A moderated mediation model: Inconsistent leader behavior and its effects on LMX, followers’ resilience and followers’ perceived stress
Naomi Kuin – 13476912
Date of submission: 23 June 2022, Master Thesis Final Version MSc in Business Administration – Leadership & Management Track University of Amsterdam
EBEC 20220622010611
Supervisors: Joanna Ritz, Almasa Sarabi
State of Originality
This document is written by Student Naomi Kuin who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of Contents
List of Tables ... 4
Abstract ... 5
Introduction ... 6
Theoretical model ... 9
Literature review ... 10
Inconsistent leader behavior ... 10
Followers’ perceived stress ... 12
Inconsistent leader behavior and followers’ perceived stress ... 14
Mediating role of followers’ resilience ... 15
Moderating effects of LMX ... 19
Data and Research methodology ... 23
Participants and procedure ... 23
Measures ... 24
Results ... 27
Data analysis ... 27
Correlation analysis ... 27
Moderated Mediation model ... 29
Discussion ... 31
Summary of findings ... 31
Theoretical contributions ... 32
Future research ... 36
Limitations ... 35
Practical implications ... 37
Conclusion ... 39
Reference list ... 40
Appendix ... 49
Appendix A – Survey measurement item scale ... 49
List of Tables
Table 1. Correlation matrix 28
Table 2. Moderated Mediation model 30
Abstract
Recently, (in)consistent leadership literature has gained interest since studies have discovered the effects of leader behavior on follower perceptions. Previous research has shown that inconsistent leadership negatively influences followers’ emotions. However, inconsistent leader behavior is not broadly investigated yet. The relatively small number of researches regarding this topic stated that inconsistency is characterized by unpredictability, which acts as a stressor for followers. Therefore, to contribute to leadership theory and practice, the purpose of this study is to prose a moderated mediation model whereby inconsistent leader behavior predicts followers’ perceived stress. It is expected that this is explained by followers’ resilience.
It will be investigated if the leader-member exchange quality relationship (LMX) moderates this relationship.
A single-source field study has been carried out, using an online survey with 216 followers in the Dutch Workforce. As expected, the moderated mediation was supported. The findings discovered that inconsistent leader behavior is related to followers’ perceived stress which could be explained by the underlying mechanism of followers’ resilience. Additionally, this study found that followers in a high LMX quality relationship weaken the negative relationship between inconsistent leader behavior and followers’ resilience, as compared to followers in a low leader-member exchange relationship. Interestingly, the correlation analysis findings showed that there is an association between inconsistent leadership and followers’
perceived stress. Nevertheless, the regression analysis did not find a direct effect between inconsistent leader behavior and followers’ perceived stress and thus also no moderation effect between inconsistent leadership and followers’ perceived stress.
Theoretical as well as practical implications are discussed. This study contributes especially to the field of inconsistent leadership and its effects. The practical implications are
Introduction
In a perfect world, leaders would always behave consistently since it has been proven as an effective leadership trait (Shamir & Eilam, 2005). However, in reality this is not the case.
Leaders also have off days and thus, might behave inconsistently. For example, at one moment leaders might behave in a supporting manner whereas at another moment they do not. For followers, this is without a clear reason. They do not have clear information about what to expect from their leader. This leader behavior can therefore be perceived as unpredictable, irregular and inconsistent (De Cremer, 2003; Klug et al., 2019; Van Gerven, 2022). Previous leadership research has shown that this varying behavior is displayed in many types of leader behavior (e.g., abusive leadership, transformational leadership). Nevertheless, literature focused on leader inconsistent behavior did not receive that much attention yet, in comparison with consistent leadership, even though it has been found to negatively impact followers’
emotions (Bass, 1985; Breevaart & Zacher, 2019; De Cremer, 2003; Johnson et al., 2012; Katz et al., 2020). Therefore, the need to investigate the potential effects of inconsistent leader behavior is necessary.
Prior research in the field of leadership discovered a theoretical link between inconsistent leader behavior and followers’ stress (Breevaart & Zacher, 2019; De Cremer, 2003;
Johnson et al., 2012; Katz et al., 2020; Monat et al., 1972). They stated that unpredictable character of inconsistent leadership is mostly experienced as a stressor, whereas predictability is desirable by followers. Predictable and consistent behavior reflects the leaders’ core values and reduces uncertainty regarding the expectations of followers. On the other hand, inconsistent behavior will increase confusion among followers and is seen as less desirable leadership (Walumbwa et al., 2010). Matta et al. (2017) agreed on the value of predictability and found that followers would rather be constantly mistreated than be treated well irregularly.
To clarify, unpredictability is an essential aspect of inconsistent leader behavior, which can be justified based on the conservation of resource (COR) theory. This theory has emerged as the leading theory for understanding the processes of stress (Hobfoll, 2001). The main idea of the theory is that individuals strive to obtain, protect and build resources. If resources are being threatened (e.g., not being able to predict the behavior of the leader), followers want to protect their resources by observing the variation in leader behavior. As a result, they get distracted by their core task (e.g., being motivated) and depletion of energetic resources will arise. Thus, this indicates that the cognitive and emotional demand of understanding the leaders’
intentions triggers followers’ perceived stress (Burger & Arikin, 1980; Hobfoll et al., 2018). In the end, it has been acknowledged that followers’ stress relates to many negative psychological and physical outcomes, such as fear and depression (Skakon et al., 2010; Lazarus, 1966).
This study proposes that the relationship between inconsistent leadership and followers’
perceived stress is explained by followers’ resilience since it has become a very important resource in widely-used studies regarding stress (Hobfoll et al., 2018; Klug et al., 2019; Nguyen et al., 2016). Resilience helps enable followers to become flexible and adaptable towards experienced stressful circumstances (Klug et al., 2019). Related research argued that leader behavior impacts followers' resilience. Logically, leaders have an influential role and formal power within the organization and therefore impacts followers’ perceptions. (Agarwal, 2019;
Paul et al., 2016). Particular for this study, unpredictable leader behavior may provoke confusion and feelings of uncertainty, which results in negative emotions. Followers who lack resilience can easily get overwhelmed by these emotions and this may negatively affect their mental state. On the other hand, followers who possess resilience can control these emotions (De Cremer, 2003). Therefore, resilience could be an important resource to limit followers’
perceived stress. Having a high level of resilience can have a positive effect on the level of the followers’ perceived stress as compared to having low resilience.
