The impact of different leadership behaviors on the
withdrawal of employees: A meta-analytic investigation
Wietske K. van der Kamp
University of Amsterdam
The Netherlands
Bachelor thesis Business Studies Student number: 10281819 Supervisor: C.K. Buengeler
STATEMENT OF ORIGINALITY
This document is written by Wietske Katharine van der Kamp, who declares to take full responsibility for the content of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used creating it. The Faculty of Economics and Business is responsible solely for the supervision of
ABSTRACT
Interest in employee withdrawal arises because employee withdrawal is detrimental, costly, and dangerous to an organization’s sustainable success. Therefore, organizations want to prevent this phenomenon. The main goal of this research is to discover to what extent leadership can help prevent withdrawal and which leadership behaviors are particularly effective in this regard. Using a systematic literature search, relationships between leadership behaviors and employee withdrawal were found. This research provides a meta-analysis of the relationships between leadership behaviors and employee withdrawal. Results (based on 53 correlations from 31 sources) revealed that person-focused leadership (r = -.16, p < .01) has a more negative relationship with employee withdrawal than task-focused leadership behaviors. The relationship between task-focused leadership and withdrawal was also negative, but non-significant. The sub-clusters, traditional task-oriented leadership, contingent punishment of transactional leadership, and contingent reward of transactional leadership, explained the variance in the task-focused leadership behavior cluster. This research contributes to the leadership literature by focusing on objective withdrawal indicators and by taking into account all leadership behaviors that have a direct relationship to employee withdrawal found in the literature. The knowledge of which kinds of leadership behaviors are (in)effective can help organizations manage employee withdrawal and reduce the cost associated with withdrawal.
Keywords: task-focused leadership behaviors, person-focused leadership behaviors, employee withdrawal
Table of contents
INTRODUCTION 4
CONCEPTUAL FRAMEWORK 5
METHOD AND RESEARCH DESIGN 9
LITERATURE SEARCH AND STUDY SOURCES 9
CRITERIA FOR INCLUSION 11
CODING OF RELEVANT INFORMATION 11
META-ANALYTIC PROCEDURES 12
ALLOCATION LEADERSHIP BEHAVIORS 13
RESULTS 14
VARIANCE IN THE TASK-FOCUSED LEADERSHIP-WITHDRAWAL RELATIONSHIP 16
DISCUSSION 17
THEORETICAL IMPLICATIONS 20
PRACTICAL IMPLICATIONS 22
LIMITATIONS AND FUTURE RESEARCH 23
CONCLUSION 23
REFERENCES 25
Introduction
Society needs effective leaders no less than society needs effective followers (Malakyan, 2014). Leadership and followership are both essential components of effective leadership influence. Nowadays, employees have more rights and freedoms than ever before in history. Employees can refuse to follow leaders who do not meet the requirements of a good leader.
It is vital for an organization that the relationship between leader and employee is positive and that the employee is satisfied with his or her leader (Malakyan, 2014). If this relationship is not sufficient, employees might withdraw. In this research, the withdrawal behavior of employees refers to voluntary absenteeism and voluntary turnover.
The interest in the withdrawal behaviors of employees occurs, because employee turnover and employee absenteeism are detrimental and costly. Mowday, Porter, and Steels (2013) stated that employee absenteeism has become a serious and costly problem for organizations. Recruiting, selecting, and training new employees to cover production deficiencies, human capital development, and employee well-being are expensive and can endanger an organization’s sustainable success (Tse,
Huang & Lam, 2013). Because withdrawal of employees reduces revenue and increases expenses (Tracey & Hinkin, 2008), organizations want to prevent employee withdrawal. In this research, the impact of different leadership behaviors on the withdrawal behavior of employees will be examined. This research contributes to the leadership literature by focusing on objective withdrawal indicators and by taking into account all leadership behaviors that have a direct relationship to employee withdrawal found in the literature. Previous research mainly focused on withdrawal intentions and the effects of one or a few leadership behaviors, while this research focuses on the actual withdrawal and takes into account as many leadership behaviors as possible. This research also has important practical implications. The knowledge of which kind of leadership behaviors are (in)effective can help organizations manage employee turnover and absenteeism. If organizations can gain a better hold on leadership behaviors that have consequences for the withdrawal behaviors of employees, efforts can be made to adapt leadership behavior and reduce withdrawal.
Withdrawal as a process needs a focus on job behaviors, primarily leadership behavior (Krackhardt, McKenna, Porter, & Steers, 1981).
According to academic and practitioner turnover experts, leadership behavior affects employees in voluntary turnover decisions and voluntary absenteeism (Maertz & Griffeth, 2004; Schreuder et al., 2011).
Previous research has shown that the way leadership is exercised and the relationship between leader and member are essential for the satisfaction of employees and, therefore, the withdrawal process (Tse et al., 2013). Certain leadership behaviors may cause withdrawal of employees and some leadership behaviors may prevent withdrawal of employees, but it is still unclear which leadership behaviors specifically cause these processes. The main goal of this research is to discover to what extent leadership can help prevent withdrawal and which leadership behaviors are particularly effective in this regard. This research will advance understanding of the effects of different leadership behaviors on employee withdrawal.
First, the different forms of leadership that may influence employee withdrawal will be explained by means of a conceptual framework, followed by a delineation of the study hypotheses. Second, the methodology and research design will provide a detailed explanation of how the literature search was performed and which literature sources were used.
Criteria for inclusion of studies, relevant information coded, and meta-analytic procedures will be explained. The methodology and research design will also include the allocation of leadership behaviors, which were found in the literature, into clusters. Third, the results will be presented. These will be summed up in the discussion and followed by the limitations. This research will end with the conclusion.
Conceptual framework
This research’s main goal is to discover if leadership can reduce or prevent employee withdrawal and which leadership behaviors are particularly effective in this regard. Why some leadership behaviors are effective and some are not can partly be explained by job satisfaction. Elshout, Scherp, and van der Feltz-Cornelis (2013) explained, in their article, that since there is a close relationship between employee job satisfaction and absenteeism, there might be a relationship between leadership behavior and absenteeism as well. Mardanov, Heischmidt, and Henson (2008) supported this thought from the turnover perspective, since they stated that a significant cause of job dissatisfaction is leadership behavior. They also said the higher the degree of job dissatisfaction, the higher the turnover rate. Frooman,
Mendelson, and Kevin Murphy (2012) also found that style of leadership influences absenteeism through job satisfaction. Consistent with this, Griffeth, Hom, and Gaertner (2000) explained that turnover is triggered by job dissatisfaction. They also stated that a factor of job satisfaction is the relationship employees have with their supervisors. This information will be taken into account while delineating the hypotheses.
