1
CHANGE AGENTS’ NARCISSISM AND CHANGE
OUTCOMES
A bilateral study on how change agents’ narcissism, shared- and
traditional leadership style and Agent-Recipient Exchange influence
recipients’ reactions to organizational change
Master thesis MSc Business Administration, specialization: Change Management University of Groningen, Faculty of Economics and Business
Word count: 12.498 14 January 2016 Theun Leentjes Student number: s2034395 Peizerweg 85A 9726 JG Groningen Phone number: +31 (0)638230983 Email address: T.Leentjes@student.rug.nl
Supervisor:
Dr. Ileana Maris - de Bresser
2
Acknowledgements
3 Abstract
4
Table of content
1. Introduction 5
1.1. Setting the scene: Research background and theoretical gaps 5
1.2. Research objectives 6
1.3. Research questions 8
1.4. Expected theoretical contributions of this study 8
2. Literature review 8
2.1. Narcissism 8
2.2. Recipient reactions 9
2.2.1. Agents’ narcissistic personality and recipient reactions 10
2.3. Leadership style 10
2.3.1. Agent narcissism and leadership style 11
2.3.2. Agents’ leadership style and recipient reaction 12
2.4. Agent-Recipient Exchange (ARX) 13
2.4.1. ARX on the relation agents’ leadership styles and recipients’ reactions 13
3. Conceptual model 15 4. Methodology 15 4.1. Data collection 15 4.1.1. Questionnaire 15 4.1.2. Participants 16 4.2. Measurement 16 4.3. Independent variables 17
4.3.1. Narcissistic personality traits 17
4.4. Dependent variables 17
4.4.1 Change agent leadership style 17
4.4.2. Recipient resistance to the change 17
4.4.3. ARX-quality 18
4.5. Control variables 18
4.6. Validity 18
4.7. Analysis 19
5. Results 20
5.1. Combined data: agents and recipients 20
5.1.1. Hypothesis 1 22
5.1.2. Hypothesis 2 22
5.1.3. Hypothesis 3 22
5.1.1. Hypothesis 4 22
5.1.2. Hypothesis 5 23
5.2. Singular data: recipients 24
5.2.1. Testing the hypotheses 25
5.3. Singular data: agents 26
5.4. Supplementary analysis 27 5.5.1. Difference in perspectives 27 5.5.2. Consensus in perspectives 27 6. Discussion 29 6.1 Hypothesis one 29 6.2 Hypothesis two 29 6.3 Hypothesis three 30 6.4. Hypothesis four 30 6.5. Hypothesis five 30 6.6. Consensus in perspectives 30
6.7. Limitations and further research 31
6.8. Theoretical implications 32
6.9. Practical recommendations 33
7. Concluding remarks 33
8. References 35
5
1. Introduction
“It is not the most intellectual of the species that survives; it is not the strongest that survives; but the species that survives is the one that is able best to adapt and adjust to the changing environment in which it finds itself.”
-Charles Darwin- (Megginson, 1963, p.4)
1.1. Setting the scene: research background and theoretical gaps
As Charles Darwin already indicated, in order to survive in today’s fast moving environment it is necessary for organizations to constantly tailor to changes in the environment (Burnes, 2014). For this reason the concept of organizational change - which is defined as “planned alterations of organizational components to improve the effectiveness of the organization” (Cawsey, Deszca & Ingols, 2012, p.2) - is of growing interest for recent research. However, the conception that organizational change programs fails too often is supported by research many times (cf. Higgs & Rowland, 2011). Often it is claimed around 70% of the change initiatives fail (Beer & Nohria, 2000; Meaney & Pung, 2008).
Specific literature tried to find determinants for this major failure percentage. For example, one reason could be a shortcoming in the planning or implementation of the change process (Hoag, Ritschard & Cooper, 2002). Other writers point out the importance of change agents (i.e., individuals who lead organizational change efforts) for bringing the change to a successful end. Higgs and Rowland (2011) found that certain behaviors from the change agent can determine whether the change leads to success or to no success. This is in line with the Upper Echelon Theory stating that the characteristics of the leader can determine to what extent the organization will be successful (Hambrick & Mason, 1984). Due to these reasons it is useful to take the characteristics of the change agents (further in the text called: agent) into account when determining the effectiveness of the change.
6 Gentry, Campbell, Hoffman, Kuhnert and DeMarree (2008) argued that managers have a greater degree of narcissistic personality traits than non-managers.
The change management literature, though, is thin on emphasizing agents’ narcissistic personality and their effects. One research from Chatterjee and Hambrick (2007) suggests that companies led by CEO’s with high narcissistic personality traits tend to make more strategic changes within the organization. This is supported by research from Wallace and Baumeister (2002) stating that narcissistic CEO’s are taking bold or challenging actions because by doing so; they can attract attention to themselves, which they thrive on. Considering the high degree of narcissistic personality traits of management (Brunell et al., 2008) and their greater tendency to execute change more often (Chatterjee et al., 2007), these narcissistic characteristics might be important in explaining particular outcomes of organizational change. Still, not much is known about the effectiveness of the changes narcissistic agents create and what kind of leadership styles they execute.
1.2. Research objectives
Following the above mentioned interest in agents’ narcissism and its influence on the outcome of organizational change, the aim of this research is to determine why organizational change fails that often and what role agents' narcissistic personality traits plays in this. Besides this aim, the current study has three specific objectives to be met by the end of this research.
The first objective is obtaining a better understanding of the consideration regarding success or failure of organizational change. Several authors determined the success or failure of organizational change by measuring financial aspects (Mellert, Scherbaum, Oliveira & Wilke, 2015), customer satisfaction (Al-Haddad & Kotnour, 2015) and predetermined objectives (Quinn, 2004). These are important features in determining the success or failure of organizational change; however this research scrutinizes the importance of social aspects and is interested in the human side of organizational change. This is reinforced by Kotter and Cohen (2002) who endorse success and failures of organizational change are usually linked to human issues, rather than technical issues.
7 change -or not- gives useful insights in how to best lead a change (Armenakis & Harris, 2009). On the grounds that this study wants to find out why organizational change initiatives fail that often, examining recipients’ reactions in terms of resistance is one way of determining change ineffectiveness (e.g., van Dam, Oreg & Schyns, 2008). Peiperl (2005, p.348) construes the term resistance as: “...active or passive responses on the part of a person or group that militate against a particular change, a program of changes, or change in general”.
The second objective is to obtain a clear understanding of which leadership styles narcissistic agents execute. Several authors (e.g., Grant, 2012) found that the leadership style agents execute is an important factor responsible for the effectiveness of change. Two particular interesting leadership styles are of growing interest for the organizational literature and its effects on organizational outcomes: (1) traditional leadership approach; emphasizing on one single person who is nominated to lead the team, along with the relationship with his/her followers (cf. Nicolaides, LaPort, Chen, Tomassetti, Weis, Zaccaro & Cortina, 2014) and (2) shared leadership approach; considering a single leader is not able to perform all management functions because of an increasing density and complexity of the work (Pearce & Manz, 2005). Most of the research done with respect to the shared- or traditional leadership style scrutinized the effectiveness of these styles on the organization as a whole (e.g., Wood, 2005). However, in the change management literature empirical evidence of these styles and how they are related to the success or failures of the change is limited.