As mentioned earlier, inconsistencies in leader behavior are likely to be perceived as unpredictable and cause followers’ stress. Prior research has already known that the quality of the leader-follower relationship influences the followers’ perception of that leader behavior (De Cremer, 2003; Van Gerven, 2022; Skakon et al., 2010; Tafvelin et al., 2019; Zhao & Zhou, 2021). This can be explained by the LMX theory, which states that followers’ work-related outcomes and behavior are dependent on how their leader treats them. LMX centralizes the dyadic relationship on a continuum of low-to high-quality exchanges. High-quality LMX relationships are based on high levels of trust, interaction, and support whereas low-quality LMX relationships are purely based on mutually agreed on duties as specified in their work contract (Dienesch & Liden, 1986). Followers in a high-quality LMX relationship are more committed to their leader, because the leader provides them with valuable resources such as trust (Klug et al., 2019; Liang et al., 2021; Piccolo & Colquitt, 2006; Zhu et al., 2019). On the other hand, followers who experience a low-quality LMX relationship are less likely to improve their resilience. This is because not getting the right resources (e.g., support) from their leader negatively impacts the followers’ capability and mental state. Therefore, this study expects that the quality of the LMX relationship influences the effects of inconsistent leader behavior on followers’ perceived stress.
The primary objective of this study is to research and extend the theory of inconsistent leadership and their effects on followers. This is underdeveloped in comparison with consistent behavior. Leader inconsistencies did not receive that much attention yet even though it has been found to negatively impact followers’ emotions. The outcome of perceived followers’ stress will be examined. I argue that leader inconsistency triggers stress and ultimately affects the psychological and physical health of the follower. It will be examined with the underlying mechanism of followers’ resilience as well as the contextual strength (LMX) that explains this relationship. Besides this, the findings contribute to the literature and practice in more ways.
To start, using followers’ resilience as a mediator contributes to resilience literature by creating more understanding of the underlying mechanism that impacts the outcomes of leader inconsistencies. Organizations could train followers’ resilience to limit or even prevent stress.
Second, this will be the first study that examines how inconsistent leader behavior negatively impacts perceived stress through followers’ resilience, since resilience is a good predictor of perceived stress. Finally, this study is contributing to the prior literature by focusing on leader behavior and the leader-follower dyadic relationship, which is often used separately. It may suggest that besides leader inconsistencies the quality of the leader-member relationship plays a critical role in followers’ perceived stress. Survey design will be used to answer the following question:
“Is the negative impact of inconsistent leader behavior on followers’ perceived stress weaker when there is a strong LMX quality relationship through followers’ resilience compared to a low LMX quality relationship?”.
The theoretical model has been visualized in the figure below (figure 1).
Figure 1
Theoretical model
LMX
Followers’ resilience
Followers’
perceived stress Inconsistent
leader behavior
Literature review
Inconsistent leader behavior
Leadership behavior literature has gained interest since studies have explored the effects of leader behavior on follower perceptions (Walumbwa et al., 2010). On one hand, leaders are expected to be consistent in their behavior to followers, but on the other hand, they need to be flexible to deal with different circumstances. Early research based on the contingency theory (e.g., Fieldler, 1967) already stated that behaviors might change over time and cannot be fully consistent. However, the effects of leaders who are displaying multiple types of leader behavior has been less investigated, compared with consistent leader behavior (e.g., Breevaart & Zacher, 2019; Johnson et al., 2012). Nevertheless, Fieldler (1967) and Yukl and Mahsud (2010) focused on multiple leader behaviors such as flexible leadership and contingency theory. They stated that the behavior of the leader will be adapted to a specific situation. For inconsistent leaders, this is not necessarily right. According to De Cremer (2003), inconsistency is defined as the different decisions a leader makes and the different behaviors a leader uses towards persons and situations. For example, one day the leader prioritizes a specific goal whereas the following day another goal is more important. To clarify, this example highlights the inconsistencies in displaying various leader behavior which is not adjusted to a specific situation. Therefore, this study focuses on leader inconsistencies in similar situations.
Until now, only a few areas of inconsistent leader behavior have received attention. One area is fairness. The research of De Cremer (2003) investigated what inconsistent leadership does to procedural fairness by applying the consistency rule. This rule centralizes the importance of consistency across people and over time within groups, organizations and interpersonal relationships (Leventhal et al., 1980). The study discovered that inconsistencies in decision making by leaders were evaluated as less fair. Inconsistencies in leader behavior lead to feelings of uncertainty by followers, where they experienced a lower self-image
compared to followers of a consistent leader (De Cremer, 2003). Subsequently, people’s intuition concludes that fairness is always better. However, this is not always the case. Matta et al. (2017) found that followers who irregularly treated well become more stressed than followers that are regularly mistreated. Also, the research of Teoh et al. (2016) stated that consistency can be seen as the desirable leader behavior for followers, in comparison with inconsistency. They examined the interaction between supportive and unsupportive leader behavior and the effect on work attitudes. Findings revealed that when leaders behave inconsistently, followers may not trust the intention of the leader and could respond in a critical way (Teoh et al., 2016). Thus, this implies that leader inconsistencies influence followers’
reactions and might hinder their perception of successful leadership. Therefore, it is of high importance that leaders should be aware of their behavior towards followers. Creating awareness of their inconsistencies may help to reduce the negative consequences for followers.
Nevertheless, some authors in the field of leadership did not agree on the impact of leader behavior and proved that leader inconsistencies only have a small effect on followers’
perceptions (Croco, 2016; Croco et al., 2015; Kruglanski et al., 2018). They found that followers’ do not always perceive their leaders negatively after experiencing inconsistent leader behavior. The study of Croco (2016) discovered that consistency in the behavior of a political leader depends on the degree to which followers wish to attain certainty. If followers do not desire certainty, they do not care about inconsistencies in leader behavior (Croco, 2016). On the other hand, when followers want to feel certain, they will be negatively affected by the leaders’ inconsistent behavior (Di Santo et al., 2020).
