While we found no
significant correlations in the agent sample, the recipient sample showed significant correlations between Age and all variables expect for Shaping. Age and Framing were negatively correlated, indicating that the older recipients perceived less Framing behavior of the agents than younger ones did. Regarding Creating, the correlation was positive, implying that older recipients perceived more Creating behaviors than the younger ones did. The positive correlation between Age and Scope might indicate that the older recipients perceived the scope of the projects to be larger than the younger recipients in our sample. Moreover, the positive correlation between Age and Resistance implies that the older recipients perceived more resistance than the younger ones did.
Table 5
Means, standard deviations, and correlations among the variables
22 3. Creating 4.45 1.17 .40** .71** -
4. Scope 4.33 1.24 .03 .04 .07 -
5. Age 39 12.13 .00 -.14* .16* .17** -
6. Resistance 2.59 1.34 -.22** -.43** -.31** .17** .32** -
Note. N =100; recipient data is aggregated to the group level
**p<.01, *p<.05
Analysis of the Research Questions
Our first open question was to what extent agents and recipients have different
perceptions regarding the level of resistance. In order to analyze this question, we compared the means of Resistance of the agent and recipient sample with the help of an independent t-test. The analysis revealed a significant difference between the mean of the agent sample (M = 3.77, SD = 1.48) and the recipient sample (M = 2.59, SD = 1.34) with t(155) = 7.04 , p<.001. Thus, the agents reported higher levels of resistance than recipients did.
Our second open question was to what extent agents and recipients have different perceptions of the leader behavior displayed and also its contribution to recipient resistance. To test this, we used hierarchical regression analysis. In the first step of the regression we controlled for Scope, Age, and Sex. In the second step we regressed Shaping, Framing, and Creating on
Resistance. Taking a bilateral view, we analyzed the agent and recipient samples separately in
23 significant main effect: Shaping: B = -.23 , p = .36; Framing: B = -.26, p = .4; Creating: B = .32,
p = .13. It is important to note, however, that both Model 1 and 2 were not significant and
therefore have to interpreted with caution.
In the recipient sample, we used the same analysis, with the only exception being that we used the aggregated scores. Here, we found an increase in R² from .04 (p <.01) in model 1 to .13 (p< .001) in model 2. Shaping was significantly negative related to Resistance (B = -.25, p <.01). This indicates that, according to the recipients in our sample, leaders engaging Shaping
leadership styles decreased the resistance to change. Framing and Creating were not significant (Framing: B = -.38, p = .09; Creating: B = -.04, p = .84), which indicates that these types of leader behavior did not significantly contribute to recipient resistance in our sample. Since we controlled for Scope, Age, and Sex in the second model, it can be said that the effects found are independent of these variables. Age also had a significant positive association with recipient resistance (B = .03, p < .01), indicating that older recipients perceive more resistance than younger ones.
These results suggest that agents and recipients have different views on the agent’s contribution to resistance. The agents in our sample do not seem to think that their leadership behavior has an influence on recipient resistance. The recipients, however, do seem to think that leaders with Shaping styles contribute to a decrease in resistance, as opposed to Framing and Creating leaders, who they do not seem to contribute. This is also reflected in the increase in R² in the recipient sample.
24 we added the interactions between the three leader styles and Scope. This way, we controlled for
Scope and the leader behaviors in the first two steps. In order to decrease possible
multicollinearity between an interaction term and its corresponding main effects and to facilitate interpretation, we first mean centered all variables (Aiken & West, 1991). The results for the agent sample did not show any significant main effect of the leader behaviors on recipient
resistance. It did indicate a slightly significant interaction effect between Shaping and Scope (B = .49, p = .03). These results indicate that Shaping leader behavior decreased recipient resistance in in change projects with low scope and increased recipient resistance in projects with high scope (see Figure 2). Additionally, we found a significant interaction effect between Framing and
Scope (B = -.72, p <.01). Interestingly, this implies that Framing behavior decreased recipient
resistance in projects with high scope and increased resistance in projects with low scope (see Figure 3).