8
1.3. Research questions
One central question is fundamental throughout this research: ‘What influence does agents’ narcissistic personality traits have on the outcome of organizational change concerning recipients’ resistance to change?’ This research will find an answer to that by asking the following sub-questions:
1. What is the relation between agents’ narcissistic personality traits and recipients’ resistance?
2. Is agents’ leadership style a mediator for the relation between agents’ narcissistic personality traits and recipients’ resistance?
3. What is the moderating role of the degree of ARX on the relation between agents’ narcissistic personality traits and recipients’ resistance?
1.4. Expected theoretical contributions of this study
This study is striving to contribute to the change management literature by finding a reason for the major organizational change initiatives failure and introducing a relatively new concept in the change management literature: the degree of agents’ narcissistic personality traits. Another contribution of this study is that it takes a bilateral perspective regarding organizational change. Several studies took a unilateral perspective by holding a -change agent- leader-centric focus, or a -change recipient- employee focus (Armenakis et al., 2009). By having such a unilateral perspective, important contributions from one party will be neglected. Therefore, this research values both the agent and recipient perspectives. The research based on a bilateral perspective done by Ford et al. (2008) gives novel insights and forces us to revise prior assumed verities.
2. Literature review
2.1. Narcissism
9 Within the organizational psychological domain, however, narcissism is tested as an individual difference variable and with another instrument (Grijalva & Newman, 2015), what results in measuring narcissistic personality traits instead of a narcissistic personality disorder. Due to his or her self-centered, self-focused and self-serving behavior, in an organization a narcissistic person positively overestimates his or her own contribution and underestimates the contribution of others. Also, they will aggress to those who provide them with criticism (Brunell et al., 2008).
2.2. Recipient reactions
Despite the growing consensus that recipients’ reactions is the primary factor in determining the success or failure of any organizational change (Bartunek, Rousseau, Rudolph & DePalma, 2006), in prior research very little attention has been given to recipients’ reactions or attitudes concerning the change (Oreg et al., 2009). The meta-analysis of Bouckenooghe (2010) reveals that there are conceptually two attitudes towards change regarding change recipients: readiness for change and resistance to change. The most cited definition of readiness for change is provided by Armenakis, Harris and Mossholder (1993, p.681): “As organizational members’ beliefs, attitudes and intentions regarding the extent to which changes are needed and the organization’s capacity to successfully make those changes”. The other side of the coin is resistance; which is probably the best known attitude towards change (Bouckenooghe, 2010). Bouckenooghe’s (2010) meta-analysis found returning fundaments in the definition of resistance to change; an intentional or behavioral fundament as a driving force behind maintaining the status quo, and hindering successful implementation of change.
10 There has been portrayed a one-sided view about resistance, which actually says that it is ‘over there, in them’ (i.e., change recipients). However, Ford et al. (2008) argue that resistance is indeed partly caused by individual factors, but the influences from context are of greater importance. Confirmed by the literature review of Appelbaum, Degbe, MacDonald and Nguyen-Quang (2015) this indicates that the degree of resistance depends on how change agents behave.
2.2.1. Agents’ narcissistic personality and recipients’ resistance. As organizational
change brings a lot of confusion and anxiety beforehand, it is important that leaders can take the emotional related behaviors into account in order to reduce the effect of resistance (Battilana, Gilmartin, Sengul, Pache & Alexander, 2010). For example, leaders can recognize the emotional states of others (Huy, 1999), offer consideration and anticipate to emotional reactions from recipients (Oreg, 2003). Since narcissistic persons have a lack of empathy (American Psychiatric Association, 2013), it is expected that leaders with a narcissistic personality will experience more resistance from their recipients. This leads to the first hypothesis of this study:
Hypothesis 1: ‘Change agent’s narcissistic personality is positively associated with
recipients’ resistance.’
2.3. Leadership style
Historically, organizations had a hierarchical structure where the one at the top of the organization ‘led’ the company (Bass & Stogdill, 1990). This form of leadership, where the influence of the team members’ as well as their actions comes from one person; the leader itself (Nicolaides, et al., 2014), is called traditional leadership (a.k.a., hierarchical-; top-down-; vertical-top-down-; authoritarian or heroic- leadership). During the last decades, the notion that one single leader takes the leadership is increasingly less accepting (Ensley, Hmieleski & Pearce, 2006).
11 A fundamental distinction between shared and traditional leadership style is that the former is a “simultaneous, ongoing, mutual influential process” (Pearce, 2004, p.48) involving lateral, peer, up- or downward influences of team members (Conger & Pearce, 2003), while the latter involves only “a downward influence on subordinates by an appointed or elected leader” (Pearce, Manz & Sims, 2009, p.234).Differences are presented in Table 1.
Table 1
Differences between Shared Leadership and Traditional Forms of Leadership
Issues Related to
Leadership Style Shared Leadership Traditional Leadership Behavior expressed Aggregated behavior Singular or multiple behavior Type of structure Lateral and decentralized Hierarchical and centralized Actions of member Autonomous and self-led Dependent and instructed Actions of team Collaborative and
consensus-driven
Responsive to desires of leader
Derived from: Wood, 2005 (p.65)
Weick and Sutcliffe (2001, p.74) note: “With every problem, someone somewhere sees it coming. But those people tend to be low rank, invisible, unauthorized, reluctant to speak up”. In a context which accentuates shared leadership, members are free giving their concerns and are motivated to get involved in decision processes (Uhl-Bien, Marion & McKelvey, 2007). Considerable research developments found that an active involvement from multiple sources can create the possibility for high quality work efforts (Stewart, Courtright & Manz, 2011). It is therefore these leadership styles that could be important in explaining the success or failure of organizational change.
2.3.1. Agents’ narcissism and leadership style. In today’s change management
12 personality traits do have a more unrealistic view about them self, and are more focused on their ‘self’ than others (Mathieu & St-Jean, 2013).
As the traditional leadership style is typically a ‘leader-centric approach’ where the emphasis lies on the leader (Uhl-Bien et al., 2007), it sounds plausible that agents with more narcissistic personality traits would exercise the traditional leadership style more often. In addition to this, persons with narcissistic personality traits have a lack of empathy, cannot handle feedback (American Psychiatric Association, 2013) and are poor listeners (Vaknin, 2001). In a shared leadership environment, people work with each other and it is expected that everyone has their own responsibility and managers allowing them to exercise their skills (Wood, 2005). This leader obviously has to possess the skills of listening, be empathetic and is able to handle feedback, consolidating the argument why narcissistic agents place a greater emphasis on the traditional leadership style, rather than the shared leadership style. Maccoby (2004, p.71) straightens this argument: “Narcissistic leaders often say that they want teamwork. What that means in practice is that they want a group of yes-men”. This leads to the second hypothesis of this study:
Hypothesis 2: ‘Change agent’s narcissistic personality is positively associated with
executing a more traditional leadership style and negatively with executing a more shared leadership style.’
2.3.2. Agents’ leadership style and recipients’ reactions. Recipients undergoing
13
Hypothesis 3: Change agents’ shared leadership style is negatively associated with
recipients’ resistance and traditional leadership style is positively related to recipients’ resistance.