The second area of inconsistent leader behavior that has received attention is the switch between contractionary leadership styles. The most researched and known styles are transformational and passive leadership (Breevaart & Zacher, 2019; Diebig & Bormann, 2020;
Johnson et al., 2012; Mullen et al., 2011). For example, Breevaart and Zacher (2019) have
investigated the effect of these opposite styles on the leaders’ perceived effectiveness. A transformational leader is seen as an inspiration and role model for followers. They are related to positive follower performance and job satisfaction. Conversely, passive leaders, also known as laissez-faire leaders, have been shown to undermine followers’ satisfaction with their job and follower motivation. They do not engage with their followers and slip away from responsibilities. This leadership style is seen as inefficient. It seems that leaders who are showing these two styles at the same time are perceived as less effective by their followers. As a result, followers lack trust in their leader (Breevaart & Zacher, 2019). Thus, transformational leaders need to understand the impact of their various behavior and be aware of their consistency. This is because displaying multiple leadership styles will negatively impact followers’ reactions and emotions such as stress (Skakon et al., 2010).
Followers’ perceived stress
Work-related stress is a prevalent and increasing problem in the modern workplace. This is reasoned by the increased complexity in organizational work environments and challenging circumstances such as COVID19 (Skakon et al., 2010). Individuals with stress perceive that the demands of an external situation are beyond their perceived ability to cope with them. Stress is therefore defined as an uncomfortable and undesirable feeling experienced by an individual.
This is associated with negative outcomes of fear, anxiety, irritation, annoyance, anger, sadness and depression (Lazarus, 1966; Skakon et al., 2010). Moreover, the topic of stress has been widely studied by researchers to understand the experienced feeling but most importantly, to limit the problem (Skakon et al., 2010). Along with this, the conservation of resource (COR) theory has been used in organizational behavior and psychology and adopted in many areas of stress. The tenet of this theory states that individuals strive to obtain, foster, retain and protect the things they value (Hobfoll et al., 2018). For example, a good relationship with their leader.
Lazarus (1996) suggests that stress occurs when followers feel threatened by the resources they value. The response to stress leads to psychological outcomes such as strains (Lazarus, 1996).
Similarly, the World Health Organization (WHO) has stated that stress negatively affects individuals’ psychological and physical health. Poor performance of an individual could be a result of stress and may influence organizational effectiveness as well (WHO, 2021).
However, people assume that stress has always had a negative effect. Interestingly, this is not always the case. Minimal amounts of stress are believed to promote followers’
performance (Liang et al., 2021; Montani et al., 2017). Previous literature has constructed a challenge-hindrance model to explain two types of stressors and the difference in terms of the impact on individual performance (LePine et al., 2005; Montani et al., 2017; Lazarus &
Folkman, 1984). To start, challenging stressors are described as demands that have the potential to promote achievement or personal growth. They require energy but are simultaneously stimulating. This type of stressor contributes to performance opportunities whereas hindrance stressors will interfere with performance or goals. Hindrance stressors are described as demands that block the followers’ process towards personal development and accomplishments in their work. It may result in deplete energy and elicit emotional-focused coping (Eatough et al., 2011;
Mazzola & Disselhorst, 2019; Montani et al., 2017). Hence, it can be concluded that challenging and hindrance stress differs. Nevertheless, followers who experience both types of stressors are unlikely to perform well. This is because hindrance, as well as challenging stressors, require followers’ energy and resources. Challenging stress has specifically the potential for short-term gains (e.g., increased motivation or positive affect), but the negative physical and psychological outcomes will overrule these gains in the long term (Mazzola & Disselhorst, 2019). The meta- analytic results of Eatough et al. (2011) confirm this perspective regarding the positive and negative effects on followers’ performance.
Inconsistent leader behavior and followers’ perceived stress
A systematic review of three decades of empirical research has stated that leader behavior is a crucial element that impacts followers perceived stress (Skakon et al., 2010). The linkage describes Skakon et al. (2010) as the following: “stress involves a crossover contagion process, where leaders’ mood is seen as being ‘contagious’” (p. 108). Previous research in this area showed that inconsistent leader behavior was mostly related to unpredictable experiences for followers (De Cremer, 2003). This is because unpredictability can be seen as undesired leader behavior since it has often been experienced as a stressor. On the other side, predictable and consistent behavior is desirable by followers, because it reflects the leaders’ core values. Also, it reduces uncertainty regarding the expectations of followers, which will be increased by inconsistent leader behavior (Walumbwa et al., 2010). In the end, if followers cannot predict the leaders’ behavior, they do not have the control anymore and do not know how they handle the varying behavior (Bormann & Diebig, 2020; Kemery, 2006; Klug et al., 2019). Burger and Arkin (1980) found that the feeling of helplessness and a lack of perceived control are causing motivational losses and perceived stress. On top of that, followers may feel uncertain about the relationship with the leader (e.g., if the leader is not giving consistent support) which evolve in followers’ stress. Furthermore, low support is known as one of the most common causes of followers perceived work-related stress (Diebig & Bormann, 2020; WHO, 2021). The research of Skakon et al. (2010) agreed on this and found that negative leader behavior (e.g., leader undermining) directly impacts follower stress. Conversely, leader behaviors support might prevent stress and improve followers’ stress coping (Skakon et al., 2010).
Consistent with the COR theory, Hobfoll et al. (2001) suggest that stress is a consequence of resource loss. They argued that people obtain, foster and protect the things they centrally value. This is reasoned by the fact that people do not want to meet stressful challenges and therefore use their key resources. Besides this, people want to build resources for future
needs, which is based on the evolutionary need to conserve resources for survival (Hobfoll et al., 2001). Therefore, leaders have to invest in followers’ resources (e.g., personal, social and material) to maintain mutual respect, interpersonal trust and predictability of their leaders behavior. Nevertheless, if a leader lacks these resources in a similar situation, followers may perceive this varying behavior as inconsistent and unpredictable (e.g., one day the leader is showing empathy to the follower whereas the next day the leader is not interested).
As mentioned before, followers have to predict the leaders’ behavior. These mixed signals of leader behavior may deplete the follower resources (Bormann & Diebig, 2021; De Cremer, 2003; Matta et al., 2017). Bormann and Diebig (2021) argued that the more resources are depleted, the more followers are looking for coping resources. Finally, more stress is perceived by the follower. This literature confirms the prior mentioned research which has shown that inconsistency in leader behavior negatively affects followers’ perception (Breevaart
& Zacher, 2019; De Cremer, 2003; Katz et al., 2020). Therefore, I propose a positive association between inconsistent leader behavior and followers’ perceived stress, whereas followers perceive more stress when inconsistencies in leader behavior increase. Thus:
Hypothesis 1: Inconsistent leader behavior is positively related to followers’ perceived stress.
