This item has the following downloads:
1 FROM NEGATIVE ACT TO NEGATIVE RELATIONSHIP: UNDERSTANDING HOW PATTERNS OF ABUSIVE SUPERVISION EMERGE AND DEVELOP OVER TIME By L AUREN S. SIMON A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011
2 2011 Lauren S. Simon
3 To my family
4 ACKNOWLEDGMENTS I thank my wonderful family and friends for the support they have given me thr oughout this process. I also thank my advisor committee members, and colleagues for their encouragement, advice, and guidance Few are fortunate enough to have been mentored by such an esteemed group.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRA CT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 2 LITERATURE REVIEW ................................ ................................ .......................... 15 Abusive Su pervision Literature ................................ ................................ ............... 15 Abusive Supervision Defined ................................ ................................ ............ 15 Related Constructs ................................ ................................ ........................... 16 Review of Empirical Findings ................................ ................................ .................. 20 Antecedents ................................ ................................ ................................ ..... 20 Consequences ................................ ................................ ................................ 27 Limitations of the Existing Literature ................................ ................................ ....... 42 3 THEORY AND HYPOTHESES ................................ ................................ ............... 45 Emotion Mediated Pathways ................................ ................................ .................. 46 The Emotion Process ................................ ................................ ....................... 47 Anger ................................ ................................ ................................ ................ 48 Fear ................................ ................................ ................................ .................. 51 Compassion ................................ ................................ ................................ ..... 53 ................................ ..... 55 Directe d Deviance and Abusive Supervision ........... 55 Directed Citizenship Behavior and Abusive Supervision ................................ ................................ ................................ ... 57 Subordin Directed Avoidance and Abusive Supervision ......... 58 Main and Moderating Effects of Personality ................................ ............................ 59 Trait Volatility ................................ ................................ ................................ .... 60 Trait Withdrawal ................................ ................................ ............................... 64 Trait Compassion ................................ ................................ ............................. 66 4 METHODS ................................ ................................ ................................ .............. 71 Sample and Procedure ................................ ................................ ........................... 71 Measures ................................ ................................ ................................ ................ 73 Personality ................................ ................................ ................................ ........ 73
6 Abusive Supervision ................................ ................................ ......................... 73 Emotions ................................ ................................ ................................ .......... 74 Behaviors ................................ ................................ ................................ ......... 75 Control Variables ................................ ................................ .............................. 76 5 RESULTS ................................ ................................ ................................ ............... 78 Analyses Overview ................................ ................................ ................................ 78 Tests of Hypotheses ................................ ................................ ............................... 79 Main Effects of Abusive Supervision on Emotions ................................ ........... 79 Emotion M ediated Pathways ................................ ................................ ............ 79 .............................. 82 Moderating Effects of Personality on Abusive Supervision and Emotions ........ 83 Moderating Effects of Personality on Abusive Supervision and Behaviors ....... 84 Main Effects of Personality on Abusive Supervision ................................ ......... 84 Mediators of the Relationship between Personality and Abusive Supervision 85 6 DISCUSSION ................................ ................................ ................................ ....... 103 Theoretical Implications ................................ ................................ ........................ 105 Limitations and Future Research ................................ ................................ .......... 108 Practical Implications ................................ ................................ ............................ 110 Conclusion ................................ ................................ ................................ ............ 111 APPENDIX A LIST OF ITEMS USED IN INITIAL SURVEY (COWORKER REPORT) ............... 113 B LIST OF ITEMS USED IN INITIAL SURVEY (SELF REPORT) ............................ 115 C LIST OF ITEMS USED IN MONTHLY SURVEY ................................ ................... 116 LIST OF REFERENCE S ................................ ................................ ............................. 119 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 133