Figure 2: Interaction effect between Shaping and Scope. 1 2 3 4 5 6
Low Shaping High Shaping
Re
sis
tance Low Scope
25 Figure 3: Interaction effect between Framing and Scope.
In the recipient sample, we found a significant main effect for Shaping on Resistance (B = -.24, p <.01). This suggests that, according to the recipients, leaders engaging in Shaping leader behaviors decrease recipient resistance to change. This finding is similar to what we found in model 2. However, we did not find an interaction effect. A summary of the regression analyses can be found in Table 6.
Table 6
Hierarchical linear regression results
RESISTANCE TO CHANGE
Agent Perspective Recipient Perspective
M1 M2 M3 M1 M2 M3 Step 1. Scope .28* .30 .26 .08 .12 .10 Age -.00 -.01 -.00 .03** .03** .03** Sex -.22 -.25 -.44 -.05 -.11 -.11 Step 2. Shaping -.24 -.31 -.25** -.24** 1 2 3 4 5 6
Low Framing High Framing
Re
sis
tance Low Scope
26 Framing -.25 -.13 -.37 -.38 Creating .32 .31 -.03 -.04 Step 3. Shaping x Scope .49* .06 Framing x Scope -.72** -.04 Creating x Scope .11 -.05 Total R²(ΔR²) .05 .09(.04) .16(.07) .04 .09(.05**) .13(.0)
Notes. Unstandardized regression coefficients are displayed. Recipient data is aggregated to the
group level **p<.01, *p<.05
DISCUSSION Summary of the Findings
Resistance to change has been studied extensively by researchers. Still, even though it might be obvious that there are two parties involved in a change process, namely the agent and the recipient, literature has mostly taken a unilateral view. This way, literature did not
27 resistance. Ford et al. (2008) also argue that agents might engage in self-serving accounts and label difficulties in the process as resistance in order to feel better about themselves. Another reason might be that agents see recipient resistance as a threat to successful change, and therefore making their job harder. This might lead them to be more aware of possible negative
consequences and thus interpret situations as resistance even though recipients themselves do not intend to resist (Ford et al., 2008; Young, 2000). This might be supported by the notion of Kotter (1995) who argues that change is often associated with more risk, pressure, and urgency than normal organizational activities. This rather stressful situation can lead agents to be more frustrated by and less tolerant of habitual behaviors displayed by recipients and thus label it as resistance.
In our second question we were interested in the relationship between the leader behaviors Shaping, Framing, and Creating (Higgs & Rowland, 2005) and recipient resistance. Our results suggested that the agents in our sample did not perceive a relationship between their leader behavior and recipient resistance at all. This is interesting because it might support the
argumentation of Ford et al. (2008), who criticized that resistance is seen as something happening ‘over there’, that is, on the side of the recipient, with the agents having no role in this. Our
28 The recipients, however, did report that leaders engaging in Shaping behaviors
contributed to a decrease in their resistance. This is remarkable, because literature supports the view that Shaping leaders are more likely to increase resistance (Higgs & Rowland, 2011). However, Higgs and Rowland (2011) also found that a lot of the more successful change projects they studied did contain a notable amount of Shaping leader behavior. These authors then
reasoned that a degree of leadership direction might be necessary for successful change implementation and that Shaping behavior might be further distinguished into behavior that serves the purpose of the organization and behavior that serves the leader only. Our findings point into the same direction. Shaping behavior that is directed to serve the organization might provide direction and structure with a common organizational goal. As change is a threat to the status quo and is often accompanied by anxiety (Conner, 1998), this direction might help the recipients cope and, in turn, decrease anxiety.
Our third question was about a possible influence of change context, specifically, the scope of the change project, on the relationship between leader behaviors and recipient resistance. Here, we found two significant interaction effects for the agents. That is, agents indicated that Shaping leader behavior decreased recipient resistance in change projects with low scope but increased recipient resistance in projects with high scope. As with the findings regarding our second question, this implies that a degree of leadership direction might be necessary in order to implement change. However, as scope and thus impact on the recipients increases, recipients seem to resist this direction a little bit more. This is consistent with Higgs and Rowland (2011) who found that Shaping leader behaviors were present in successful changes as well, but that the most successful change projects were predominantly characterized by the absence of Shaping.