In this research, the leadership behaviors are clustered following Burke et al.’s (2006) classification. Burke et al. (2006) created two clusters for the leadership behaviors: Task-focused leadership and person-focused leadership.
The first cluster, task-focused leadership, contains three categories of leadership behavior: Transactional, initiating structure, and boundary spanning (Burke et al., 2006). By linking job performance to rewards and by ensuring employees have all the resources they need to complete a job, transactional leadership can help organizations achieve their current goals more efficiently (Zhu, Chew & Spangler, 2005). According to Fleishman (1973), initiating structure is the degree to which a leader defines and organizes his or her role and the roles of followers, is oriented toward goal attainment, and establishes patterns and channels for communication. Structuring
leaders are oriented toward tasks (Bass & Stogdill, 1990). Boundary spanning was described by Ernst and Yip (2009) as leadership that bridges boundaries between groups in service of a larger organizational vision, mission, or goal.
Given that all three types of leadership are characterized by task accomplishment, they can be subsumed under task-focused leadership.
Transactional leadership consists of three dimensions: Contingent reward, active management by exception, and passive management by exception. Transactional leadership focuses on exchanges in which the leader praises, rewards, or withholds a punishment from an employee who meets expectations (Burke et al., 2006). A transactional leader depends on the environment in which he or she operates, to be a leader with whom employees are satisfied (Hamstra, Van Yperen, Wisse & Sassenberg, 2011). Overall, transactional leadership may not be the most employee-interacting leadership behavior, but it does encourage employees to fulfill their tasks because transactional leaders give clarity about standards and rules. This leadership behavior may be successful for those who prefer this behavior, which can be reflected in less employee withdrawal (Hamstra et al. 2011).
directive leadership and autocratic leadership (Pearce & Conger, 2003). Initiating structure behavior emphasizes the completion of task goals by minimizing role ambiguity and conflict (Burke et al., 2006). According to Burke et al. (2006), initiating structure sets clear goals and gives good directions for a certain goal, which can be a driving force for employees.
Boundary spanning implies scanning of the environment and collaborating with other people outside the team (Burke et al., 2006). Through boundary spanning, employees can be empowered at all levels and can develop cross-organizational learning abilities (Lee, Horth & Ernst, 2012). This empowerment can be a reason for enhanced job satisfaction and thereby less withdrawal.
Overall, task-focused leadership behavior mainly concerns task accomplishment (Burke et al., 2006). If employees perceive task-focused leadership behavior as positive, this perception can have a positive influence and decrease the withdrawal of employees (Minavand, Mokhtari, Zakerian & Pahlevan, 2013). By combining the knowledge of the mostly positive effects of task-focused leadership behavior, the following hypothesis was derived:
Hypothesis 1:
The relationship between task-focused leadership behavior and employee withdrawal is negative.
The second cluster, person-focused leadership, contains four categories of leadership behavior: Transformational, consideration, empowerment, and motivational (Burke et al., 2006).
A common theme among person-focused leadership behaviors is that the leader acts in a way that builds employee respect and encourages employees to focus on the well being of the group (DeRue, Nahrgang & Wellman, 2011).
Transformational leadership behavior is described as a creative and meaningful interaction between leaders and employees, with the goal to bring about vision driven change in employees (Burke et al., 2006). As described by Bass (1999), “the leader moves the follower beyond immediate self-interest through idealized influence (charisma), inspiration, intellectual stimulation, or individualized consideration” (p. 11). Previous studies have suggested that transformational leadership has a negative relationship with employee turnover intention (Tse et al., 2013). Frooman et al. (2012) also concluded that transformational leadership could help reduce voluntary absenteeism due to the positive effect transformational leadership has on job satisfaction.
The characteristics of consideration as leadership behavior are mutual open communication, respect, trust, and the fulfillment of employee needs (Burke et al., 2006). Leaders who score highly on consideration show respect and concern for individual group members, are approachable and friendly, are open to the ideas of others, and treat all employees as equals (Bass & Stogdill, 1990). Pool (1997) stated that there is a positive relationship between consideration leadership behavior and employee job satisfaction, which might mean that a smaller number of employees will withdraw.
Pearce and Conger (2003) explained empowerment as leadership behavior, emphasizing the development of follower self-management or self-leadership skills. With this self-leadership behavior, employees will be actively involved in business processes. With empowerment, employees might experience more commitment to an organization and will be less inclined to leave the organization (Albrecht & Andreetta, 2011).
Motivational leadership behavior refers to behavior that promotes continuous efforts by employees, especially in difficult times. It includes rewards and recognition for achievements, but it also ensures that the values and
needs of employees are met by providing support (Burke et al., 2006). The combination of reward and support can be satisfying for employees and with that employee withdrawal might be reduced.
Overall, person-focused leadership facilitates team interaction and/or development (Burke et al, 2006). Graen Liden and Hoel (1982) stated that members tend to remain in the organization when they see themselves actively exchanging support, effort, resources, etc. with their leaders. By combining information and findings of previous studies about person-focused leadership behaviors and withdrawal of employees, the following hypothesis was derived:
Hypothesis 2:
The relationship between person-focused leadership behavior and employee withdrawal is negative.
Finally, the last hypothesis will focus on the differential effects of the two clusters, person-focused and task-focused leadership behavior. Hypotheses 1 and 2 state that task-focused and person-focused leadership behaviors each have negative relationships with employee withdrawal. This research will examine which of the two clusters, defined by Burke et al. (2006), is more effective in decreasing the level of turnover and absence of employees.
Minavand et al. (2013) concluded, in their research, that employees who are led by leaders with person-focused leadership behaviors have higher levels of job satisfaction and a lower tendency to leave their jobs than employees that are led by leaders with task-focused leadership behaviors. Also, in the presence of leaders with person-focused leadership behaviors, employees reported fewer days of absence. Due to the appearance of collective identities that person-focused leadership entails, employees feel more committed to an organization and will have a lower tendency to withdrawal. This result was not observed for leaders who performed task-focused leadership (Mellor, Arnold, & Gelade, 2009). It is important that the relationship between leader and follower functions properly; otherwise, it can cause employees to lose commitment to or satisfaction with their jobs, which can eventually lead to employee withdrawal. Taking this fact into account, Hypothesis 3 was derived as follows:
Hypothesis 3:
The relationship between person-focused leadership behavior and employee withdrawal is more negative than the relationship between task-focused leadership behavior and employee withdrawal.