7 LIST OF TABLES Table page 5 1 Variance decomposition of Level 1 variables ................................ ..................... 87 5 2 Means, standard deviations and intercorrelations among study variables .......... 88 5 3 HLM regression results f or the effects of abusive supervision on ................................ ................................ ....................... 89 5 4 HLM regression results for the effects of abusive supervision on ................................ ................................ ...................... 90 5 5 HLM regression results depicting emotions as mediators of the relationships ................................ 91 5 6 HLM regression supervision ................................ ................................ ................................ ......... 92 5 7 HLM regression results depicting personality traits as moderators of the relationship between abusive supervis .............. 93 5 8 HLM regression results depicting personality traits as moderators of the .............. 94 5 9 abusive supervision ................................ ................................ ............................ 95 5 10 HLM regression results for t emotions ................................ ................................ ................................ ............. 96 5 11 behaviors ................................ ................................ ................................ ............ 97 5 12 mediators of the relationships between personality and abusive supervision ..... 98 5 13 S ummary of study findings ................................ ................................ ................. 99
8 LIST OF FIGURES Figure page 3 1 Conceptual model ................................ ................................ ............................... 70 5 1 P lot of the moderating effect of trait compassion on the relationship between abusive supervision and state compassion ................................ ...................... 102 5 2 Plot of the moderating effect of trait withdrawal on the relations hip between abusive supervision and supervisor directed avoidance ................................ .. 102
9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FROM NEGATIVE ACT TO NEGATIVE RELATIONSHIP: UNDERSTANDING HOW PATTERNS OF ABUSIVE SUPERVISION EMERGE AND DEVELOP OVER TIME By Lauren S. Simon May 2011 Chair: Timothy A. Judge Major: Business Administration Abusive super to which supervisors engage in the sustained display of hostile verbal and nonverbal definition is the n otion that abusive supervision is sustained ; rather than being occasionally exposed to uncivil acts, subordinates subjected to abusive supervision are repeatedly mistreated. Yet, little is known regarding the processes through which these destructive relat ional patterns emerge. How do generally functional supervisor subordinate relationships, characterized by occasional negative events, transform into dysfunctional relationships consumed by such events? In attempting to help answer this question, I integrat e reciprocal determinism (Bandura, 1978, 1983), victim precipitation (Elias, 1986), and cognitive appraisal perspectives (Lazarus, 1991a, 1991b) to investigate the processes through which sustained abuse occurs and the extent to which personality influence s these processes. may or may not reinforce the likelihood of future abuse thus facilita ting or inhibiting sustained
10 abusive patterns. To test th is model, longitudinal data were collected over the course of a five month period ( six waves) among a sample of 159 employed individuals and their coworkers. Results offered some support for the abov ementioned propositions In particular, subordinates who possessed high levels of trait withdrawal a facet of neuroticism were more likely to endure abusive supervision in general, and also, were more likely to avoid perpetrator s in response to mistreat me nt Moreover, a reciprocal relationship directed avoida nce emerged. Taken together, thi s set of findings suggest s that high trait withdrawal individuals might find themsel ves in abusive relationships in part because their own reactions to mistreatment can knowingly or unknowingly reinforce such behavior. Theoretical and practical implications are discussed.
11 CHAPTER 1 INTRODUCTION Humans have a fundamental need to belong to form lasting friendships replete with positive interaction and mutual concern (Baumeister & Leary, 1995). T he sheer amount of time individuals spend at work coupled with the increased use of teams in organizations (Gordon, 1992) make the workplace a likely breeding ground for suc h relationships. Indeed, a recent Gallup Poll ( Rath, 2006 ) suggests that nearly one third of are benign. Numerous researchers have found evidence to suggest that, in ad dition to cultivating friendship, the workplace can be a major source of interpersonal harm. Particularly common are acts of nonphysical mistreatment (Barling, Dupre, & Kelloway, k, ridicule, and social exclusion. Porath and Pearson (2010), for example, estimated that nearly half of American employees were treated rudely at least once per week in 2005, up from 25% in 1998. Moreover, 96% of employees reported experiencing incivility at least once, and 99% of employees reported witnessing incivility. Likewise, roughly 13.6% of U.S. workers endure an abusive supervisor (Tepper, 2007), and Basch and Fisher (20 00 ) found that, of the myriad events sparking negative emotion in the workplac e, 59% involve d acts of colleagues or management Although scholars have identified many types of interpersonal mistreatment, including abusive supervision (Tepper, 2000), incivility (Andersson & Pearson, 1999), petty tyranny (Ashforth, 1994), bullying (H o el & Cooper, 2001 ), mobbing (Zapf, Knorz, & Kulla, 1996), emotional abuse (Aquino & Douglas, 2003), and social undermining (Duffy, Ganster, & Pagon, 2002), these behaviors share grave consequences. Tepper