29 situations. Again, this is consistent with Higgs and Rowland (2011), who found that most
successful leader styles were enabling. It might be argued that Framing leaders help recipients to cope with the impact that projects with large scope have on them. As a consequence, they might be less resistant. What is not clear, however, is why Framing leaders increased resistance in low scope situations. It might be an idea that recipients prefer direction in these situations because they do not think that careful consideration of a problem, which is emotionally and cognitively exhausting (Ford et al., 2008), is necessary in low scope change projects. If they can simply follow the rules of the Shaping leader, they do not have to put a lot of effort into the change, which does not affect their work life much anyway. Thus, if they are encouraged to take a more active role by a Framing leader, they might feel bothered by the emotional and cognitive work that this active role demands and then feel the need to resist. However, a more elaborate study is needed to analyze this relationship.
Another interesting finding during our explorative analysis was the main effect of age on recipient resistance. We found that the older recipients were more likely to resist than younger ones. This might be the case because younger people, who grew up with computers and
innovations on a daily basis might be more used to a constantly changing environment than older people, who grew up in a somewhat slower environment. It is likely that the change has a greater impact on the older generation than the younger one, which in turn makes them more likely to resist.
Theoretical Implications
30 taking a bilateral view. Another theoretical contribution is that we did not find the finer grained leader behaviors as described by Higgs and Rowland (2011). This might indicate that this categorization may not hold.
Practical Implications
Even though the fact that agents and recipients have different perceptions of the change process might not be new to practitioners, we do think that our study stresses the importance of this fact. Our results suggest that recipients grant agents more influence on resistance than agents thought. This important finding can be used to train agents to better communicate and also to be more aware of recipient reactions. If they understand the needs of their recipients better, they are more able to tailor their behavior to them.
Limitations and Suggestions for Future Research
Naturally, our study had a few notable limitations. First of all, our sample of 100 change projects was decent but a larger sample size might allow for more meaningful insights. Another limitation was the way our participants were sampled. Due to strict privacy rules in the
companies, the agents had to give their consent before we were allowed to approach the participants. As a consequence, the agents were the ones who chose the participants. It might have been the case that the agents chose those participants that were most likely to agree with them, which might have biased the results. Additionally, even though we ensured all participants that the questionnaire was confidential, the participants might have not believed that and
31 A fourth limitation is the way we measured resistance to change. We used only the
behavioral dimension of Oreg (2006). However, the full model has three dimensions: cognitive, affective, and behavioral. Due to the bilateral nature of our study, it was not possible to include all three dimensions, as it was not possible for agents to rate how recipients think or feel.
Naturally, they could only rate the behavior of others. Therefore, it is important to note that there is more to the concept of resistance than just the behavioral part. Additionally, the questions we asked to measure resistance were framed rather negatively. As was argued in the literature and also in this study, resistance has its positive sides as well. For example, it can make change leaders aware of possible flaws in their process or desired outcome (Ford et al., 2008). Questions that cater the positive side of resistance as well could have made a difference.
32 Another factor might be that recipients may have been frustrated by the situation before the change and are thus happy about the change. An example would be an old-fashioned
computer program that made collaboration with other departments in the company cumbersome. A new system like an ERP which helps connect the departments might therefore be relieving to the recipients. Thus, future research should include questions about satisfaction with the status before the change and also about the perceived usefulness of the change.
33 REFERENCES
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Sage.
Armenakis, A. A., & Bedeian, A. G. (1999). Organizational change: A review of theory and research in the 1990s. Journal of Management, 25, 293-315.
Armenakis, A. A., Harris, S. G., & Mossholder, K. (1993). Creating readiness for organizational change. Human Relations, 46, 681–703.
Avolio, B. J., Walumba, F. O., Weber, T. J. (2009). Leadership: Current theories, research and future directions. Annual Review of Psychology, 60, 421-449.
Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis. Klein, K.J., Kozlowski, S.W.J. (Eds), Multilevel theory,
research, and methods in organizations: Foundations, extensions, and new directions. San Francisco, CA, US: Jossey-Bass, pp. 349-381.