Mediating role of followers’ resilience
Following the COR theory, the relationship between inconsistent leader behavior and followers’ perceived stress is impacted by resilience. Prior research argues that resilience is one of the most important mediating variables in the leadership literature. Resilience is a personal resource and part of the psychological capital (PsyCap) which is known for the potential drive to influence organizational outcomes (Agarwal, 2019; Karatepe & Talebzadeh, 2016; Liao, 2015; Luthans et al., 2007). However, no study has examined followers’ resilience as an
explanation for the relationship between inconsistent leader behavior and followers’ perceived stress. Therefore, this paper applies resilience as an effective way to create more clarity about why followers perceive stress differently as a response to inconsistencies of their leader.
Resilience is commonly described as “the ability to bounce back or overcome some form of adversity”, which is based on the Latin verb resilire which means “bounce back”.
However, the concept of resilience is still unclear (Hu et al., 2015). Resilience considered as a personal characteristic can be conceptualized as a trait, outcome or process (Hu et al., 2015;
Kossek & Perrigino, 2016; Luthans et al., 2007). Näswall et al. (2015) described resilience as
“an employee capability, facilitated and supported by the organization, to utilize resources to continually adapt and flourish at work, even if/when faced with challenging circumstances” (p.
5). This description focuses on the dynamic process of resilience which is characterized by constant change and context-specific situations. Further, it highlights that resilience is an interactive concept that involves interaction between the individual, adversity, environment and the outcome, which create a positive outcome (Vella & Pai, 2019).
Specifically, followers’ resilience linked to the leader, has gained more interest over the past few years. This is because leaders play a vital role in affecting followers’ resilience due to their formal power in the organization. For example, leaders can provide followers with resources such as giving support (Zhu, 2019). However, when leader behavior contains inconsistencies, it is likely to negatively impact an individual capability and mental state.
Followers who are working for inconsistent leaders are often confused by the different decisions a leader makes. This ambiguity in leader behavior makes the followers less optimistic about their own individual capability (Agarwal, 2019). For instance, followers do not dare to give their opinion during a meeting anymore. According to COR theory, inconsistent actions of a leader could deplete followers’ resources and possibly lose future resources due to the negative responses (Hobfoll, 2001). To maintain a positive perception about the decision-making of their
leader, followers need to cope with the consequences of inconsistent leader behavior that might influence their own belief of ability. Thus, resilience is a personal resource that can cope with the consequences of inconsistencies in leader behavior. This can be executed by refueling followers’ energy in challenging circumstances and providing their ability to identify effective solutions (De Clercq et al., 2021).
Furthermore, Borden et al. (2018) stated that followers with a high level of resilience will suffer less from the inconsistent decision-making of the leader. Resilient people are able to function better cognitively, make better decisions and have higher job satisfaction. However, these mixed signals in behavior are still likely to be viewed as incompatible which results in a vulnerable position for the follower. On the other hand, followers with lower levels of resilience suffer more from inconsistent leader behavior (Berardi, 2020). Additionally, they find it harder to adapt to this behavior and recover slower from difficult circumstances (De Clercq et al., 2021). Thus, inconsistencies in leader behavior may lead to draining the energy reserves of followers, because they feel worse about their own capability to handle challenging circumstances. These negative actions could lead to more depleted resources and decrease followers’ resiliency (Agarwal, 2019; Youssef & Luthans, 2007). In line with the aforementioned, I expect that inconsistent leader behavior harms followers’ resilience.
Although leader inconsistencies may have a negative effect on followers, research stated that resilience emerges from early life experiences with stress. Dagnino-Subiabre (2022) describes resilience as a biological behavioral process that is continuously generated over time to regulate stress responses. Therefore, the adaption of stressful events can be gradually developed and thus inconsistencies as well. Good management practice and organizational interventions can develop followers’ resilience (Youssef & Luthans, 2007). The benefit of resilient followers is that they are able to effectively bounce back from stress. Additionally, they can easily adapt to challenging circumstances whereas less resilient followers cannot
translate the learnings of stressful events in the past into something positive. As a result, less resilient followers get quicker affected by long-term stress. Continuous stress increases the risk of mental health problems (Liu et al., 2013).
In line with this statement, the meta-analysis of Hu et al. (2015) showed that resilience was negatively correlated to negative indicators of mental health. In particular stress, which is the opposite for positive indicators. Therefore, it can be argued that resilient followers’ can easier limit the negative impact of stressful events. This explains that followers who are having a high level of resilience can have a positive effect on the level of the followers’ perceived stress. On the other hand, low resilient followers can increase the followers’ perceived stress.
Thus, considering the COR theory, I can conclude that inconsistent leader behavior results in the loss of followers’ resilience. Followers will suffer from perceived stress whereby necessary actions need to be taken to prevent any further resource loss (Agarwal, 2019). Therefore, I expect that resilience will impact the positive relationship between inconsistent leader behavior and followers’ perceived stress. Hence:
Hypothesis 2: The positive relationship between inconsistent leader behavior and followers’
perceived stress is mediated by followers’ resilience. Followers who experience inconsistent leader behavior are expected to experience a lower level of resilience which lead to more perceived stress.
Moderating effects of LMX
Looking at the previous relationship, the COR theory suggests that people react less to stress when followers are protected from the anticipated loss of valued resources. This indicates that the resources from the environment (e.g., support from the leader) increase the followers’
ability to cope with challenging events (Hobfoll, 1989). It seems that the quality of the social exchange relationship between the leader and follower, also known as LMX, is one of these resources. The resource shows to have a resource-protective function for followers who experiences stress (Montani et al., 2017). Therefore, this study will look at the leader-member exchange quality in the relationship abovementioned relationship.
The Leader-member exchange theory proposes that a dyadic relationship between the leader and subordinate develops over time (Graen & Cashman, 1975). The theory is describing a continuum based on low- to high-quality exchanges, because it focuses on the nature and quality of the relationship. The characteristics of a high-quality relationship are mutual respect, communication, support, trust and loyalty. On the other hand, a low-quality relationship is mainly based on an employment contract. It is characterized by poor information exchange and close supervision due to the low levels of trust (Dansereau et al., 1975; Dienesch & Liden, 1986; Wayne et al., 1997). Montani et al. (2017) stated that followers in high LMX relationships exhibit superior job performance, enjoy higher levels of trust and support and receive more rewards from the leader, compared to low LMX relationships. When followers have a good relationship with the leader, it is less likely to threaten the follower’s resources (such as resilience). Consequently, it can help to buffer followers against the damaging effects of work- related stress (Dansereau et al., 1975; Harland et al., 2005; Hobfoll et al., 2018).