12 (2007) estimated that organizations lose $23.8 bill ion annually as a result of the effects of harmful interpersonal behavior on individuals. Among these effects are decreased performance (Harris, Kacmar, & Zivnuska, 2007) and commitment (Aryee, Chen, Sun, & Debrah, 2007), increased stress (Grandey, Kern, & Frone, 2007), absenteeism (Bowling & Beehr, 2006), and even post traumatic stress disorder (Leymann & Gustafsson, each ot her without running t he risk of being taken to court, In this dissertation, I primarily explore one type of interpersonal mistreatment extent to which super visors engage in the sustained display of hostile verbal and nonverbal behaviors, excluding physical contact focus to the supervisor subordinate relationship because evidence suggests ment depend to some degree on who enacts it (Inness, seem particularly important to understand, because the power and control of valuable resources typically afforded to suc h individuals (Magee & Galinsky, 2008) can greatly Perhaps partly for this reason, research examining abusive supervision is quite popular relative to research examining other types of employee mistreatment. Notwithstanding this popularity our understanding of abusive supervision remains fairly limited The majority of studies examining abusive supervision have been cross sectional (Tepper, 2007) Cross sectional studies of abusive supervision provide insight
13 into relations invo lving the aggregated quantity of behaviors that participants remember experiencing over a designated time period. They do not, however, allow researchers to assess growth in the quantity or variety of abusive behaviors, nor do they allow researchers to exa future mistreatment. Additionally, cross sectional studies also do not permit researchers to draw causal conclusions. Thus, although previous research provides valuable the dynamic nature of the supervisor subordinate relationship. This is especially problematic, given that, central to the definition of abusive supervision is the notion of a sustained pattern of behavior that, necessarily, unfolds over time. Another gap in the abusive supervision research involves the role of individual differences. To be sure, some research has examined individual differences in who is targeted for and who is most reac tive to abuse (c.f., Mitchell & Ambrose, 2007; Tepper, Duffy, Henle, & Lambert, 2006 ). Yet, the mechanisms underlying these relationships, as coping strategies for prevent ing further mistreatment, remain to be explored. Taken together, the aforementioned gaps suggest that we have a poor understanding of the dynamic processes through which generally functional supervisor subordinate relationships, perhaps characterized by oc casional negative events, transform into dysfunctional relationships consumed by such events. In the current study, then, I collect longitudinal data to help address these gaps. In doing so, I integrate reciprocal determinism (Bandura, 1978, 1983), victim precipitation (Elias, 1986), and cognitive appraisal (Lazarus, 1991a, 199b) theories and perspectives,
14 behavioral reactions to abuse as well as the overall levels of ab use subordinates experience Additionally, I examine whether reactions to abuse increase or decrease the likelihood of future mistreatment. These findings should shed considerable light on the processe s through which more discrete negative in teractions among supervisors and subordinates potentially transform into negative relationships, while simultaneously addressing calls to incorporate the role of time into applied research design (Koslowski, 2009). Using the results of this research, it is my hope that researchers and practitioners will be better able to develop interventions effective in preventing abusive relationships in the workplace.
15 CHAPTER 2 LITERATURE REVIEW Abusive Supervision Literature In order to understand how patterns of abu sive supervision unfold, it is useful to begin by reviewing the literature on abusive supervision. In doing so, I first thoroughly define abusive supervision. I then compare abusive supervision to other popular mistreatment concepts. Following this, I revi ew the antecedents and consequences of abusive supervision, and conclude by discussing the gaps in this work. Abusive Supervision Defined In his seminal work on the topic, Tepper (2000) coined the term abusive tions of the extent to which supervisors engage in the sustained display of hostile verbal and nonverbal behaviors, excluding physical contact erceptions, abusive supervision is subjective, based on the perspective of the individual potentially experiencing mistreatment. Abusive supervision is also enduring in that it is likely to continue until the supervisor subordinate relationship is terminat ed or until the supervisor modifies his or her behavior. Finally, abusive supervision involves behaviors enacted with a purpose in mind, although, the purpose is not necessarily to harm subordinates. For example, supervisors may engage in abusive behaviors in an effort to increase subordinate s performance. Examples of behaviors characteristic of abusive supervision include, others, and being rude or lying to a subordinate. Mo re physical behaviors, such as shoving, do not fall within the conceptual realm of abusive supervision.