Method and research design
Literature search and study sources
This research involved a systematic search and review of the existing literature linking leadership and withdrawal, which means the relevant literature for this research was systematically collected and combined. Systematic techniques were used to increase the probability that all relevant studies were identified, and to accurately gather information (Cooper, 2009). The information flow through the different phases of the research is shown in Figure 1. The literature collection process will be explained using this figure.
Figure 1. Flow of information through the different phases of a systematic review (Moher, Liberati & Altman, 2009, p. 267)
For a systematic literature search, it is highly important that every activity is noted. In the first phase of the systematic review, the literature was extracted from two databases supported by the University of Amsterdam, PsycINFO and Business Source Premier.
In PsycINFO a multi-field search was used to look for a combination of the following key terms:
• Leadership and withdrawal (8) • Leadership and turnover (85) • Leadership and absence (40) • Leadership and absenteeism (7) The number of hits per key term are shown in parentheses. The search in the database was limited to “keywords in abstract,” so the hits would be more specific and useful. Because this research was conducted with reference to an already existing database, the timeframe of the multi-field searches was limited to January 1, 2014 to the present.
At Business Source Premier, an advanced search was used to look for a combination of the same key terms:
• Leadership and withdrawal (1) • Leadership and turnover (38) • Leadership and absence (23) • Leadership and absenteeism (3) The number of hits per key term are shown in parentheses. Here also, the search in the database was limited to “keywords in
abstract”. Again, this search timeframe was limited to January 1, 2014 to the present, because of the existing database. The source type was set on “academic journals’’ because, for this research, academic journals were the relevant source. These searches combined resulted in 205 articles.
In addition to the representative literature search, which consisted of the multi-database search, this research involved a quality control. This quality control consisted of searching in a composite list of relevant, peer-reviewed journals to ensure the articles’ quality. To obtain as many relevant articles as possible, a manual search of each issue, from 2000 to the present, of relevant academic journals known to publish leadership research (Journal of Applied
Psychology, Leadership Quarterly,
Journal of Management, Journal of
Organizational Behavior, European
Journal of Work and Organizational Psychology, Academy of Management Journal, Organizational Behavior and Human Decision Processes, Journal of
Occupational and Organizational
Psychology, Group & Organization
Management, Management Science,
Organization Studies and Applied
Psychology: A review) was conducted. The
search was adjusted for articles written in the last 15 years to include papers from a
reasonably long time range. Nine relevant articles were found through this additional search.
All the articles found were imported into Mendeley, a desktop and web program that can manage and share research papers. From there, the articles were imported into MS Excel. In the screening and the eligibility phases (Figure 1), the duplicates were removed and by looking at the title and the abstract, the relevant literature was selected. The collected articles were added to the existing database. The articles from this existing database could not all be used, so these articles were also sorted and checked for relevance. The articles in the existing database were used as a comparison to ensure that the coding was completed correctly. The coding and documenting of the relevant data was completed using SPSS.
If articles were not accessible, an email was sent to the first author, kindly asking for permission to obtain the article. In the last phase, the included phase (Figure 1), it was clear how many studies would be included in the systematic literature review. With this information, a basic meta-analytic examination of the relationship between leadership behaviors in relation to withdrawal was performed. A meta-analysis is a statistical technique used to combine the results and findings of studies (Cooper, 2009). With these
outcomes, it would become clear whether the hypotheses were supported.
Criteria for Inclusion
For a study to be included in this research, five criteria had to be met. First, the study had to include a measure of leadership and a measure of employee withdrawal. Second, the study needed to be about actual withdrawal behavior. Withdrawal intention was excluded from this research. Third, the study had to contain statistical results expressed in a Pearson’s product-moment correlation coefficient or an effect size, which could be converted into a Pearson’s product-moment correlation coefficient. Fourth, there had to be a direct relationship between the leadership behavior and the withdrawal behavior. This direct correlation had to be present for the meta-analysis. Fifth, the studies were, most preferably, of reasonable quality. Therefore, most studies in the database were extracted from reputable journals. This research yielded 31 studies that contained 53 correlations between some form of leadership behavior and employee withdrawal behavior.
Coding of relevant information
After the criteria for inclusion were met, each study was assessed and coded for independent variables, dependent variables, moderators, and mediators
related to leadership and withdrawal. There was also interest in some additional variables that could influence the relationship between leadership and withdrawal, so these measurements and correlations were also coded. In some cases, odd ratios were presented instead of correlations. Odd ratios were transformed into correlations using the following formula:
The source of measurements was also interesting due to evidence suggesting that information obtained from the same source may differ from information obtained from different sources (Fuller, Patterson, Hester, & Stringer, 1996; Judge & Piccolo, 2004; Lowe, Kroeck, & Sivasubramaniam, 1996). Study characteristics, sample characteristics, and other relevant characteristics of the studies were also listed in the coding scheme. These are represented in appendix 1. The size of the sample was incorporated as the sample used to compute the correlation coefficient in the respective primary study. The majority of the studies were conducted at the individual level. In the case of the individual level, the sample size was registered as the total number of employees in the study. When a study was conducted at the team or organizational level, sample size reflected the number of
units on the respective level. The leadership behaviors were measured by the leaders themselves or by their employees. In the coding scheme, a distinction was made as to which of these measured the leadership behavior.
Meta-analytic procedures
The meta-analytic methods of Hunter and Schmidt (1990) were used to quantitatively summarize previous research findings between leadership behaviors and employee withdrawal. The average correlation across individual studies was computed by weighing the sample size of the study.
Confidence intervals of 95% were computed. A confidence interval consists of a lower and an upper bound. The width of the confidence interval indicates the precision (Akobeng, 2005). Confidence intervals reflect variability in the estimated mean correlation, which indicates the probable amount of error in the estimated mean correlation due to sampling (Woo, Chernyshenko, Stark, & Conz, 2013).
If the Q statistic between leadership behaviors and withdrawal was significant, it meant there was a significant variability in the effects. A moderator analysis had to explain this variance.
For the moderator analysis, the studies were divided into categories. Separate meta-analyses were conducted to
reveal the possible effects of moderators. Taking Judge and Piccolo (2004) as an example, Q statistics were used to test for the presence of moderator effects. A significant Q statistic indicates the probability that moderators explain variability in the correlations across the studies.