Moreover, there are several reasons why high LMX relationships may protect the follower’s resources during stressful events. First, research shows that high LMX relationships enhance communication between leader and subordinate. Clear communication will induce the
leaders’ varying behaviors and make it more predictable for a follower. For example, the leader and follower will evaluate the stressful conditions in a specific goal together. Conversely, low LMX relationship followers are disadvantaged in communication with their leader since they get less attention for the stressors they are dealing with. The chance that a follower experiences a loss of resources is therefore higher. As a result, they cannot deal with the inconsistencies in leaders’ behavior (Harris & Kacmar, 2006; Montani et al., 2017).
Second, the support, resources and encouragement in high LMX relationships creates a high personal control over stressful events by followers. This is necessary to reduce the risk of resource loss, because high personal control of followers increases coping resources, build resilience and prevents resource depletion (Hobfoll et al., 2018; Montani et al., 2017; Schmidt
& Diestel, 2015). Leader support, which is an essential characteristic of the high LMX relationship, can help to build individuals’ resilience against the harmful effects of work-related stress. On top of that, it can buffer followers from resource loss (Agarwal, 2019; Montani et al., 2017). Subsequently, Skakon et al. (2010) concluded that leadership support might prevent followers’ perceived stress and improve their coping and affective wellbeing. Both are key determinants of individual and organizational health. Conversely, followers in a low LMX relationship may distrust their leader. This relates to the leaders coping abilities as well as their unwillingness to offer help and resources. Therefore, the perception of resource loss might arise. Followers will appraise stressful events as personally uncontrollable (Skakon et al., 2010).
To clarify, the relationship between the leader and follower is one of the most common sources of stress in organizations (Skakon et al., 2010). Followers need to perceive the relationship with their leader as psychologically healthy to prevent stress-related symptoms and illness. Subsequently, high LMX relationships will help followers to view challenging events as less threatening to their (personal) resources. However, critical research stated that resource
loss could also arise in high LMX relationships (Skakon et al., 2010). For example, Harris &
Kacmar (2006) showed that at a certain point high LMX relationships also increase followers’
perceived stress. This is reasoned by the fact that leaders overwhelm their followers with demanding requests. Followers feel extra pressure and go beyond their duties to meet the expectations of their inconsistent leader, which may result in abusive leadership (Harris &
Kacmar, 2006; Liang et al., 2021).
Finally, high levels of trust in high LMX relationships protect followers’ resources.
Researchers found that trust changes the way followers experience, interpret and react to events (Dirks & Ferrin, 2001). Although the leader could behave inconsistently, followers in a high LMX relationship feel confident that the leader cares about them. As a result, the needed resources (e.g., trust) will be available. Followers will perceive their current work conditions as less stressful and protect them from resource loss. Contrarily, followers in a low LMX relationship perceive their inconsistent leader as vague and selfish. The leaders’ behavior becomes unpredictable and ultimately drain individual resource (Montani et al., 2017).
To conclude, consistent with the COR theory of Hobfoll (1989), the aforementioned arguments suggest that followers who experience a high LMX relationship are better able to handle their leaders’ inconsistencies. Followers gain resources and feel capable of facing stressful challenges in comparison to followers in a low LMX relationship. Furthermore, it also suggests that a high LMX relationship is an interesting predictor to increase the followers’
resilience and finally decreasing the followers’ perceived stress. Therefore, I propose:
Hypothesis 3: The quality of the leader-member exchange relationship moderates the relationship between inconsistent leader behavior and followers’ resilience, such that the negative relationship between inconsistent leader behavior and followers’ resilience is weaker for followers with a high LMX relationship as compared to followers with low LMX relationship.
Hypothesis 4: The quality of the leader-member exchange relationship moderates the relationship between inconsistent leader behavior and followers’ perceived stress, such that the negative relationship between inconsistent leader behavior and followers’ perceived stress is weaker for followers with high LMX as compared to followers with low LMX relationship.
Data and Research methodology Participants and procedure
The total sample consisted of 216 followers in the Dutch workforce. Their age ranged from 20 to 66 years (M = 35.63; SD = 14.036) whereas a small majority of these followers identified as female (55% female, 43.8% male and 1.3% I rather do not say). The respondents spent between 0 and 38 years at their current organization (M = 7.080; SD = 8.709) and spent an average of 3.3 years working together with their leader (SD = 4.403). This quantitative research was carried out using a survey design, because this method can gather data on a convenient way so that the findings are interpretable. This online survey was administrated digitally with a group of four master students via the online survey software Qualtrics. The survey took approx. 10 minutes to complete. The respondents could complete the survey in Dutch (84.6% of respondents) or English (15.4% of respondents). The translation of the survey was checked by four individuals, to avoid bias resulting from interpretation. Additionally, to make sure that they had a clear understanding of the items.
This single-source study had only two requirements for participation. The population needs to be followers in the Dutch workforce and has a leader (e.g., manager, supervisor, boss).
The survey was only sent to followers and excluded leaders in participation (multi-source study) since the given timeframe was relatively short. I believed that testing a single-source study provides a higher participation rate in comparison to a multi-source study. Therefore, a combination of non-probability sampling has been used, because the population is quite large and diverse. The primary sampling technique used was convenience sampling whereby employees have been contacted via the network of the researcher. Second, snowball sampling was utilized by a small number of employees who already filled out the survey and forward the survey to colleagues. The survey was distributed via WhatsApp to individual workers whereby respondents could have filled in their answers on their laptops or mobile phone.
Out of the 312 followers’ that opened the survey, 216 respondents completed it.
Therefore, the response rate resulted in 69.23%. 96 respondents did not finish the survey, which is relatively high. This could have to do with the length of the survey where fatigue effects arise. This effect indicates that participants become tired or bored whereby their performance with the prolonged task declines (Dörnyei & Taguchi, 2009). As a result, fewer participants have completed the survey as indicated. However, more than 200 participants have filled in the survey, which indicates a fairly good survey accuracy to interpret the results. Furthermore, the data was collected in 13 days. This target was succeeded since multiple reminders in the second week were sent. Every approached follower received one reminder via WhatsApp, starting from day 8 till day 13.