Hypothesis 4: The relation between agents’ narcissistic personality and recipients’
resistance is mediated through the leadership style of the agent.
2.4. Agent-Recipient Exchange (ARX)
In the organizational literature some particular research is done on Leader-Member Exchange (LMX). LMX can be characterized as the quality of the exchange relationship between the agent and the recipient (Schriesheim et al., 1999). The exchange relationship can be seen as a form of reciprocity, so a ‘give-and take’ relationship (Dienesch & Liden, 1986). Gouldner (1960, p.171) refers to the norm of reciprocity as: “(1) people should help those who have helped them, and (2) people should not injure those who have helped them” (cf. Bernerth, Armenakis, Feild, Giles & Walker, 2007). LMX is different from other leadership theories because of its focus on the dyadic relationship between member and leader (Gerstner & Day, 1997).
The change management literature emphasizes the importance of good relationships between agents and recipients (e.g., Ford et al., 2008); however, the concept of Leader-Member Exchange and its outcomes on organizational change has received little attention. Current research addresses Agent-Recipient Exchange (ARX) instead of LMX, due to its emphasis on a changing organization, not a ‘regular’ organization. A high degree of exchange component (i.e., high-quality ARX), indicates a strong tendency for agents to provide the recipient with clearly stated goals, feedback on their performance and better information about their role (cf. Geertshuis et al., 2015). Recipients believe they have been treated well by their agents, will reciprocate by helping both the organization- and the agent in return (Dulebohn, Bommer, Liden, Brouer & Ferris, 2012).
2.4.1. ARX on the relation between agents’ leadership styles and recipient reactions. Research on LMX found that it contributes to positive organizational outcomes
14 organizational change. Several reasons why ARX can moderate the relationship between agents’ leadership style and recipients’ resistance are mentioned.
First, high-quality ARX relationships will most likely be characterized by high levels of relationship commitment (i.e., long-term orientation, intent to persist and psychological contract). Hence, parties will be more likely to engage in the maintenance of relationship acts resulting in motivation (Thomas, Martin, Epitropaki, Guillaume & Lee, 2013). As in a shared leadership style, parties have to work with each other and there is a high form of interaction (cf. Coyle & Foti, 2015), it is important that there is a good relationship. Therefore, it is argued that a high-quality ARX relationship strengthens the negative relation between shared leadership and recipients’ resistance.
Second, the social exchange theory, which explains how “social exchange behaviors contribute to a business-to-business (B2B) exchange governance structure that is characterized as a form of relational exchange” (cf. Chang, Tsai, Chen, Huang & Tseng, 2015, p.868), acknowledges this exchange is an important factor in organizations. For instance, the social exchange theory advocates that when a perceived imbalance in the exchange is present, the trust of the employee decreases (Khazanchi & Masterson, 2011). Trust influences all kind of important properties in an organization, for example motivation (McEvily, Perrone & Zaheer, 2003). By the above mentioned reasons, the following hypothesis is formulated:
Hypothesis 5: Agent- Recipient Exchange (ARX) has a moderating role between
15
2. Conceptual model
Figure 1 below illustrates the conceptual model of this study.
Agent narcissistic personality Traditional leadership Framing change Agent leadership style
Resistance Recipient reaction Shared leadership High quality ARX + + -+ -+
-Figure 1. Conceptual model
4. Methodology
Current research follows the theory-testing approach: establishing a research gap in already existing theories and applying a conceptual model which is followed throughout the paper. Several steps are necessary: (1) indicate there is a literature gap and this gap can be tailored to business phenomena, (2) generate a conceptual model through recent- and of high quality research and develop specific hypotheses, (3) gather questionnaires and execute data collection, (4) analyze the data and (5) draw conclusions based on this data aligned with the developed hypothesis (van Aken, Berends & Van der Bij, 2012).
4.1. Data collection
4.1.1. Questionnaire. Online questionnaires via the program Qualtrics obtained the
16 combined, because the number of participants included in the analysis is higher and it is important that both questionnaires measure the same construct in order to compare them. This could be an issue, because of the difference in perspectives from the agent and the recipient. Therefore the hypotheses are tested again by separately applying the data from an agent and recipient perspective. Now two resources of data exist and this ensures the data to be even more valid.
4.1.2. Participants. Participants were selected by participating in free network
conferences, sending emails, promotion on the internet and conducting a snowball technique (i.e., pursuing references of references). To persuade, each participant had a chance to win a 20 euro-voucher from an online store. This could have result in the issue of attracting participants who entered only for the price and not taking it seriously. To solve this problem, participants who fulfilled the questionnaire for less than four minutes were deleted from the database.
A power analysis will predetermine the minimal sample size. Standardized and conventional use will set the alpha level at α=0.05 and the effect size at Cohen’s f2=0.15
(Cohen, Cohen, West & Aiken, 2003). The power is set at γ=0.80. This is deemed as acceptable and common usage (Cunningham & Gardner, 2007). With six numbers of predictors in set A (i.e., control variables) and two in set B as its maximum (two predictors including the moderator), the minimal sample size is set at 73.
Eventually, 124 employees (excluding four who fulfilled the questionnaire in less than four minutes) participated in the survey of which 48 were agents (38.7%) and 76 were recipients (61.3%). The agents consisted of 24 men (50.0%) and 24 women (50.0%) with a mean age of 45 (SD=13). The recipients consisted of 28 men (38.8%) and 48 women (63.2%) with a mean age of 40 (SD=15).
4.2. Measurement
17
4.3. Independent variables
4.3.1. Narcissistic personality traits (NNS). In order to measure the narcissistic
personality traits of the agent, fifteen items from the Narcissistic Personality Inventory (NPI) (Raskin & Hall, 1979) were used. This original questionnaire is a validated (self-report) measurement which ought to measure the extent of narcissistic personality traits and exist out of 40 statements where the agent has to indicate whether he/she agrees or disagrees with the statement. Barelds and Dijkstra (2010) argue that these questions can be measured with a 7-point likert scale having no apparent effect on the responses. To be consistent with the other questions, the questionnaire in this study also used a 7-points likert scale.
It should be noted the recipients’ questionnaire excluded cognitive and affective statements, because it is problematic for respondents to answer questions about the cognitions and emotions of someone else. The agent questionnaire contains all the original fifteen items. The higher the total scores on the statement which ought to measure narcissism, the higher the extend of narcissistic personality traits. An example of such statement is: “Modesty does not fit me/the leader” where the participant has to indicate to what extend this is true or not true on a 7–point Likert scale.
4.4. Dependent variables
4.4.1. Change agents’ leadership style (SHQ). Whether the change agents adhere to
a shared-leadership style or to a traditional-leadership style was measured by the eighteen questions of the Shared Leadership Perception Survey (Wood, 2005). The original eighteen items of these surveys measure the perception of shared leadership on a scale from 1 till 7. Three items are aimed at organizations in global and are removed. A low score emphasizes a more traditional leadership style and a high score emphasizes a more shared leadership style. An example question is: “There is a ‘pecking order’ within this leadership team.”
4.4.2. Recipient resistance to the change (RR). The degree of recipients’ resistance
18 his behavior on a 7-point likert scale. An example is: “During the change process... The employees sought ways to obstruct the change”.