Allocation leadership behaviors
During the coding process, several leadership behaviors were identified. All these different leadership behaviors were placed in the two clusters, according to Burke et al. (2006). Additionally, there were some leadership behaviors that could not be placed in the clusters task-focused leadership and person-focused leadership. Evidence shows that abusive, coercive, and autocratic leadership behaviors have negative relationships with positive outcomes. Placing these leadership behaviors in the task-focused cluster would confound the effect (Chi & Liang, 2013; Nyberg, Westerlund, Hanson, & Theorell, 2008; Sheridan & Vredenburgh, 1978). Therefore, another cluster was made called “abusive, coercive, and autocratic leadership.” This cluster includes abusive leadership, coercive power, autocratic leadership, and leadership X. Leadership X is explained as authoritarian behavior based on the assumptions that the power of the leader comes from the position he or
she occupies and that people are basically lazy and unreliable (Carter, 2010).
Additionally, for the meta-analysis, a few leadership behaviors were excluded for several reasons. First, Average Leadership Style (ALS) was excluded because this leadership behavior was used as an averaged value across a unit in the studies measuring dyadic leadership style (LMX). Therefore, ALS was not specified at the same level as LMX. To avoid relationships being coded twice, the focus was on the individual-level relationship.
Second, effective leadership behavior was excluded because this leadership behavior could not be assigned to the three clusters visible in Tables 2, 3, and 4. To avoid the meta-analysis was run with a cluster containing only one correlation this leadership behavior was excluded.
Table 1. Excluded leadership behaviors Effective leadership
Average leadership style (ALS)
In the following tables, the allocation of the leadership behaviors used for the meta-analysis is presented.
Table 2. Task-focused leadership behaviors
Production-oriented leadership Transactional leadership Reward power Legitimate power Expert power Passive-avoidant leadership Initiating structure
Active management by exception
Table 3. Person-focused leadership
behaviors
Empowering leadership Dyadic leadership (LMX) Transformational leadership
Attentive-managerial leadership (AML) Transformational-reward leadership Employee-oriented leadership Supportive leadership Change-oriented leadership Supportive leadership Empowering leadership Ethical leadership Leadership Y Referent power Relation-oriented leadership Inspirational leadership Consideration Referent power
Table 4. Abusive, coercive, and
autocratic leadership behaviors
Abusive leadership Coercive power
Autocratic leadership Leadership X
To explain the variability of the task-focused leadership cluster, three sub-clusters were created:
• Traditional task-oriented leadership • Contingent punishment of
transactional leadership
• Contingent reward of transactional leadership
The first sub-cluster can be described as typical traditional task-oriented leadership behavior. This sub-cluster contains performance-oriented leadership, production-oriented leadership, and initiating structure. The second sub-cluster contains leadership behaviors that are described in the studies as slightly more negative forms of task-focused leadership behavior. This sub-cluster contains passive-avoidant leadership and active management by exception. The third sub-cluster can be described as the positive forms of task-focused leadership behavior. This sub-cluster contains transactional leadership, reward power, legitimate power, and expert power.
Results
The results of the overall meta-analysis, the relationship between leadership behavior and withdrawal, and
the results of the moderator analysis of the clusters are provided in Table 5. The meta-analysis results show that leadership overall has a significant negative relationship with withdrawal (r = -.09, p < .01), meaning if leadership increases, withdrawal decreases. The Q statistics (Q[52] = 1017.46, p < .01) between overall leadership and withdrawal were highly significant, which indicates there was a significant amount of variance in the correlations across studies. Therefore, it was tested if a distinction should be made between the withdrawal types (i.e., turnover, absenteeism, job withdrawal, and work withdrawal). Results revealed that type of withdrawal did not moderate the relationship between leadership and withdrawal because the Q-statistics were not significant (Q[52] = 7.96, p > .05), so no distinction had to be made between types of withdrawal.
To explain the variance, Burke’s clusters were tested as a moderator. The Q statistics between these clusters were significant, so leadership clusters, as a moderator, do matter (Q[52] = 28.54, p <
.01). The significant Q-statistic between the moderator clusters became non-significant within the clusters person-focused and abusive leadership behavior, which indicates that this moderator explains a significant amount of the variability in these clusters.
Hypothesis 1 suggested that the relationship between task-focused leadership behavior and employee withdrawal is negative. The results showed that task-focused leadership had a negative relationship with employee withdrawal (r = -.02, p > .05). However, Hypothesis 1 is not supported because the results showed that task-focused leadership is not significant. Task-focused leadership had marginally significant Q statistics, suggesting the presence of moderators across these studies. This variance will be explained later.
Hypothesis 2 predicted a negative relationship between person-focused leadership and employee withdrawal. The results showed that person-focused leadership had a negative relationship with employee withdrawal (r = -.16, p < .01).
Table 5. Relationships of leadership behaviors with employee withdrawal
k N p r 95% CI
Leadership 53 33999 < .01 -.09 -.1443, -.0368 Task-focused 13 2883 > .05 -.02 -.1032, .0652 Person-focused 34 25903 < .01 -.16 -.2102, -.1097 Abusive 6 5213 < .01 .17 .0495, .2843 Note. k = number of correlations; N= sample size; p= p-value; r =mean correlation; CI= confidence interval
Thus, Hypothesis 2 is supported. The Q statistics were not significant, which indicates homogeneity of correlations in this leadership cluster.
Although not hypothesized, the extra cluster, abusive, coercive, and autocratic leadership, was examined. Abusive, coercive, and autocratic leadership had a positive significant relationship with withdrawal (r = .17, p < .01), meaning that in the presence of abusive, coercive, and autocratic leadership, employee withdrawal increases. Hypothesis 3 stated that the relationship between person-focused leadership behaviors and employee withdrawal is more negative than the relationship between task-focused leadership behaviors and employee withdrawal. As shown in Table 5, the relationship between person-focused leadership behaviors and employee withdrawal (r = -.16, p < .01) is more negative than the relationship between task-focused leadership behaviors and employee withdrawal (r = -.02, p > .05). The withdrawal relationship is only significant for person-focused leadership, not for task-focused leadership. Therefore, Hypothesis 3 is supported.
Variance in the task-focused leadership-withdrawal relationship
The Q-statistics of the relationship between task-focused leadership and withdrawal were significant, which indicates there was a significant amount of variance in the correlations across studies (Q[12] = 77.99, p < .01). To explain the variance between task-focused leadership and employee withdrawal, the cluster task-focused leadership was divided in three sub-clusters: Task-focused leadership, contingent punishment of transactional leadership, and contingent reward of transactional leadership. These sub-clusters were used in the moderator analysis. The significant Q-statistics between the moderator clusters became non-significant within these clusters, which indicates that this moderator explains a significant amount of the variability (Q[12] = 17, p < .01). The results showed that the sub-cluster traditional task-focused leadership had a marginally significant positive relationship with employee withdrawal (r = .10, p < .10). Sub-cluster two (contingent punishment of transactional leadership) had a significant positive relationship with employee withdrawal (r = .22, p < .05). Sub-cluster three (contingent reward of transactional leadership) had a significant negative relationship with employee withdrawal (r = -.14, p < .01). The
variance in task-focused leadership was explained by the sub-clusters.