Measures
The variables inconsistent leader behavior, LMX, followers’ resilience and followers’
perceived stress were used to test the hypothesis. All measures have used a 5-point scale, ranging from 1 to range 5. However, the scale names differ. For the variables inconsistent leadership and follower resilience is the most common Likert scale name “strongly disagree to strongly agree” utilized. For followers’ perceived stress was “never to very often” used whereas the scale name differs for each question for LMX (e.g., very bad to very good). All scales and questions were directly translated to Dutch, to gather more respondents in the chosen Dutch sample. After that, the survey was checked by a supervisor and sent to four friends to make sure that the translation was correct. All scales translations in Dutch were already available from researchers in the field, except for LMX. Appendix A visualizes the translations.
Perceived stress. Perceived followers’ stress was measured with the ten-items of Cohen et al. (1983) scale, where participants were asked to rate the frequency of stressful events that occurred in the past month. The scale proved reliable with a Cronbach’s alpha of 0.78. One example item is “In the last month, how often have you felt nervous and ‘stressed’?” and “In the last month, how often have you been upset because of something that happened unexpectedly?”.
Inconsistent leadership. Inconsistent leadership was measured with the relatively new four-items of Van Gerven et al. (2022) scale. This scale reflects best the definition of inconsistent leader behavior due to the focus on unpredictable and various behavior. The scale proved reliable with a Cronbach’s alpha of 0.87. Two example items are “ My supervisor is inconsistent in his/her behavior” and “My supervisor behaves differently in comparable situations”.
Employee resilience. Followers’ resilience was measured with the nine-item Employee Resilience (EmpRes) of Näswall et al. (2015) scale, which is a revision of the scale presented in Näswall et al. (2013). It reflects the best behavioral focus on the employee resilience construct and is facilitated in organizational context. The scale proved reliable with a Cronbach’s alpha of 0.86. Two example items are “I effectively collaborate with others to handle challenges at work” and “I learn from mistakes at work and improve the way I do my job”.
LMX. The Leader-member quality exchange relationship was measured with the seven- item of Graen and Uhl-Bein. (1995) scale. I have used this scale because the LMX variables (trust and respect) correspond with the definitng characteristics of LMX. The scale proved reliable with a Cronbach’s alpha between 0.8 and 0.9. Two examples item are “How effective is your working relationship with your leader?” and “How well does your leader understand your job problems and needs?”
Control variables. Demographic information such as gender (nominal variable), age, years working with the current leader and years working for the current organization (ratio variables) are gathered at the beginning of the survey. The tenure with the leader and organization (in years) is included as control variable, because the negative effects of inconsistent leadership behavior might grow over time (Paulhus, 1998). Second, age is used as a control variable considering the assumption of Huang et al. (2010) that age differences may affect stress-related feelings coming from leader behavior. Finally, gender is a common control used in leadership research. Recent research by Van Gerven et al. (2022) stated that female (narcissistic) leaders tend to behave more inconsistent compared to a male. They emphasized the importance of gender and therefore used in this research.
Results Data analysis
I used IBM SPSS version 27.0 to perform the data analysis. First, the correlation matrix has been performed, followed by a moderated mediation model. The PROCESS macro version of Hayes (2018) has been used to evaluate each hypothesis. The independent variable is inconsistent leader behavior, the dependent variable is followers’ perceived stress and followers’ resilience was used as mediating variable in this model. LMX was included as a moderator between inconsistent leader behavior and followers’ resilience as well as inconsistent leader behavior and followers’ perceived stress. Therefore, model 8 has been chosen with 5000 bootstrap samples (Hayes, 20218). The effects are statistically significant when the 95%
confidence intervals (CI) did not contain zero. Gender, age and tenure organization variables were included as control variables. Furthermore, I have operationalized high and low levels of LMX as one standard deviation below and above the mean score.
Correlation analysis
Table 1 shows the means, standard deviators and correlations from all variables including the control variables. The table shows that inconsistent leader behavior has a moderate positively significant correlation with followers’ perceived stress (r = 0.359, p <0.01). In other words, when inconsistent leadership behavior increases, the followers’ perceived stress increases as well. The variable followers’ perceived stress is also correlated with follower’s resilience (r = -0.235, p <0.01) and the quality of the LMX relationship (r = -0.314, p <0.01). Nevertheless, both were negatively related, which explains that followers in a good relationship and/or are resilient, experience less stress. The quality of the LMX relationship is negatively correlated with inconsistent leader behavior (r = -0.554, p <0.01) and is positively correlated with followers’ resilience (r = 0.332, p <0.01). This means that higher quality in the LMX
relationship is associated with lower levels of inconsistent leader behavior and a higher level of followers’ resilience.
The only correlation which is not significant (excluding the control variables) is followers’ resilience and inconsistent leader behavior. Furthermore, by looking at the demographics, age, gender and tenure organization are significantly correlated with followers’
perceived stress. Age (r = -0.347, p <0.01) and organization tenure (r = -0.199, p <0.01) are negatively correlated with followers’ perceived stress. This means that younger people experience more stress in comparison with older people. Additionally, people who work have worked for the organization for a shorter period experience less stress than people who have been working there for a longer time. Lastly, the positive correlation suggests that gender is associated with followers’ perceived stress (r = 0.160, p < 0.05). Inconsistent leader behavior, followers’ resilience and the LMX quality relationship did not correlate with the demographics.
Table 1
Means, Standard Deviation, Correlations
Note. N = 240. **p < .01, *p < .05. Target gender is coded as 1 male, 2 female, 3 rather do not say.
Means, Standard Deviations, Correlations
Variable M SD 1 2 3 4 5 6 7
1. Inconsistent Leader Behavior 2.45 .861 -
2. Resilience 3.97 .405 -.066 -
3. LMX 3.72 .578 -.554** .332** -
4. Perceived Stress 2.40 .540 .359** -.235** -.314** -
5. Age 35.63 14 -.118 -.008 -.011 -.347** -
6. Gender 1.59 .565 .112 .023 -.097 .160* -.133* -
7. Tenure leader 3.336 4.403 .019 -.101 -.009 -.101 .408** -.141* -
8. Tenure organization 7.080 8.709 -.030 -.065 -.101 -.199** .674** -.079 .557**
Moderated Mediation model
The results of the moderated mediation model are described in table 2. To start with hypothesis 1, inconsistent leader behavior is expected to negatively relate to followers’ perceived stress.
Evidence of the analysis showed that there is a no statistically significant direct effect in the relationship between inconsistent leader behavior and followers’ perceived stress (ß = .280, SE
= .212, t = 5.562, p = 0.187, 95% CI: -.137 to .698). In other words, it cannot be concluded that followers who experience a leader that behaves inconsistent, experience more stress. Therefore, hypothesis 1 is not supported.