4.4.3. Agent-Recipient Exchange (ARX). Eight items measured the quality of ARX
relationship based on the LMSX-scale of Bernerth et al., (2007). These items are adapted to a changing context and in form of a statement where the participant has to indicate whether he/she thinks this statement is referring to the truth on a 7-point likert scale. An example for such a statement is: “During the change process… I had a two-way exchange relationship with the employees”.
4.5. Control variables
Controlling for variables is incorporated because it reduces the effects of confounding variables. These variables could lead to the incorrect assumption that the dependent and independent variable are related to each other because the estimate fails to account for the confounding variable. The age, gender, level of education, tenure and role (i.e., agent or recipient) of the participant could serve as a compound variable and are controlled for. As well as how ground-breaking the change was.
4.6. Validity
Several predetermined actions were dedicated ensuring validity (i.e., measure what is intended to measure) and reliability (i.e., consistency in answer pattern): minimal sample size through power analysis, use of validated questionnaires and control variables, taking a bilateral perspective, anonymity of the test results, voluntary participation and executing a pilot study. Generalisability was guaranteed by asking participants from different companies and setting a minimal sample size through power analysis.
19 factor were removed and in this way validity is ensured. There are as many factors as variables; four in this study (narcissistic personality traits, shared- or hierarchical-leadership style, recipients’ resistance and ARX).
All items left have a factor-loading of more than .70 and no cross-loadings higher than .40; hereby the measurement is declared valid. Appendix 1 presents the factor analysis with the remaining items for the combined data of both agent and recipient, and the singular perspectives, respectively. This study is aware of the small sample size concerning agents (Nagent=48), however, different researchers (e.g., de Winter, Dodou & Wieringa, 2009; Jung &
Lee, 2011) stated PCA could be done with sample sizes below N=<50. Also, all the assumptions for performing a factor analysis are met (KMO=.79, Bartlett test=.000 in the combined version; KMO=.82, Bartlett test=.000 in the recipients version; KMO=.67, Bartlett
test=.000 in the agents version).
A Cronbach’s Alpha analysis is performed determining the reliability of the items. In Table 3 the Cronbach’s Alpha of the several variables are presented. Variables with a Cronbach's Alpha lower than <.70 are not used, because it is not certain whether the items measure the same construct. None of the variables have a Cronbach’s alpha lower than <.70 as Table 3 illustrates.
Table 3
Cronbach’s alpha: before and after removal of malfunctioning items detected by factor analysis
Questionnaire Used data Number of items questionnaire Number of items after removal Cronbach’s alpha before removal items Cronbach’s alpha after removal items
NNS Combined 8 3 .75 .81 Recipient 8 4 .75 .83 Agent 15 4 .75 .85 SHQ Combined 15 4 .87 .81 Recipient 15 2 .86 .75 Agent 15 5 .80 .83 RR Combined 4 4 .87 .87 Recipient 15 8 .91 .90 Agent 4 4 .92 .92 ARX Combined 8 5 .88 .89 Recipient 8 6 .91 .94 Agent 8 4 .74 .82 4.7. Analysis
20 correlations between the constructs. Then, the hypotheses are tested by (1) using the data of the recipient and the agent combined, (2) using the singular data of the recipient and (3) using the singular data of the agent executing a hierarchical multiple regression analysis. According to Lewis (2007) multiple regression analysis is appropriate for the social sciences, because constructs often have high correlations with each other.
In this research a hierarchical multiple regression analysis is performed using SPSS. In this analysis first (model 1), the control variables were added (i.e., age, gender, level of education, tenure, role, and groundbreaking), then (model 2) the main effect variables were added and at last (model 3) the whole model including the moderator (i.e., ARX). In this way, it can be detected how much the second model extra explains with respect to the previous model, and whether this difference is significant. After testing the fundaments (i.e., hypotheses) of this study, supplementary analysis inquiry interesting effects outside the hypotheses.
In multiple regressions analysis a problem that could arise is multicollinearity. This occurs when there are high correlations between predictor variables, leading to unreliable and unstable estimates of the regression coefficients. One way of measuring it is obtaining the variance inflation factor (VIF) in every analysis. If none of the predictors are correlated, the VIF will be equal to VIF=1. However, most of the time there will be a moderate relation between predictors, but should not be higher than VIF=<4.0. All the predictor variables had a multicollinearity lower than VIF=<4.0, which will be presented per analysis. It should be noted that multicollinearity among control variables are harmless (Voss, 2004). The assumptions homoscedasticity, linearity, independence of the data and normal distribution of the residual variances, are met for executing the analyses.
5. Results
5.1. Combined data: agents and recipients
Table 4 illustrates the descriptive results for the variables. It appears that the agents in this study (M=3.95, SD=1.30) significantly differ on their degree of narcissistic personality in comparison to the average score on the NPI-16 (M=2.45, SD=1.40) (see: Ames, Rose & Anderson, 2006) based on the ANOVA-analysis (F(1,899)=119.47, p=.000). It can be concluded that the agents have a higher extend of narcissistic personality traits than average.
21 personalities from agents. A Kurtosis between -2 and +2 (George & Mallery, 2010) and a Skewness between -1 and +1 (Bulmer, 1979) are considered acceptable in order to prove normal distribution. The variables were all normal distributed (Table 4).
Table 4
Descriptive results per variable.
Variable Mean SD Min Max Skewness Kurtosis
NNS 3.95 1.30 1.00 6.67 -.02 -.48
SHQ 4.65 1.22 1.75 7.00 -.58 -.51
RR 3.44 1.61 1.00 7.00 -.03 -.93
ARX 4.04 1.36 1.00 7.00 -.26 -.54
Notes. N=124; Scores are on a range from 1-7.
The correlation-coefficient matrix (Table 5) illustrates a positive relation between agents’ narcissistic personality and traditional leadership style (r=.20, p=.028) and a negative relation with ARX (r=-.20, p=.024). Also a strong correlation between SHQ and ARX is found (r=.47, p=.000). Noteworthy is the significant correlations from the role of the participant and the degree of narcissistic personality of the agent (r=-.27, p=.002), his/her leadership style (r=.40, p=.000) and the degree of ARX (r=-.21, p=.017) indicating a difference in perspective concerning these variables.
Table 5
Correlation coefficient matrix (Pearson’s r)
Variable 1 2 3 4 5 6 7 8 9 1. Agec 2. Genderac -.17 3. Educational levelc -.16 -.01 4. Tenurec -.92** -.12 -.11 5. Rolebc -.16 -.13 -.08 -.11 6. Groundbreakc -.07 -.07 -.13 -.06 -.13 7. NNS -.05 -.05 -.15 -.01 -.27** -.05 8. SHQ -.09 -.07 -.05 -.10 -.40** -.06 -.20* 9. RR -.11 -.05 -.01 -.14 -.15 -.10 -.10 -.12 10. ARX -.04 -.19* -.23** -.11 -.21* -.03 -.20* -.47** -.16 Notes. Ntotal=124
*.Correlation is significant at .05 level (2-tailed); **.Correlation is significant at .01 level (2-tailed).
22
5.1.1. Hypothesis 1
The first hypothesis tested is ‘Change agent’s narcissistic personality is positively
associated with recipients’ resistance’. The hierarchical multiple regression analysis showed
no significant relationship between agents’ narcissistic personality traits and recipients’ resistance (β=.16, R²∆=<.02, VIF=1.13, p=.091) and therefore no support is provided for hypothesis 1.