More tests were conducted to find other moderators that could have influenced the relationship between leadership behaviors and employee withdrawal. A meta-analysis was conducted to discover whether the different branches would have a moderating effect on the relationship between leadership behaviors and withdrawal.
The Q statistics of this test were significant (Q[47] = 25.89, p < .01), which means branches do moderate the relationship between leadership and withdrawal. The relationship between overall leadership behavior and withdrawal in different branches was significantly negative (r = -.08, p < .01). Most notably, in business and industry, there was a significant negative relationship (r = -.16,
p < .05), meaning, in general, leadership
has a negative effect on withdrawal in business and industry. The business and industry branch refers to private manufacturing firms or private service firms. When examining the clusters individually, the Q-statistics of branches as a moderator between task-focused leadership and withdrawal were significant (Q[12] = 7.65, p < .05). The results showed task- leadership had a negative effect on withdrawal in the business and industry branch (r = -.17, p < .05). The Q-statistics
of branches as a moderator between person-focused leadership and withdrawal were also significant (Q[30] = 11.94, p < .05). Person-focused leadership behaviors also had a negative relationship with withdrawal in the business and industry branch (r = -.15, p < .01). In the health sector, person-focused leadership also seemed to prevent withdrawal (r = -.17, p < .01). Public and private companies, as a moderator on the relationship between leadership and withdrawal, were tested too. The Q-statistics of this test were non-significant (Q[44] = 2.27, p > .05), which indicates that this moderator does not explain a significant amount of the variability. No further research was conducted due to this fact.
Additionally, it was tested if there was a difference between cross-sectional and longitudinal studies, but the results did not seem significant (Q[52]= .07 p > .05), so there was also no further research needed.
Discussion
This research was conducted to determine to which extent leadership can help prevent withdrawal and which leadership behaviors would be particularly effective in this regard. The results revealed there is a significant negative relationship between leadership and employee withdrawal. After clustering the
leadership behaviors, results showed that task- focused leadership behaviors have a non-significant negative relationship with withdrawal, and person-focused leadership behaviors have a significant negative relationship with employee withdrawal, while abusive, coercive, and autocratic leadership behaviors have a significant positive relationship with withdrawal. With this information, Hypothesis 1 was not supported because of the non-significance, contrary to Hypothesis 2, which was supported. The results also showed that person-focused leadership had a more negative relationship than task- focused leadership to employee withdrawal, which means person-focused leadership is considered to have a stronger effect on employee withdrawal. Thus, Hypothesis 3 was also supported.
In the cluster task-focused leadership the variance was explained by the creation of sub-clusters as a moderator. Because the literature used in this research represented different forms of task-focused leadership, which had a negative or positive relationships with withdrawal, the variance in this cluster was affected. The results showed contingent reward of transactional leadership and contingent punishments of transactional leadership have different relationships to employee withdrawal. Contingent reward of transactional leadership was seen as
effective in preventing employees from withdrawal, while contingent punishment of transactional leadership seems to encourage withdrawal.
Statistics showed task-focused leadership had significant negative relationships with withdrawal in the business and industry branch. Additionally, person-focused leadership had a significant negative relationship with withdrawal in the health sector and the business and industry branch. Summarized, both task-focused leadership and person-task-focused leadership behaviors can be effective when it comes to reducing withdrawal. Person-focused leadership had a more positive effect on employee withdrawal, as it had a more significant negative relationship than task-focused leadership to employee withdrawal and had a significant effect in more branches. However, results showed that task-focused leadership did have a more negative effect than person-focused leadership in the business and industry branch.
It can be discussed why person-focused leadership had a negative relationship to withdrawal in health care. Garman and Tran (2006) explained, in their article, that health care functions in a socioeconomic environment and can be called a relation business. Communication with personnel and adaptability to certain situations is vital. As stated by Burke et al.
(2006), person-focused leadership facilitates team interaction and/or development. For a leader to be effective in this sector, he or she needs to be more relation-focused, which explains why person-focused leadership had a negative relationship to withdrawal in the health care sector. Due to the socioeconomic environment, leaders and employees must work closely together, so a leader that focuses only on tasks does not create a positive relationship with an employee. As the results showed, this relationship is different in the business and industry branch, where person-focused and task-focused leadership both had negative relationships with employee withdrawal. This work environment differs considerable from the health care sector. Bass (1986) explained that the business and industry sector has usually discouraged charismatic leadership. For leaders in this branch, it is important to use their positions as a method to make sure employees meet the job requirements by handing out rewards or punishments. However, Bass (1986) stated that fully task-focused leadership has limits when it comes to making employees feel confident and valued in the environment in which they work in, which can be further increased by person-focused leadership behaviors. Through this behavior, leaders can become inspirations to employees.
Hereby, employees might make extraordinary efforts and feel connected to their work environment by the creation of trust and confidence (Bass, 1986). Albrecht and Andreetta (2011) showed that if employees feel more committed to an organization, they are less inclined to leave the organization.
Task-focused and person-focused leadership are completely different but can both have positive influences on employees. When a company is highly focused on achieving tasks, and employees are satisfied with these terms, task-focused leadership can be extremely effective. If employees prefer to have better relationships with their leaders, a leader should exhibit person-focused leadership behavior.
According to Elshout et al. (2013), there is no single leadership behavior that is appropriate in all situations. An effective leader is one who can adapt his or her leadership behavior to the readiness level of employees. Readiness is the ability and motivation of employees for a given task (Silverthorne, 2000). Also, other studies clearly showed there is no single leadership behavior that is universally successful (Hersey & Blanchard, 1969). In addition to Korman (1966) and Fiedler and Chemers (1967), other investigators have shown that different situations require different leadership behaviors (Hersey &
Blanchard, 1969). In short, for a leader to be effective, he or she should have characteristics of task-focused leadership behaviors and person-focused leadership behaviors. Elshout et al. (2013) concluded, from their research, that effective leadership reduces absence. Sellgren, Ekvall and Tomsen (2007) clarified that effective leadership can also reduce employee turnover. In summary, effective leadership is not fully task-focused or fully person-focused. It depends on the situation and what kind of leadership is desired. However, person-focused leadership is predominant in effective leadership and will, therefore, be more effective in decreasing the level of turnover and absence of employees.