Nevertheless, hypothesis 2 expected that the relationship between inconsistent leader behavior and followers’ perceived stress is mediated by followers’ resilience. The conditional indirect effects of the moderation mediation model seem to appear as significant for both paths (a and b). The bootstrap CI of the index of the moderated mediation does not contain zero, thus supporting Hypothesis 2. Therefore, I could argue that follower resilience fully mediates and thus explains the relationship between inconsistent leadership behavior and follower perceived stress. In other words, followers who experience a leader that is behaving inconsistently perceive more stress, only when they experience a lower level of resiliency.
For hypothesis 3, I expected that LMX moderates the relationship between inconsistent leader behavior and followers’ resilience. As predicted, results have shown that there is a positive significant direct relationship between LMX and follower resilience (ß = .566, SE = .123, t = 4.600, p = 0.000, 95% CI: .324 to .809). On top of that, results showed that the interaction effect is negatively significant, whereby there can be concluded there is a moderating effect (ß = -.101, SE = .044, t = -2.28, p = 0.024, 95% CI: -.189 to -.014). This indicates that with 1 unit increase in the behavior of the inconsistent leader, the difference between LMX and follower’s resilience increases by 0.10 units. In other words, the quality of the LMX relationship impacts the relationship between inconsistent leader behavior and
followers’ resilience, suggesting that the relationship was stronger for high LMX followers. To summarize, evidence of the analysis proved that there is a moderated mediation. Thus, hypothesis 3 is supported.
Lastly, for hypothesis 4 was the expectation that the quality of the LMX relationship directly moderated the relationship between followers’ resilience and followers’ perceived stress. Unfortunately, the direct effect of inconsistent leadership behavior and followers’
resilience without the moderator (hypothesis 1) was already rejected and thus hypothesis 4 is rejected as well. Results confirmed this whereby the interaction effect is negatively insignificant (ß = -.025, SE = .058, t = -,440, p = .660, 95% CI: -.139 to .088).
Table 2
Moderated Mediation results for the relationship between Inconsistent Leader Behavior and Followers’
Perceived Stress with 95% Bias-corrected Confidence Intervals (N = 212).
Note. LMX: Leader-member exchange quality relationship, LLCI: Lower Limit Confidence Interval;
ULCI: Upper Limit Confidence Interval
Followers' Perceived Stress (Y)
Antecedent Coeff. SE p Coeff. SE p
Inconsistent Leader Behavior .453 .165 <.01 .176 .194 .365
Followers' Resilience - - - -.237 .080 <.01
LMX .564 .126 <.001 -.096 .152 .528
Inconsistent Leader Behavior x LMX -.103 .046 <.05 -.002 .053 .975
Gender .048 .047 .314 .063 .055 .250
Age .002 .003 .535 -.011 .003 <.001
Tenure Organization -.002 .004 .680 -.001 .005 .806
Constant 1.545 .510 <.01 3.562 .601 <.001
R2 = .156 R2 =.301
F(6.205) = 6.335, p<.001 F(7.204) = 12.555, p<.001
LMX Unstandardized
Boot Effects Boot SE Boot LLCI Boot ULCI
3.136 -.030 .016 -.064 -.005
3.723 -.016 .012 -.043 .003
4.309 .002 .012 -.029 .022
Consequent
Follower Resilience (M)
Conditional indirect effect at Interaction to follower perceived stress
for the levels of LMX
Discussion
Summary of findings
This study contributes to existing leadership literature, especially in the field of inconsistent leadership. It aimed investigation, because previous research briefly stated that leader behavior can affect followers’ (personal) resources such as resilience and perceived stress. Therefore, the LMX quality relationship and followers’ resilience were examined to see whether they influenced the relationship between inconsistent leader behavior and followers’ perceived stress. Thus, I gathered from the leadership literature, employed the assumptions of the COR theory (Hobfoll, 1989) and proposed a moderated mediation model. The study hypothesized that inconsistent leadership results in followers’ perceived stress through followers’ resilience.
Additionally, I expected that the effect of inconsistent leader behavior on followers’ resilience and perceived stress was dependent on the LMX quality relationship.
The results showed that inconsistent leader behavior is only correlated to followers’
perceived stress but does not show a significant relationship (hypothesis 1). However, interestingly is that this relationship is fully mediated by followers’ resilience, supporting hypothesis 2. Next to this main finding, the quality of the LMX quality relationship also seems to moderate (weakens) the negative relationship between inconsistent leader behavior and followers’ resilience (hypothesis 3). Thus, we could argue for a significant moderated mediation model. The study did not find a moderation effect between inconsistent leader behavior and followers’ perceived stress, rejecting hypothesis 4 (hypothesis 1 was already rejected). The findings will be elaborated on in the next paragraph.
Theoretical contributions
To start, this research has discovered that the relationship between inconsistent leader behavior and followers’ perceived stress is explained by followers’ resilience. It is found that followers’
facing inconsistent leader behavior experience a loss of (personal) resources, which results in perceived stress. This is in line with the scientific proof that resilience is a good predictor of stress (Agarwal, 2019; Karatepe & Talebzadeh, 2016; Liao, 2015; Luthans et al., 2007).
Specifically to this research, previous studies determined that leader inconsistencies are likely to negatively impact an individual capability and mental state and therefore could deplete followers’ resources (Agarwal, 2019; De Clerq et al., 2021). So, the less resilient, the more stress a follower perceives. The current study adds value to the existing resilience as well as leadership literature by proving new information that resilience could also serve as a buffer for leader inconsistencies. Thus, using resilience as an underlying mechanism has filled in the literature gap on why followers perceive stress differently as a response to inconsistencies.
Though, the current study did not find the direct impact of inconsistent leader behavior on followers’ perceived stress. This indicates that followers do not perceive more stress when a leader is showing inconsistencies. It seems like inconsistent leader behavior is not immediately seen as negative leader behavior, which was suggested in this study. The findings of Skakon et al. (2010) are opposed to this and found that negative leader behavior directly impacts followers’ stress. It could be reasoned by the fact that this empirical research combined more than 100 studies regarding leadership behavior and styles. Still, only a few studies in this research on negative leader behavior have addressed inconsistent leadership behavior. Another explanation could be that inconsistent leadership is known for its ambiguous characteristic and its difficulty to predict (Walumbwa et al., 2010). Therefore, the direct effect of inconsistent leader behavior on followers’ perceived stress could be harder to find. It is advised for future research to dive deeper into the categorization of inconsistent leadership as negative behavior.