Boohene, R., & Williams, A. (2012). Resistance to organizational change: A case study of Oti Yeboah Complex Limited. International Business & Management, 4(1), 135-145. Bouckenooghe, D. (2010). Positioning change recipients’ attitudes toward change in the
organizational change literature. Journal of Applied Behavioral Science, 46(4), 500-531. Bouckenooghe, D., & Devos, G. (2008). Ready or not…? What’s the relevance of a meso level
approach to the study of readiness for change (Working Paper). Retrieved from University of Gent website: http://www.feb.ugent.be/nl/ondz/wp/Papers/wp_08_512.pdf
Bouckenooghe, D., Devos, G. & van den Broeck, H. (2009). Organizational Change
34 Burke, W.W. (2002), Organization Change: Theory and Practice. Foundation of Organization
Science, Sage Publications Series: London.
Burnes, B. (2014). Managing change: A strategic approach to organisational dynamics (6th ed.). Harlow: Pearson Education Limited.
Burnes, B., James, H. (1995). Culture, cognitive dissonance, and the management of change.
International Journal of Operations and Product Management, 15, 14– 33.
Conner, D. (1992). Managing at the speed of change. New York, NY: John Wiley.
Conner, D. (1998). Managing at the speed of change. How resilient managers succeed and
prosper where others fail. Chichester: Wiley.
Conner, D. (1999). Leading at the edge of chaos. New York, NY: John Wiley.
Del Val, M. P. Fuentes, C. M. (2003). Resistance to change: a literature review and empirical study. Management Decision, 41(2), 149-155.
Dent, E., & Goldberg, S. (1999). Challenging “resistance to change.” Journal of Applied
Behavioral Science, 35, 25–41.
Field, A. (2009). Discovering statistics using SPSS (3rd ed.). London: SAGE Publications Ltd. Ford J. D., Ford L. W. (1995). The role of conversations in producing intentional change in
organizations. Academy of Management Review, 20(3), 541–70.
Ford, J. D., & Ford, L. W. (2009). Decoding resistance to change: Strong leaders can hear and learn from their critics. Harvard Business Review, 87, 99 – 103.
Ford, J. D., Ford, L. W., & D’Amelio, A. (2008). Resistance to change: The rest of the story.
Academy of Management Review, 33(2), 362 - 377.
35 Hage, J. T. (1999), Organizational innovation and organizational change, Annual Review of
Sociology, 25, 597-622.
Hauenstein, N. M., & Foti, R. J. (1989). From laboratory to practice: Neglected issues in implementing frame-of-reference rater training. Personnel Psychology, 42, 359-378. Higgs, M. J., & Rowland, D. (2001). Developing change leaders: Assessing the impact of a
development programme. Journal of Change Management, 2, 47-64.
Higgs, M. J., & Rowland, D. (2005). All changes great and small: Exploring approaches to change and its leadership. Journal of Change Management, 5, 121-151.
Higgs, M. J., & Rowland, D. (2011). What does it take to implement change successfully? A study of the behaviors of successful change leaders. The Journal of Applied Behavioral
Science, 47(3), 309-335.
Higgs, M. J. (2003). Developments in leadership thinking. Leadership and Organization
Development Journal, 24, 273-284.
Holt, D. T., Armenakis, A., Harris, S. G., & Feild H. S. (2007). Toward a comprehensive definition of readiness for change: A review of research and instrumentation. In W. A. Pasmore & R. M.Woodman (Eds.), Research in organizational change and development (pp. 289–336). Oxford, England: Elsevier.
Judge, T. A., & Piccolo, R. F. (2004). Transformational and transactional leadership: A metaanalytic test of their relative validity. Journal of Applied Psychology, 88, 755-768. Kanter, R. M., Stein, B. A., & Jick, T. D. (1992). The challenge of organizational change: How
companies experience it and leaders guide it. New York: Free Press.
Kee, J.E. & Setzer, W. (2006). Change-centric leadership. Working paper. Washington DC: Center for Innovation in the Public Service, George Washington University.