Theoretical implications
The most important discussion point is the classification of the clusters. For this research, the leadership behaviors were clustered according to Burke et al. (2006) into person-focused and task-focused leadership behaviors. DeRue et al. (2011) divided person-focused leadership in relation-focused leadership and change-focused leadership. DeRue et al. (2011) decided to make this division because he thought relation-focused leadership behavior is mainly focused on the interaction between leader and employee. The leaders want to build follower respect
and encourage employees to think about the well-being of the team, while change-focused leadership behavior focuses more on development and the communication of a vision for change. Here, risk taking and innovative thinking is encouraged (Yukl, Gordon, & Taber, 2002). Relation-focused leadership behavior is supporting and helping. Change-focused leadership behavior is encouraging and facilitating. This cluster, on its own, could be effective because change-focused leadership behavior is encouraging and facilitating, and in addition to having a good relationship with employees, it also focuses on the performance of employees (DeRue et al., 2001). This leadership behavior cluster does have it own features and might even be more effective than relation-oriented leadership behavior.
To test if this clustering would have been a better alternative, an additional analysis was conducted where change-focused leadership was removed from the person-focused cluster and added to the two clusters of Burke et al. (2006). The Q-statistics were significant (Q[52] = 29.98, p < .01). Results showed that the change-focused cluster had a significant negative relationship with employee withdrawal (r = -.18, p < .05). The correlation of change-focused leadership and withdrawal was even more negative than person-focused leadership and withdrawal.
Change-focused leadership behaviors have their own features and the results showed change-focused leadership can function on its own as a separate cluster.
Why active management by exception was assigned to the task- focused cluster should also be discussed. Active management by exception was originally not intended to be a negative form of task-focused leadership, as it is part of the full-range of leadership theory and assumed to be part of effective leadership as well. In contrast, Yukl (1999) concluded, from his study, that active management by exception is not related to leader effectiveness. Northouse (2015) stated that active management by exception uses more negative reinforcement patterns than other task-focused leadership behaviors. Therefore, active management by exception might have positive effects on withdrawal. Active management by exception was also coded with a positive relationship to withdrawal (r = .26). Still, this leadership behavior was measured with the Multifactor Leadership
Questionnaire (MLQ), so it had to be assigned to task-focused leadership. An additional meta–analysis was completed without active management by exception. The mean correlation of task-focused leadership and withdrawal was somewhat more negative, but the relationship was still not significant (r = -.04, p > .05).
Another discussion point is the sample sizes used because, in this research, some sample sizes of studies were used multiple times. Hereby, it was necessary to conduct an extra check to see if the research was still reliable if only one correlation per study was used. If a study contained several correlations, the correlations that had a smaller number in the leadership clusters were retained, which means many person-focused leadership behaviors were excluded from the extra check. This was done so there would be a better balance in correlations per leadership cluster. In Table 6, the statistics of this additional analysis are displayed. The Q-statistics of this test were significant (Q[31] = 274.03, p < .01).
Table 6. Extra Check: Relationships of leadership behaviors with employee withdrawal
k N p r 95% CI
Leadership 32 23620 < .05 -.05 -.1003, -.0024 Task-focused 5 1540 < .05 .13 .0257, .2394 Person-focused 22 18983 < .01 -.13 -.1739, -.0840 Abusive 5 3097 < .01 .14 .0481, .2518 Note. k = number of correlations; N= sample size; p= p-value; r= mean correlation; CI= confidence interval
There were some differences in the results, but the mean correlation (r = -.05, p < .05) of leadership and employee withdrawal was still marginally significant. The mean correlations of the person-focused leadership and abusive leadership clusters did not differ much from the original meta-analysis in the extra check. The only large difference was in the mean correlation of task-focused leadership, probably because there was originally much variance in the task- focused cluster, so by excluding some studies, the mean correlation achieved a noticeably different value.
Additionally, there was one mediator that could not be ignored. The meta-analysis showed there were direct linkages between leadership and withdrawal. It is also important how leadership influences withdrawal. The literature revealed that job dissatisfaction is a strong indicator of employee withdrawal. Previous research also revealed that leadership behavior has influence on job satisfaction. In this research, this mediator also stood out. This relationship will only be described. The correlations clearly showed that job satisfaction and person-focused leadership behaviors always had a strong positive correlation, which means that in the presence of person-focused leadership, job satisfaction increased. Job satisfaction and withdrawal in general had a negative relationship, which indicates if
job satisfaction decreased, employee withdrawal increased.
Practical implications
This research confirmed that leadership behavior does have an impact on employee withdrawal. Person-focused leadership behavior is predominant in an effective leader in regard to employee withdrawal. Organizations can utilize these findings with the aim of decreasing employee withdrawal and its associated costs. The relationship between employee and leader was proven to be important.
This research showed that person-focused leadership behavior might help prevent withdrawal in the health sector. Close communication is needed in this sector due to the socioeconomic environment, so it is important for health care organizations that leaders exercise person-focused leadership behavior. For business and industry, effective leadership type depends on the situation. Person-focused and task-Person-focused leadership behaviors both can be effective regarding employee withdrawal, but the combination of both might be most effective.
Taking this information into account, organizations might be able to decrease the cost of recruiting, selecting, and training new employees by making sure employees are satisfied with their
leader and thus remain with the organization.
Limitations and future research
First, during the literature search it became clear that many of articles about the leadership-withdrawal link focused on turnover intentions instead of actual turnover behavior. Several studies reported that turnover intention is only moderately linked to actual turnover. Broach and Dollar (2007) concluded that intent to leave is not a good indicator of actual turnover. Cohen, Blake, and Goodman (2015) confirmed this finding this by arguing that turnover intention rate is not significantly associated with the actual turnover rate. This research only focused on objective withdrawal indicators, so a large amount of articles were excluded. Not much research was conducted about the relationship between leadership behaviors and the actual withdrawal of employees. Even though this is a strong point of this research, the numbers of articles that could be included were reduced.
Second, some articles contained the actual withdrawal process but did not contain a direct correlation between leadership behavior and employee withdrawal. In these articles correlations from leadership behavior to turnover intention and from withdrawal intention to
withdrawal were included, but a direct correlation was not. These articles had to be excluded.