5.1.2. Hypothesis 2
Secondly, the hypothesis: ‘Change agent’s narcissistic personality is positively
associated with executing a more traditional leadership style and negatively with executing a more shared leadership style’ is tested. No significant relationship is found between change
agents’ narcissistic personality and adhering to a leadership style (β=-.09, R²∆=<.01,
VIF=1.13, p=.309), establishing no support for hypothesis 2.
5.1.3. Hypothesis 3
Next, the hypothesis ‘Change agents’ shared leadership style is negatively associated
with recipients’ resistance and traditional leadership style is positively related to recipients’ resistance’ is examined. Support is found that traditional leadership is positively related to
recipients’ resistance (and shared leadership style significantly negative related to recipients’ resistance) (β=-.21, R²∆= < .04, VIF=1.24, p=.035).
5.1.4. Hypothesis 4
Also the fourth hypothesis: ‘The relation between agents’ narcissistic personality and
recipients’ resistance is mediated through the leadership style of the agent’ is examined.
23 Table 6
Results (β) from the hierarchical multiple regression analysis testing the mediator. Dependent variable: recipient resistance
Model 1 Model 2 Model 3
Control variables Direct effects Mediation
β VIF β VIF β VIF
Age -.19 7.24 -.17 7.26 -.19 7.28 Gender -.06 1.06 -.08 1.07 -.10 1.08 Educational level -.01 1.08 -.03 1.11 -.03 1.11 Tenure -.30 6.92 -.28 6.95 -.27 6.95 Role -.14 1.07 -.19 1.16 -.27* 1.37 Groundbreaking -.07 1.05 -.08 1.05 -.08 1.05 NNS .16 1.13 -.14 1.14 SHQ -.20* 1.25 F-value 1.11 1.38 1.73 R2 .05 .08 .11
Notes. Ntotal=124; *.β is significant at p<.05 level (2-tailed); **.β is significant at p<.01 level (2-tailed).
5.1.3. Hypothesis 5
The last hypothesis tested is: ‘Agent- Recipient Exchange (ARX) has a moderating role
between leadership style and recipients’ reactions, in that way that ARX will strengthen the positive effect of shared leadership style on resistance and weakens the effect of traditional leadership style on resistance.’ As Table 7 below illustrates, no significant effect is found
24 Table 7
Results (β) from the hierarchical multiple regression analysis testing the moderator. Dependent variable: recipient resistance
Model 1 Model 2 Model 3
Control variables Direct effects Moderation
β VIF β VIF β VIF
Age -.19 7.24 -.12 7.78 -.09 7.83 Gender -.06 1.06 -.12 1.14 -.12 1.14 Educational level -.01 1.08 -.03 1.17 -.04 1.31 Tenure -.39 6.92 -.20 7.40 -.19 1.18 Role -.14 1.07 -.24* 1.30 -.23* 7.42 Groundbreaking -.07 1.05 -.07 1.06 -.08 1.06 SHQ -.14 1.53 -.17 1.56 ARX -.16 1.52 -.18 1.57 SHQ ARX (moderator) -.13 1.15 F-value 1.11 1.69 1.74 R2 .05 .11 .12
Notes. Ntotal=124; *.β is significant at p<.05 level (2-tailed); **.β is significant at p<.01 level (2-tailed).
5.2. Singular data: recipients
Also in this data, as Table 8 illustrates, the agents (M=4.39., SD=1.22) have significant higher narcissistic personality traits than average (M=2.45, SD=1.40) (Ames et al., 2006) based on a ANOVA analysis (F(1,851)=135.80, p=.000).
Table 8
Descriptive results per variable.
Variable Mean SD Min Max Skewness Kurtosis
NNS 4.39 1.22 1.25 6.75 -.36 -.33
SHQ 4.13 1.38 1.00 7.00 -.06 -.70
RR 3.50 1.43 1.00 6.88 -.00 -.83
ARX 3.84 1.48 1.00 7.00 -.11 -.89
Notes. Nrecipients=76; Scores are on a range from 1-7.
Table 9 below illustrates significant positive correlations for agents’ narcissistic personality (r=.36, p=.001) and agents’ traditional leadership style (r=-.25, p=.029) with recipients’ resistance (hypothesis one & three). No significant correlation is found between agents’ narcissistic personality and their leadership style (hypothesis two) (r=-.18, p=.113). Other interesting significant negative correlations are between the concepts ARX with agents’ narcissistic personality (r=-.28, p=.014) and ARX with recipients’ resistance (r=-.30,
p=.009). A strong significant positive correlation is found between ARX and shared
25 Table 9 Correlationcoëfficientmatrix Pearson’s (r) Variable 1 2 3 4 5 6 7 8 1. Agec 2. Genderac -.11 3. Educational levelc -.03 -.04 4. Tenurec -.95** -.07 -.10 5. Groundbreakc -.19 -.09 -.02 -.20 6. NNS -.07 -.02 -.01 -.01 -.10 7. SHQ -.00 -.18 -.10 -.03 -.08 -.18 8. RR -.20 -.11 -.02 -.24* -.24* -.36** -.25* 9. ARX -.15 -.30** -.24* -.21 -.06 -.28* -.41** -.30**
Notes. Nrecipients=76; *.Correlation is significant at .05 level (2-tailed); **.Correlation is significant at .01 level (2-tailed).
aGender: 1=men; 2=women; cControl variables
5.2.1. Testing the hypotheses. Agents’ narcissistic personality is significantly positive
related to recipients’ resistance (β=.39, R²∆=.14, VIF=1.07, p=.001), supporting hypothesis 1. Second (hypothesis 2), no association is found for change agents’ narcissistic personality adhering to a leadership style (β=-.19, R²∆=.03, VIF=1.07, p=.114). However, a significant negative relation is for hypothesis 3 (β=-.28, R²∆=.07, VIF=1.05, p=.016), which indicate that shared leadership is negatively related to recipients’ resistance.
Table 10 illustrates no mediation effect (hypothesis 4) from leadership style on the relation between agents’ narcissistic personality with recipients’ resistance.
Table 10
Results (β) from the hierarchical multiple regression analysis testing the mediator. Dependent variable: recipient resistance
Model 1 Model 2 Model 3
Control variables Direct effects Mediation
β VIF β VIF β VIF
Age -.29 10.44 -.01 11.07 -.04 11.14 Gender -.09 11.03 -.07 1.04 -.03 11.07 Educational level -.01 11.06 -.01 1.07 -.03 11.07 Tenure -.46 10.52 -.17 11.14 -.23 11.21 Groundbreaking -.19 11.05 -.23 1.06 -.25* 11.07 NNS -.39** 1.07 -.35** 11.11 SHQ -.21 11.09 F-value 1.69 3.76* 3.91** R2 0.11 0.25 0.29
26 No evidence is found for a moderating effect of ARX (hypothesis 5) on the relationship between leadership style and recipients’ resistance (β=-.10, R²∆=<.01, VIF=1.06,
p=.395), as is presented in Table 11.