36 King, N., & Anderson, N. (1995). Innovation and change in organizations. London: Routledge.
Klonek, F. E., Lehmann-Willenbrock, N., Kauffeld, S. (2014). Dynamics of Resistance to Change: A Sequential Analysis of Change Agents in Action. Journal of Change
Management, 14(3), 334-360.
Kotter, J. P. (1995). Leading change: Why transformation efforts fail. Harvard Business Review,
May-June, 11-16.
Kramer, R. J. (2005). Developing global leaders: Enhancing competencies and accelerating the
expatriate experience. New York, NY: The Conference Board.
Lewin, K. (1952). Group decision and social change. In G. E. Swanson, T. M. Newcombe, & E. L. Hartley (Eds.). Readings in social psychology (2nd ed.). 459-473. New York: Holt. Matsunaga, M. (2010). How to Factor-Anayze your Data Right: Do’s, Don’ts, and How-To’s.
International Journal of Psychological Research, 3(1), 97-110.
Maurer, R. (1997). Transforming resistance. HR Focus, 74(10), 9-10.
Michel, A., By, R. T., & Burnes, B. (2013). The limitations of dispositional resistance in relation to organizational change. Management Decision, 51(4), 761-780.
Nadler, D.A. (1993). Concepts for the management of organizational change. In Maybe, C, and Mayon-White, B. E (Eds), Managing Change, Paul Chapman, London, 85-98.
Oreg, S. (2006). Personality, Context, and Resistance to Organizational Change. European Journal of Work and Organizational Psychology, 15(1), 73-101.
Oreg, S., & Berson, Y. (2011). Leadership and employees’ reactions to change: the role of leaders’ personal attributes and transformational leadership style. Personnel Psychology,
37 Oreg, S., Vakola, M., & Armenakis, A. (2011). Change recipients’ reactions to organizational
change: A 60-year review of quantitative studies. The Journal of Applied Behavioral
Science, 47(4), 461-524.
Palmer, I., Dunford, R., & Akin, G. (2009). (2nd Ed.). Managing organizational change: A
multiple perspectives approach. Boston: McGraw-Hill.
Plowman, D. A., Baker, L. T., Beck, T. E., Kulkarni, M., Solansky, S. T., & Travis, D. V. (2007). Radical change accidentally: The emergence and amplification of small change. Academy
of Management Journal, 50, 515-543.
Piderit, S. K. (2000). Rethinking resistance and recognizing ambivalence: A multi-dimensional view of attitudes toward an organizational change. Academy of Management Review, 25, 783-794.
Rogers, E. M. (1962). Diffusion of innovations. New York: Free Press.
Saksvik, I., & Hetland, H. (2009). Exploring dispositional resistance to change. Journal of
Leadership & Organizational Studies, 16(2), 175-183.
Senge, P. M. (1997). Communities of leaders and learners. Harvard Business Review, 75, 30-31. Street, C. T., Gallupe, R. B. (2009). A proposal for operationalizing the pace and scope of
organizational change in management studies. Organizational Research Methods, 12, 720-737.
Thomas, J. B., Clark, S. M., & Gioia, D. A. (1993). Strategic sensemaking and organizational performance: Linkages among scanning, interpretation, actions, and outcomes. Academy of
Management Journal, 36, 239–270.
38 Vos, J. F. J., Brand, M. J. (2012). How do change agents in SMEs approach change? A
cross-sectional study into change projects within Dutch SMEs. Paper presented at the Euram Conference 2012, Rotterdam.
Waddell, D. & Sohal, A. S. (1998). Resistance: a constructive tool for change management.
Management Decision, 36(8), 543-548.
Weick, K. E., & Quinn, R. E. (1999). Organizational change and development. Annual Review of
Psychology, 50, 361-386.
Young, A.P. (2000). “I’m just me”: A study of managerial resistance. Journal of Organizational
39 APPENDIX
Appendix A – Additions to the Dataset
For this Master’s Thesis, the author was allowed to make use of an existing database to which she added cases. The following paragraph will provide an overview of the additional cases. The additional participants were seven change agents with 20 corresponding recipients. All projects were of different organizations and selection criteria were the same as for the existing dataset and as described in the main text. The group of change agents (N = 7) consisted of 86% males and 14% females, with a mean age of 37 (SD = 8.2). All agents were German. Regarding education, 57% of the agents reported having received higher vocational training, while 43% reported to have received ‘other’ education. The agents in our sample had an average experience of 3 years (SD = 2.8) in their current position and an overall work experience of 12 years (SD = 6.8).