Third, as mentioned before in the discussion, an additional analysis was completed to test DeRue et al.’s (2011) leadership clusters. The cluster change-focused leadership had a significant negative mean correlation, so for future research, this cluster should be considered an independent cluster.
Fourth, even though a meta-analysis allows one to minimize certain biases, these biases cannot be fully excluded. While reading articles, it seemed some authors had a negative or positive bias about task-focused or person-focused leadership behaviors. In advance, a distinction was made between the clusters, reflecting person-focused leadership behaviors as more positive and task-focused leadership behaviors as more negative.
Conclusion
The idea of this research was to investigate to what degree leadership can help prevent withdrawal and which leadership behaviors are especially effective in accomplishing this.
Overall, it seemed that person-focused leadership behavior had a more constant negative effect on withdrawal, so person-focused leadership behavior should
be most present in the character of a leader to decrease the amount of employee withdrawal. The results showed that both person and task-focused leadership behaviors have positive features. There is no leadership behavior that is completely effective. Effective leaders should have traits of both a task and person-focused leader.
This research improved understanding of the effects of different leadership behaviors on employee withdrawal. The results can be of value for organizations that want to reduce or prevent withdrawal of employees by changing leadership behavior.
This research contributed to the leadership literature and has results that can be important to organizations, but effectiveness of leadership behaviors in regard to employee withdrawal remains an interesting field of research where much still needs to be unraveled.
References
-Akobeng, A. K. (2005).
Understanding systematic reviews and meta-analysis.Archives of Disease in
Childhood, 90(8), 845-848.
-Albrecht, S. L., & Andreetta, M. (2011). The influence of empowering leadership, empowerment and engagement on affective commitment and turnover intentions in community health service workers: Test of a model. Leadership in
Health Services, 24(3), 228-237.
-Bass, B. M., & Stogdill, R. M. (1990). Handbook of leadership (Vol. 11). New York: Free Press.
-Bass, B. M. (1999). Two decades of research and development in transformational leadership. European
journal of work and organizational psychology, 8(1), 9-32.
- Bass, B. M. (1986). Leadership: Good, better, best. Organizational
dynamics,13(3), 26-40.
-Broach, D., & Dollar, C. (2007). Comparison of Intent-to-Leave With Actual Turnover Within the FAA.
-Cooper, H. (2009). Research
synthesis and meta-analysis: A step-by-step approach (Vol. 2). Sage Publications.
-Burke, C. S., Stagl, K. C., Klein, C., Goodwin, G. F., Salas, E., & Halpin, S. M. (2006). What type of leadership behaviors are functional in teams? A meta-analysis. The leadership quarterly, 17(3), 288-307.
-Carter, J. (2010). A
CORRELATIONAL STUDY OF
PRINCIPALS’LEADERSHIP STYLE AND
TEACHER ABSENTEEISM (Doctoral
dissertation, The University of Alabama TUSCALOOSA).
-Chi, S. C. S., & Liang, S. G. (2013). When do subordinates' emotion-regulation strategies matter? Abusive supervision, subordinates' emotional exhaustion, and work withdrawal. The Leadership
Quarterly, 24(1), 125-137.
-Cohen, G., Blake, R.S., Goodman, D. (2015). Does turnover intention matter? Evaluating the usefulness of turnover intention rate as a predictor of actual turnover rate. Review of Public Personnel
Administration, 1-24.
-DeRue, D. S., Nahrgang, J. D., Wellman, N. E. D., & Humphrey, S. E.
(2011). Trait and behavioral theories of leadership: An integration and meta-‐‑ analytic test of their relative validity. Personnel Psychology, 64(1), 7-52.
-Elshout, R., Scherp, E., & van der Feltz-Cornelis, C. M. (2013). Understanding the link between leadership style, employee satisfaction, and absenteeism: a mixed methods design study in a mental health care institution.Neuropsychiatric disease and
treatment, 9, 823.
-Ernst, C., & Yip, J. (2009). Boundary spanning leadership: Tactics to bridge social identity groups in organizations.
-Fiedler, F. E., & Chemers, M. M. (1967). A theory of leadership effectiveness. Korman, A. K. (1966). “Consideration,”“Initiating Structure,” and Organizational Criteria-‐‑a Review. Personnel Psychology, 19(4),
349-361.
-Fleishman, E. A. (1973). Twenty years of consideration and structure. In E. A. Fleishman & J. G. Hunt (Eds.), Current developments in the study of leadership.
Southern Illinois University Press, 1–40.
-Frooman, J., Mendelson, M. B., & Kevin Murphy, J. (2012). Transformational and passive avoidant leadership as determinants of absenteeism. Leadership & Organization
Development Journal, 33(5), 447-463.
-Fuller, J. B., Patterson, C. E., Hester, K., & Stringer, D. Y. (1996). A quantitative review of research on charismatic leadership. Psychological
reports, 78(1), 271-287.
- Garman, A. N., & Tran, L. (2006). Knowledge of the Healthcare Environment. Journal of Healthcare
Management, 51(3), 152.
-Graen, G. B., Liden, R. C., & Hoel, W. (1982). Role of leadership in the employee withdrawal process. Journal of
Applied Psychology, 67(6), 868.
-Griffeth, R. W., Hom, P. W., & Gaertner, S. (2000). A meta-analysis of antecedents and correlates of employee turnover: Update, moderator tests, and research implications for the next millennium. Journal of management, 26(3), 463-488.
-Hamstra, M. R., Van Yperen, N. W., Wisse, B., & Sassenberg, K. (2011).
Transformational-transactional leadership styles and followers’ regulatory focus: Fit reduces followers’ turnover intentions. Journal of Personnel Psychology,10(4), 182.
-Hersey, P., & Blanchard, K. H. (1969). Life cycle theory of leadership. Training & Development Journal.
-Hunter, J. E., & Schmidt, F. L. (1990). Dichotomization of continuous variables: The implications for
meta-analysis. Journal of Applied
Psychology, 75(3), 334.
-Judge, T. A., & Piccolo, R. F. (2004). Transformational and transactional leadership: a meta-analytic test of their relative validity. Journal of applied
psychology, 89(5), 755.
-Korman, A. K. (1966). “Consideration,” “Initiating Structure,” and Organizational Criteria-‐‑a Review. Personnel Psychology, 19(4),
349-361.
-Krackhardt, D., McKenna, J., Porter, L. W., & Steers, R. M. (1981). Supervisory behavior and employee
turnover: A field experiment. Academy of
Management Journal, 24(2), 249-259.