Table 11
Results (β) from the hierarchical multiple regression analysis testing the moderator Dependent variable: recipient resistance
Model 1 Model 2 Model 3
Control variables Direct effects Moderation
β VIF β VIF β VIF
Age -.29 10.44 -.17 11.71 -.15 11.73 Gender -.09 11.03 -.00 1.17 -.08 11.17 Educational level -.01 11.06 -.06 1.19 -.06 11.19 Tenure -.46 10.52 -.32 12.37 -.32 12.37 Groundbreaking -.19 11.05 -.21 1.06 -.22 11.06 SHQ -.21 1.28 -.23 11.31 ARX -.17 1.63 -.17 11.63 SHQ ARX (moderator) -.11 11.04 F-value 1.69 2.39* 2.22* R2 .15 .23 .24
Notes. Nrecipient=76; *.β is significant at p<.05 level (2-tailed); **.β is significant at p<.01 level (2-tailed)
5.3. Singular data: agents
The descriptive results of agents’ narcissistic personality (Table 12) is higher (M=3.80,
SD=1.19) than the average score (M=2.45, SD=1.40) (Ames et al., 2006). Due the low sample
size of the agents (i.e., Nagent=48) no ANOVA- and multiple regression analysis will be
executed on this data.
Table 12
Descriptive results per variable
Variable Mean SD Min Max Skewness Kurtosis
NNS 3.80 1.19 1.00 6.50 -.23 -1.03
SHQ 5.18 1.08 2.00 7.00 -.93 -1.06
RR 3.73 1.79 1.00 6.75 -.27 -1.09
ARX 4.68 1.13 2.50 7.00 -.05 -.75
Notes. Nagent=48; Scores are on a range from 1-7.
Non-parametric technique such as Spearman’s (rho) correlation is used for
27 personality traits and recipients’ resistance (rho=-.25, p=.088). However, Figure 2a below indicates a slight negative trend concerning the relationship between agents’ narcissistic personality and recipients’ resistance. Also, hypothesis 2, no relationship between agents’ narcissistic personality traits and leadership style is found (rho=.05, p=.755).
Although not significant, a strong negative trend is visible (Figure 2e) between
traditional leadership style and recipients’ resistance, and shared leadership negatively related with recipients’ resistance (hypothesis 3) (rho=-.28, p=.051).
Table 13
Correlationcoëfficientmatrix Spearman’s (rho)
Variable 1 2 3 4 5 6 7 8 1. Agec 2. Genderac -.28 3. Educational levelc -.40** -.12 4. Tenurec -.86** -.17 -.13 5. Groundbreakc -.20 -.33* -.00 -.22 6. NNS -.09 -.16 -.15 -.03 -.08 7. SHQ -.21 -.04 -.18 -.22 -.04 -.05 8. RR -.08 -.21 -.23 -.01 -.03 -.25 -.28 9. ARX -.07 -.05 -.05 -.17 -.17 -.18 -.30* -.18
Notes. Nagent=48; *.Correlation is significant at 0.05 level (2-tailed); **.Correlation is significant at 0.01 level (2-tailed). aGender: 1=men; 2=women; cControl variables
5.4. Supplementary analysis
5.5.1. Difference in perspectives. Opposite perspectives are found on the relation
between agents’ narcissistic personality traits and ARX (Figure 2c below). The degree of agents’ narcissistic personality traits is negatively related to the degree of ARX; confirmed by regression analysis from the singular recipients data (β=-.22, R²∆=.043, VIF=1.11, p=.046). Although not significant from agent perspective (rho=.18, p=.225), a positive trend is distracted from Figure 2c.
5.5.2. Consensus in perspectives. Consensus between relationships is presented in
Figure 2d, 2e and 2f below. The relationship between ARX and recipients’ resistance is significantly negatively related based on the combined data (β=-.22, R²∆=.04, VIF=1.24,
p=.023) and singular recipient data (β=-.27, R²∆=.05, VIF=1.35, p=.039). No significant
relation is found based on the Spearman’s correlation (rho) regarding the data from the agent (rho=-.18, p=.225). But again, a negative trend is present.
Shared leadership and ARX are significantly positive related to each other based on the combined data (β=.45, R²∆=.16, VIF=1.28, p=.000), the recipients data (β=.37, R²∆=.13,
29
6. Discussion
This study makes a number of contributions through the analysis of why organizational change fails that often and what role agents' narcissistic personality traits plays in this based on a bilateral (from the agent and recipient) perspective. Several analyses respecting the hypotheses and interesting supplementary findings are discussed.
6.1 Hypothesis one
Controversy in perspectives among agents and recipients is found concerning the relation between agents’ narcissistic personality and recipients’ resistance. From a recipient’ perspectives, agents’ narcissistic personality is associated with recipients’ resistance (i.e., negative outcomes of agents’ narcissistic personality). A possible explanation is that an organization is a social system (Romme, 2003) and changing features of that organization requires dealing with other people. Feedback from recipients is needed to be aware of certain threats which could hinder change outcomes. Narcissistic persons are more focused on their self than others (Mathieu et al., 2013) and cannot handle feedback (Brunell et al., 2008). Therefore, that agents’ narcissistic personality leads to recipients’ resistance is likely concerning a recipient’ view.
Through the eyes of the agent, however, agents’ narcissistic personality shows a negative trend with recipients’ resistance (i.e., positive outcomes of agents’ narcissistic personality). A reason for this difference could be that narcissistic persons are more inclined to see their self in a positive light (Wallace et al., 2002) and think they did a good job; leading to the perception of low resistance.
6.2. Hypothesis two
30
6.3. Hypothesis three
No evidence is found that narcissistic agents’ adhere to either a shared- or traditional leadership style. One explanation for this could be that the leadership styles were defined and measured as a process (Pearce, 2004) or a team structure (Wood, 2005), not particularly as intensive actions from the leader. Because “structure only exists in and through the activities of human agents” (Giddens, 1989, p.256), the concept of how leadership styles were defined, emphasize its actions based on organizational members views rather than to focus on the intentions from the leader. So, it could be that a narcissistic agent desires a traditional leadership style, but cannot due to other contextual factors such as the acceptance of this leadership style from organizational members.
6.4. Hypothesis four
No support is found that agents’ leadership style mediate for the relation between agents’ narcissistic personality and recipients’ resistance. A mediation effect only occurs when the dependent variable (in this case agents’ narcissistic personality) is in relation with the mediator (Hayes, 2009) and in this study no support is found for the relationship between agents’ narcissistic personality and their leadership style. This could be a reason why no mediation effect is found.
6.5. Hypothesis five
No moderating effect from Agent-Recipient Exchange (ARX) on the relation between agents’ leadership style and recipients’ resistance is found. Specifically, this study found consensus in perspectives that ARX and shared leadership are highly related to each other. As shared leadership is emphasizing collaboration (Wood, 2005) and collaboration leads, obviously, to more interaction, the chances that a high-quality exchange relationship arise (i.e., ARX, see: Schriesheim et al., 1999) is likely to be present. The reason that shared leadership and ARX co-exist, could serve as reason why no moderating effect is found from ARX on the relation between shared leadership and recipients’ resistance.