The group of change recipients (N = 20) consisted of 60% males and 40% females, with a mean age of 37 (SD = 10.5). All recipients were German. The educational level ranged from higher vocational training or university (60%) to lower vocational training (5%), or other (10%). The recipients in our sample had an average experience of 7.4 years (SD = 6.8) in their current position and an overall work experience of 13.4 years (SD = 11).
40 Appendix B – Factor Analysis of Leader Behavior Scale
The initial scale consisted of a total of 40 items reflecting all five leadership styles:
Shaping (eight items, example: “I often clarified the direction of change.”), Attractor (eight
items, example: “I was very aware of my role as the change agent.”), Edge and Tension (nine items, example: “I did not shy away from difficulties.”), Container (eight items, example: “I set high goals.”), and Creator (seven items, example: “I organized discussions with the employees to come up with different solutions.”). The example questions given reflect the agent version of the questionnaire. The recipient version consisted of the same questions but posed from the
perspective of the recipient. An example of this would be “The change agent set high goals” instead of “I set high goals.” All answers were given on a Likert type scale ranging from 1 (“strongly disagree”) to 7 (“strongly agree”).
All 40 items were significantly correlated. The values of the anti-image correlation matrix were all above the minimum of .5 (Field, 2009), with the lowest value being .939. Bartlett’s test of sphericity was significant χ² (780) = 11365.426; p = .000, and in combination with the KMO value of .965, the sampling adequacy was verified (Field, 2009). After verifying that factor analysis is appropriate, we ran a principal component analysis with Oblimin rotation method. This analysis extracted five factors, which cumulatively explained 61.54% of the variance. Here, however, the pattern matrix could not be shown due to low convergence. Therefore, we were not able to validate the leader behavior scale concerning five distinct leader behaviors as suggested by Higgs and Rowland (2011).
41 .5/.2 rule. That is, a factor is retained only if its primary loading is higher than .5 and its second highest factor loading is smaller than .2. This way, as opposed to the traditional cut-off of .4, the larger pattern of the factor loadings is considered and the issue of cross-loadings is addressed (Matsunaga, 2010). Following this rule, we removed ten items3 (ET_7, Crea_1, Crea_7, Cont_7, Crea_2, ET_6, SB_4, SB_1, Cont_1, Crea_3). The analysis extracted the fixed three factors, cumulatively explaining 59.39% of the variance. The pattern matrix (Table A) suggests the existence of three distinct leader behavior styles. As the leader behavior scale distinguishing five different components was developed on the basis of the one with three different components (see theory section), we decided to revert to the theoretical origins of the scale, namely the three original leader behaviors: Shaping, Framing and Creating (Higgs, & Rowland, 2005).
Table A
Pattern matrix leader behavior scale
Item Factor loadings
1 2 3 22. SB_2 .542 23. SB_3 .661 24. SB_5 .645 25. SB_6 .761 26. SB_7 .813 27. SB_8 .769 28. ET_1 .750 29. ET_2 .696 30. ET_3 .497 31. ET_4 .604 32. ET_5 .464 33. ET_8 34. ET_9 .768 35. Crea_4 -.695
3 The variable names are the same names used in the initial dataset. This is to ensure that the assessors of this thesis
42 36. Crea_5 -.699 37. Crea_6 -.709 38. Attr_1 -.410 39. Attr_2 .484 40. Attr_3 -.407 41. Attr_4 42. Attr_5 .528 43. Attr_6 .442 44. Attr_7 .506 45. Attr_8 .565 46. Cont_2 .642 47. Cont_3 -.651 48. Cont_4 .829 49. Cont_5 .783 50. Cont_6 -.610 51. Cont_8 .539
43 Table B
Factor loadings of all survey items
Item Factor loadings
44