-Lee, L., Horth, D. M., & Ernst, C. (2012). Boundary Spanning in Action: Tactics for Transforming Today's Borders into Tomorrow's Frontiers. Center for
Creative Leadership, Greensboro, NC.
http://www. ccl.
org/leadership/pdf/research/boundarySpan ningAction. pdf.
-Lowe, K. B., Kroeck, K. G., & Sivasubramaniam, N. (1996). Effectiveness correlates of transformational and transactional leadership: A meta-analytic review of the MLQ literature. The Leadership Quarterly, 7(3), 385-425
-Maertz, C. P., & Griffeth, R. W. (2004). Eight motivational forces and voluntary turnover: A theoretical synthesis with implications for research. Journal of
Management, 30(5), 667-683.
-Malakyan, P.G. (2014). Followership in Leadership studies: A case of Leader Follower Trade Approach. Journal of leadership Studies, 7(4), 6-22.
-Mardanov, I. T., Heischmidt, K., & Henson, A. (2008). Leader-member exchange and job satisfaction bond and
predicted employee turnover. Journal of
Leadership & Organizational Studies.
- Mellor, N., Arnold, J., Gelade, G. (2009). The effects of transformational leadership on employee absenteeism in four UK public sector organizations.
Health and Safety Executive.
-Minavand, H., Mokhtari, S. E., Zakerian, H., & Pahlevan, S. (2013). The impact of project managers’ leadership style on employees’ job satisfaction, performance and turnover. IOSR journal of
business and management, 11(6), 43-49.
-Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred reporting items for systematic reviews and
meta-analyses: the PRISMA
statement.Annals of internal
medicine, 151(4), 267.
-Mowday, R. T., Porter, L. W., &
Steers, R. M. (2013). Employee—
organization linkages: The psychology of commitment, absenteeism, and turnover.
Academic Press.
-Northouse, P. G. (2015).
Leadership: Theory and practice. Sage
publications.
-Nyberg, A., Westerlund, H., Hanson, L. L. M., & Theorell, T. (2008). Managerial leadership is associated with self-reported sickness absence and sickness presenteeism among Swedish men and women. Scandinavian journal of public
health, 36(8), 803-811.
-Pearce, C. L., & Conger, J. A. (2003). Shared leadership: Reframing the hows and whys of leadership. Thousand Oaks, CA: Sage.
-Pool, S. W. (1997). The relationship of job satisfaction with substitutes of leadership, leadership behavior, and work motivation. The
Journal of Psychology, 131(3), 271-283.
-Schreuder, J. A., Roelen, C. A., Van Zweeden, N. F., Jongsma, D., Van der Klink, J. J., & Groothoff, J. W. (2011). Leadership effectiveness and recorded sickness absence among nursing staff: a cross-‐‑sectional pilot study. Journal of
nursing management, 19(5), 585-595.
-Sellgren, S., Ekvall, G., & Tomson, G. (2007). Nursing staff turnover: does leadership matter?. Leadership in
-Sheridan, J. E., & Vredenburgh, D. J. (1978). Usefulness of leadership
behavior and social power variables in predicting job tension, performance, and turnover of nursing employees. Journal of
Applied Psychology, 63(1), 89.
-Silverthorne, C. (2000). Situational leadership theory in Taiwan: a different culture perspective. Leadership & Organization Development Journal, 21(2),
68-74.
-Tracey, J. B., & Hinkin, T. R. (2008). Contextual factors and cost profiles associated with employee turnover. Cornell Hospitality Quarterly, 49(1), 12-27.
-Tse, H. H., Huang, X., & Lam, W. (2014). Why does transformational leadership matter for employee turnover? A multi-foci social exchange perspective (vol 24, pg 763, 2013). LEADERSHIP
QUARTERLY, 25(3), 628-629.
-Woo, S. E., Chernyshenko, O. S., Stark, S. E., & Conz, G. (2014). Validity of six openness facets in predicting work behaviors: A meta-analysis. Journal of
personality assessment, 96(1), 76-86.
-Yukl, G., Gordon, A., & Taber, T. (2002). A hierarchical taxonomy of
leadership behavior: Integrating a half century of behavior research. Journal of
Leadership & Organizational
Studies, 9(1), 15-32.
-Yukl, G. (1999). An evaluative essay on current conceptions of effective leadership. European Journal of Work and
Organizational Psychology, 8(1), 33-48.
-Zhu, W., Chew, I. K., & Spangler, W. D. (2005). CEO transformational leadership and organizational outcomes: The mediating role of human–capital-enhancing human resource management. The Leadership Quarterly, 16(1), 39-52.
Appendix 1.
Coded variables 1 TN
2 Leadership Style
3 Cor Leadership Withdrawal 4 Three clusters 5 Four clusters 6 Leader Behavior 7 Leader Trait 8 Author 9 Year 10 Title 11 Journal 12 Level 13 Published 14 Year data 15 Research purpose 16 Country first author 17 Design 18 Type 19 Research type 20 Mediators 21 Moderators 22 Independent variable 23 Dependent variable 24 Comment variables 25 Total Study N 26 Sample Country 27 Sample Size 28 Weight 29 Comment SS 30 Sample Type
31 Sample Size Employee
33 Mean Age 34 SD Age
35 Mean Age Employees 36 SD Age Employees 37 Tenure Employee 38 Tenure Employee SD 39 Mean Age Leader 40 SD Age Leader 41 Tenure Leader 42 Tenure leader SD 43 Percentage Male
44 Percentage Male Employee 45 Percentage Male Leader 46 Comment on Sample 47 Leadership Measurement 48 Type Leadership Measure 49 LM Tool
50 LM Tool subscales 51 LM Tool items 52 LM Tool scale from 53 LM Tool scale to 54 LM comment 55 Perspective LM 56 Mean LM 57 SD LM 58 CA LM 59 Withdrawal 60 Type Withdrawal 61 W Tool 62 W Tool items 63 W Tool scale from 64 W Tool scale to 65 W Tool alpha
67 W SD
68 Comment W Measure 69 Cor Turnover Absenteeism 70 Cor LM LM
71 Cor LM Mediator 72 Cor LM Mediator 2 73 Cor LM Moderator 74 Cor LM Moderator 2 75 Cor Med Mod 76 Cor Med Mod 2 77 Cor Med Withdrawal 78 Cor Med2 Withdrawal 79 Cor Mod Withdrawal 80 Cor Mod2 Withdrawal 81 Extra Correlations 82 Time point Measurement 83 Comment 84 Source of Measurement 85 Comment 86 Public 87 Job complexity 88 Level Leadership 89 Sample 90 Setting 91 Comments