6.6. Consensus in perspectives
31 relationship to occur. In the literature review, some reasons were mentioned for this relationship, namely the enhancement of trust and organizational citizenship behaviors (OCB). Another justification why a high-quality ARX relationship and shared leadership lead to less resistance is the agents’ ability to deal with emotional reactions. Resistance to change is partly attributable to the recipients’ emotional reactions, which emerge from threats to their self-esteem (Nadler, 1982), stress as a result of uncertainty (Olson & Tetrick, 1988) or typically anxiety and confusion (Kanter, 1984). High-quality ARX relationships and shared leadership leads to more interaction and, obviously, members are more inclined to speak up. Leaders are ultimately able to explain why the change is needed through which the communication is increased. As clear communication is typically profitable for change outcomes (Battilana et al., 2010), less recipients’ resistance is apparent.
6.7. Limitations and further research
The limitations of present study could serve as an entrance for further research. First, this study made no distinction between certain characteristics of the agent. The participant could choose whether her or her role was recipient or agent, however, there is also the possibility that someone is both. This is often the case in middle manager functions (Balogun & Johnson, 2005), where the manager (i.e., change agent) follow the reports of upper managers and also must lead the change. This may have created discrepancy choosing the role. Some middle managers may have chosen to see themselves as an agent and some saw themselves as a recipient. Future research must make the distinction between those function differences and determine its influence. Also, agents can work inside the organization (internal agents), or agents can be hired externally (external agents). These distinct agents could have different intentions, knowledge and perspectives, impacting the research findings. Future research has the possibility to make a distinction between internal and external agents.
The shared leadership measurement is seen as one of the limitations, as it measures how the current structure of the change ‘team’ is, and not particularly measuring the intentions from the agent. Future research can use or construct another instrument that measures to what extent the intentions of the leader are experienced as shared- or traditional leadership. Then again, it may be examined whether controversy or consensus between perspectives among agent and recipient exist.
32 compare these outcomes with outcomes based on other sources; for example the fulfilment of predetermined outcomes, financial aspects and customer satisfaction. This is in order to increase the validity of the results even more. Another aspect to increase validity of this research is to enhance the response rate. Due to the time constraint the response rate of the agents was quite low (48 participants). This study established highly valuable indications concerning difference in perspectives in determining the effectiveness of change outcomes and what influence agents’ narcissistic personality has in this. Future research could execute the same analyses with a higher response rate of agents.
At last, this study examined agents and recipients independent from each other. This decision is made because it is more assessable to quickly access a sufficient sample size. However, it may lead to loss of important information. Future research may examine the processes between colleague agent and recipient. Then it can be tested whether there are differences in perspectives concerning the effects of agents’ narcissistic personality traits. This means future research can test the effects of the (a)symmetry in ARX. For example, asymmetry occurs when a difference in perspective regarding the quality of the ARX exists.
6.8. Theoretical implications
This study is one of the first in the change management literature focusing on agents’ narcissistic personality and its effects on change outcomes. In accordance with previous studies that managers have a higher level of narcissistic personality (Brunell et al., 2008), this study found that change agents have a higher degree of narcissistic personality traits than average. Specifically, this study found that agents’ narcissistic personality is related to organizational change failure on accounts of a high degree of recipients’ resistance and from a recipients’ viewpoint. As recipients are the ones executing the change and determining whether the outcome of the change is (non-)successful, it is important to analyze their perspective, which is done in previous research as well (e.g., Oreg et al., 2009).
Particular research (Nevicka, Velden, De Hoogh & Van Vianen, 2011) found that narcissistic leaders and information exchange are negatively related to each other. This could be due to the fact that narcissistic leaders, agents in this particular context, and recipients do not have high-quality ARX relationship, what is found in this research.
33 concepts of prototypically (Wang et al., 2014) and taking in the emotions of others (Huy, 1999; Oreg, 2003).
Also low-quality ARX relationship has negative effects on change outcomes, which verifies the work of Van Dam et al. (2008). They noticed that this relationship was mediated by strong development climate, when received more information, and opportunities for participation. Members therefore perceived more trust in management, leading to less resistance (Van Dam et al., 2008). Based on these findings this study is emphasizing its importance of the change process regarding the reactions of recipients, as was demonstrated by previous research (e.g., Oreg, 2006).
6.9. Practical recommendations
Given the global question of why organizational change fails that often and what role agents' narcissistic personality traits plays in this, this study found that it depends on which perspective is relied on. The findings show that agents’ narcissistic personality traits are related to more recipients’ resistance from a recipients’ viewpoint. However, not known is why this is the case. The findings show, however, that the interaction (i.e., ARX) among agents and recipients is of utmost importance and, as this study shows, is not the case if a narcissistic agent is on the lead.
Due to the discrepancy in perspectives between the parties concerning the agents’ narcissistic personality, a second reason why it is important for a narcissistic agent to increase the quality of its ARX relationship is because the agents could be aware of the reactions from the recipients due to the high interaction. Specifically, practitioners must when leading the change, emphasize a highly interactive context where all the individuals have input in the decision. This is the case in a high-quality ARX and when executing a shared leadership style. According to the findings of this study, this leads to better outcomes in terms of low recipients’ resistance, which is a particular important contribution to practitioners in the field.
7. Concluding remarks
Ending up with the quote of Charles Darwin where this study started with: “It is not
the most intellectual of the species that survives; it is not the strongest that survives; but the species that survives is the one that is able best to adapt and adjust to the changing environment in which it finds itself” (Megginson, 1963, p.4). What should be kept in mind
35
8. References
Al-Haddad, S., & Kotnour, T. (2015). Integrating the organizational change literature: a model for successful change. Journal Of Organizational Change Management, 28(2), 234-262.
American Psychiatric Association (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.
Ames, D. R., Rose, P., & Anderson, C. P. (2006). The NPI-16 as a short measure of narcissism. Journal Of Research In Personality, 40(4), 440-450.
Appelbaum, S. H., Degbe, M. C., MacDonald, O., & Nguyen-Quang, T. (2015). Organizational outcomes of leadership style and resistance to change (Part Two). Industrial & Commercial Training, 47(3), 35-144.
Armenakis, A. A., & Harris, S. G. (2009). Reflections: our Journey in Organizational Change Research and Practice. Journal Of Change Management, 9(2), 127-142.
Armenakis, A. A., Harris, S. G., & Mossholder, K. W. (1993). Creating Readiness for Organizational Change. Human Relations, 46(6), 681-704.
Balogun, J., & Johnson, G. (2005). From intended strategies to unintended outcomes: The impact of change recipient sensemaking. Organization studies, 26(11), 1573-1601.
Bass, B. M., & Stogdill, R. M. (1990). Bass & Stogdill's handbook of leadership: Theory, research, and managerial applications. Simon and Schuster.
Battilana, J., Gilmartin, M., Sengul, M., Pache, A., & Alexander, J. A. (2010). Leadership competencies for implementing planned organizational change. Leadership
Quarterly, 21(3), 422-438.
Barelds, D. P., & Dijkstra, P. (2010). Narcissistic personality inventory: structure of the adapted dutch version. Scandinavian journal of psychology, 51(2), 132-138.
Bartunek, J. M., Rousseau, D. M., Rudolph, J. W., & DePalma, J. A. (2006). On the receiving end sensemaking, emotion, and assessments of an organizational change initiated by others. The Journal of applied behavioral science, 42(2), 182-206.