Nevertheless, although the moderated mediation model did not show a direct effect between inconsistent leader behavior and followers’ perceived stress, the correlation analysis found an association between these variables. The correlation finding demonstrate that when leaders show more behavioral inconsistencies, followers experience more perceived stress.
Thus, it can be argued that the gap between the correlation and relation of the followers’
perception of stress depends on other resources. This current study has discovered that resilience is one of these resources, which can be seen as an interesting contribution to the previous literature. Besides this, when looking at existing literature with regards to follower perceptions, it has been found that the negative effects of inconsistencies only have a small effect on followers’ emotions (Croco, 2016; Di Santo et al., 2020). They found that the followers’ perception of stress is dependent on the certainty level of the follower. Specifically, only followers’ who want to desire sureness (e.g., a good relationship with the leader) suffer from leader inconsistencies whereas the follower who does not desire certainty will be not affected (Di Santo et al., 2020). This study supports these findings knowing that certainty can be developed by resilience, which will be elaborated in the practical implications.
Furthermore, I aimed to find a moderation effect on the quality of the LMX relationship between inconsistent leader behavior and followers’ resilience. As expected, the results indeed provide evidence that followers in a high LMX relationship weaken the negative relationship between inconsistent leader behavior and followers’ resilience, as compared to followers in a low LMX relationship. In other words, the high-quality LMX relationship will increase followers’ resilience, as followers perceive more valuable resources compared to followers in a low LMX quality relationship. This finding is aligned with earlier studies, because the relationship between the leader and member is seen as an important resource (characterized by e.g., support and trust) to influence followers’ capability for resilience (e.g., Harland et al., 2005; De Clerq et al., 2021). The supported moderated mediation model suggests that the
quality of the LMX relationship moderates the conditional indirect effects of inconsistent leader behavior on followers’ perceived stress via followers’ resilience. Therefore, this study contributes to the prior literature by focusing on leader behavior and the leader-follower dyadic relationship, which is often used separately. On top of that, using a positive mechanism (follower resilience) to explain negative leader behavior (inconsistent leadership) has been a significant addition to the corresponding literature. The study by Fredrickson (2003) concludes that “positivity can transform people for the better, making them more resilient and socially connected” (p. 334). Positivity may even limit or remove the recurring negative effect, which is in line with the current study.
Finally, as mentioned before is hypothesis four automatically rejected considering no direct effect in the relationship between inconsistent leader behavior and followers’ perceived stress. Though, this is contrary with earlier studies, where is stated that the LMX quality relationship influences the association between negative leader behavior and follower reactions regarding this behavior (Agarwal, 2019; Piccolo & Colquitt; 2006). However, a possible explanation might be reasoned that Agarwal (2019) measures abusive leadership instead of inconsistent leader behavior. Although both are categorized as negative leader behavior, abusive leadership may have a more detrimental effect on followers. Furthermore, the combination of using two moderating effects of the LMX quality relationship (direct and indirect) could have been an interesting contribution to the literature. It could show that organizational context can affect followers with inconsistencies of the leader, resilience and perceived stress in more ways. Future research may focus on other negative leader behaviors to examine if moderation will exist. Besides this, another possible explanation could be stated in the study of Harris and Kacmar (2010), which found a curvilinear effect of LMX quality relationship on stress. In other words, followers in a high LMX quality relationship feel extra pressure and go beyond the required duties of the job. Followers that want to meet the
expectations of the leader, experience more stress. For future research, it might be interesting to examine this curvilinear effect further.
Limitations
Like every other study, the current study also has limitations. To start, all measures of this research used self-reports and may lead to common-method/common-source bias (Conway &
Lance, 2010). Although the correlations of this study are not that strong, they may impact the results. Another option would be to use behavioral observation (such as leader reports), which are increasingly being used after self-report measurement (Podsakoff & Organ, 1986).
However, this method costs more time and was not feasible in the given period of this study.
To minimize the effect of common method bias, future research needs to focus on the design and administration of the survey by mixing the order of the questions and using different scale types (Podsakoff & Organ, 1986). Additionally, future research could also look to dyads (leader and follower) respondents to provide a deeper understanding of views and perspectives.
Furthermore, a self-reported survey may provoke also another bias, namely socially desirable answers. An introduction part is added in the survey to cover the bias risk and emphasizes the need for honesty and guaranteeing anonymity. However, this bias could be the possible explanation for the insignificant relation between inconsistent leader behavior and followers’ perceived stress. Stress could be seen as socially sensitive for followers and thus answer more socially desirable. Future studies could continue with elaboration on honesty and guaranteeing anonymity and may implement the social desirability bias scale. This scale distinguishes true responses from false (Lanz et al., 2022). Though, it cannot assure complete avoiding of the bias.
Thirdly, another limitation of this study could be the survey length. The results have shown that participants have finished the survey in more than ten minutes. The feedback
received on the survey indicated that participants found it too long. In line with previous research, fatigue effects could have arisen whereby participants have become tired or bored (Dörnyei & Taguchi, 2009). Consequently, participation decreases in completing the survey. It could be a logical explanation for why more than 100 participants started the survey but did not end it. As a result, this impacts the reliability of the data, which is essential to measure the consistency of scores over time (Dülmer, 2016). Despite, this study has provided a fairly good survey accuracy (216 participants) to interpret the results. Therefore, I believe that the benefits of survey design overcome the disadvantages of other designs (e.g., diary study). For future researchers that work together, another option is to include fewer variables compared to this study. This adjustment results in a shorter survey length and will possibly limit the dropout rate (Revilla & Ochoa, 2017).
Finally, the use of non-probability convenience sampling is seen as a limitation. There were only two criteria in the sampling method (working in the Netherlands and having a leader), so almost everyone could participate in the study. Due to the network sample, it reduces the generalizability of the results. Furthermore, under of over-representations of subgroups may not represent the true picture of the sample (Vehovar et al., 2016). Future studies could try random or probability sampling if the timeframe allows that.
Future research
This study has primarily focused on the negative effect of inconsistent leadership on followers.
Therefore, I have answered the call to extend the research on inconsistencies in behavior and related outcomes (Simons, 2002). Nevertheless, a suggestion for future research is to focus on the leader instead of the follower. It could be interesting to know which type of leader shows more often inconsistencies to limit the effects. Research can take leader personality traits into account. For example, impulsivity and instability (Van Gerven et al., 2022). On top of that, this