Beer, M., & Nohria, N. (2000). Cracking the code of change. If you read nothing else on
change, read thesebest-selling articles., 15.
Bernerth, J.B., Armenakis, A.A., Feild, H.S., Giles, W.F., & Walker, H.J. (2007). Leader-member social exchange (LMSX): development and validation of a scale. Journal of
36 Bouckenooghe, D. (2010). Positioning Change Recipients’ Attitudes Toward Change in the Organizational Change Literature. Journal Of Applied Behavioral Science, 46(4), 500-531.
Bulmer, M. G. (1979). Principles of Statistics. Edinburgh: Oliver & Boyd. Burnes, B. (2014). Managing change (6th ed.). Edinburgh: Pearson Education.
Brunell, A. B., Gentry, W. A., Campbell, W. K., Hoffman, B. J., Kuhnert, K. W., & DeMarree, K. G. (2008). Leader emergence: The case of the narcissistic leader.
Personality and Social Psychology Bulletin, 34(12), 1663-1676.
Campbell, W. K., Goodie, A. S., & Foster, J. D. (2004). Narcissism, confidence, and risk attitude. Journal Of Behavioral Decision Making, 17(4), 297-311.
Cawsey, T., Deszca, G., & Ingols, C. (2012). Organisational Change: An action-oriented
toolkit. Los Angeles: SAGE.
Chang, H. H., Tsai, Y., Chen, S., Huang, G., & Tseng, Y. H. (2015). Building long-term partnerships by certificate implementation: a social exchange theory perspective. Journal
Of Business & Industrial Marketing, 30(7), 867-879.
Chatterjee, A., & Hambrick, D. C. (2007). It's all about me: Narcissistic chief executive officers and their effects on company strategy and performance. Administrative Science
Quarterly, 52(3), 351-386.
Cohen, J., Cohen, P., West S.G., & Aiken, L.S. (2003). Applied multiple regression/
correlation analysis for the behavioural sciences. 3rd ed. New York: Routledge.
Conger, J. A., & Pearce, C. L. (2003). A landscape of opportunities. Shared leadership.
Reframing the hows and whys of leadership, 285-303.
Coyle, P. T., & Foti, R. (2015). If You’re Not With Me You’re...? Examining Prototypes and Cooperation in Leader–Follower Relationships. Journal Of Leadership & Organizational
Studies, 22(2), 161-174.
Cunningham, J. B., & Gardner, E. (2007). Power, effect and sample size using GPower: practical issues for researchers and members of research ethics committees. Evidence
Based Midwifery, 5, 132-136.
de Winter, J. F., Dodou, D., & Wieringa, P. A. (2009). Exploratory Factor Analysis With Small Sample Sizes. Multivariate Behavioral Research, 44(2), 147-181.
37 Dulebohn, J. H., Bommer, W. H., Liden, R. C., Brouer, R. L., & Ferris, G. R. (2012). A
Meta-Analysis of Antecedents and Consequences of Leader-Member Exchange: Integrating the Past With an Eye Toward the Future.Journal Of Management, 38(6), 1715-1759.
Dumaine, B. (1994). The Trouble with Teams. Fortune, 130(5), 86.
Ensley, M. D., Hmieleski, K. M., & Pearce, C. L. (2006). The importance of vertical and shared leadership within new venture top management teams: Implications for the performance of startups. The Leadership Quarterly, 17(3), 217-231.
Ford, J. D., Ford, L. W., & D'amilio, A. (2008). Resistance to change: the rest of the story. Academy Of Management Review,33(2), 362-377.
Geertshuis, S. A., Morrison, R. L., & Cooper-Thomas, H. D. (2015). It’s Not What You Say, It’s The Way That You Say It: The Mediating Effect of Upward Influencing Communications on the Relationship Between Leader-Member Exchange and Performance Ratings. Journal Of Business Communication, 52(2), 228-245.
Georgalis, J., Samaratunge, R., Kimberley, N., & Lu, Y. (2015). Change process characteristics and resistance to organisational change: The role of employee perceptions of justice. Australian Journal Of Management (Sage Publications Ltd.), 40(1), 89-113. George, D., & Mallery, P. (2010). SPSS for Windows Step by Step. A simple Guide and
Reference, Boston, MA. USA: Ed.
Gerstner, C. R., & Day, D. V. (1997). Meta-Analytic Review of Leader-Member Exchange Theory: Correlates and Construct Issues.Journal Of Applied Psychology, 82(6), 827-844. Gerstner, W., König, A., Enders, A., & Hambrick, D. C. (2013). CEO narcissism, audience
engagement, and organizational adoption of technological discontinuities. Administrative
Science Quarterly, 58(2), 257-291.
Giddens, A. (1989). A reply to my critics. D. Held, J. B. Thompson, eds. Social Theory of Modem Societies: Anthony Giddens and his Critics. Cambridge University Press,
Cambridge, U.K. 249-301.
Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. The leadership quarterly, 6(2), 219-247.
Grant, A. M. (2012). Leading with Meaning: Beneficiary Contact, Prosocial Impact, and the Performance Effects of Transformational Leadership. Academy Of Management
38 Grijalva, E., & Newman, D. A. (2015). Narcissism and counterproductive work behavior (CWB): Meta‐analysis and consideration of collectivist culture, Big Five personality, and narcissism's facet structure. Applied Psychology: An International Review, 64(1), 93-126. Hambrick, D. C., & Mason, P. A. (1984). Upper echelons: The organization as a reflection of
its top managers. The Academy Of Management Review,9(2), 193-206.
Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs, 76(4), 408-420.
Higgs, M., & Rowland, D. (2011). What Does It Take to Implement Change Successfully? A Study of the Behaviors of Successful Change Leaders.Journal Of Applied Behavioral
Science, 47(3), 309-335.
Hoag, B. G., Ritschard, H. V., & Cooper, C. L. (2002). Obstacles to effective organizational change: The underlying reasons. Leadership & Organization Development Journal,
23(1), 6-15.
Hogg, M. A., van Knippenberg, D., & Rast III, D. E. (2012). The social identity theory of leadership: Theoretical origins, research findings, and conceptual developments.
European Review of Social Psychology, 23(1), 258-304.
Hoch, J. E., & Dulebohn, J. H. (2013). Shared leadership in enterprise resource planning and human resource management system implementation. Human Resource Management
Review, 23(1), 114-125.
Huy, Q. N. (1999). Emotional capability, emotional intelligence, and radical change. Academy
Of Management Review, 24(2), 325-345.
Jung, S., & Lee, S. (2011). Exploratory factor analysis for small samples. Behavior research
methods, 43(3), 701-709.
Kanter, R. M. (1984). Change masters. Simon and Schuster.
Kendra, K. A., & Taplin, L. J. (2004). Change Agent Competencies for Information Technology Project Managers. Consulting Psychology Journal: Practice & Research,
56(1), 20-34.
Khazanchi, S. and Masterson, S.S. (2011), “Who and what is fair matters: a multi-foci social exchange model of creativity”,Journal of Organizational Behavior,32(1), 86-106.
Kotter, J., & Cohen, D. (2002). The Heart of Change. Boston, MA: Harvard Business School Press.
Maccoby, M. (2004). Narcissistic Leaders: The Incredible Pros, the Inevitable Cons. Harvard