16 Related Constructs Numerous labels have been used to refer to mistreatment at work, including abusive supervision (Tepper, 2000), incivility (Andersson & Pearson, 1999), petty tyranny (Ashforth, 1994), bullying (Hoel & Cooper, 2001), mobbing (Zapf et al., 1996), emotional abuse (Keashly, 1998), general hierarchical abuse (Rospenda, Richman, Wislar, & Flaherty, 2000) victimization (Aquino, 2000), workplace harassment (Bjrkqvist, sterman, & Hjelt Bck, 1994), identity threat (Aquino & Douglas, 2003), and social undermining (Duffy et al., 2002). I do not extensively compare and contrast each of these behaviors individually, as this has been done comprehens ively elsewhere (e.g., Aquino & Thau, 2009; Einarsen, 2000; Pearson, Andersson, & Porath, 2005; Tepper, 2007). Instead, I highlight general similarities and differences between abusive supervision and other mistreatment constructs, and provide detailed des criptions only for those constructs that bear the closest resemblance to abusive supervision. Two frameworks are particularly useful in assisting with this undertaking. First, Tepper (2007) compared the definitions of several mistreatment constructs using four dimensions: whether mistreatment is restricted to being directed downward from supervisor to subordinate, whether mistreatment can involve sexual or physical hostility, whether non hostile behaviors can be included, and whether the definition refers to directed downward and excludes physical and sexual hostility, non hostile content, and references to intended outcomes, other constructs evaluated by Tepper (2000) di ffered from abusive supervision on at least one dimension. In a similar effort, Aquino and Thau (2009) identified three broad dimensions and seven sub dimensions in order to compare mistreatment constructs. These dimensions
17 involve the type of needs (physi ological or psychological) thwarted by mistreatment, whether mistreatment consists of behaviors that directly (overtly) or indirectly (covertly) harm victims, and the status of the perpetrator (i.e., higher than victim, coworker, or lower than victim). Abu sive supervision was classified as thwarting psychological, rather than physiological needs, as consisting of both direct and indirect harm, and as being enacted by a superior toward a subordinate. Like Tepper (2007), Aquino and Thau (2009) found abusive s upervision to differ from every other construct examined on at least one dimension. Thus, abusive supervision appears to possess unique properties that conceptually distinguish it from other mistreatment constructs. Notwithstanding this differentiation, sc holars have questioned whether each of the abovementioned constructs necessitates separate theoretical and empirical examination, or whether the literature would be better served by the adoption of a broader perspective (Aquino & Thau, 2009). Consistent wi th the latter viewpoint, aggression (Neuman & Baron, 2005), counterproductive behavior (Spector & Fox, 2005), and victimization (Aquino & Thau, 2009) have each been nominated as potential of these constructs encompass multiple mistreatment behaviors. However, consensus regarding which, if any, of these labels is appropriate has not yet been established, nor has consensus regarding whether aggregation is, in fact, appropriate at all. Recogn izing this, many recent works opt for middle ground by focusing on specific, narrow constructs, while simultaneously drawing from literatures involving when reviewin g the literature. That is, in addition to reviewing the literature focusing
18 explicitly on abusive supervision, I also incorporate findings involving constructs closely related to abusive supervision. Tepper (2007) identified two such constructs in his rece nt review: supervisor undermining and supervisor aggression. Additionally, because very few studies examine antecedents to abusive supervision, Tepper (2007) advocated drawing from the workplace victimization literature, which has disproportionately focuse d on examining antecedents to mistreatment. In the following sections, I discuss these constructs in detail, and compare and contrast them to abusive supervision. Supervisor undermining Supervisor undermining is part of the broader social undermining cons establish and maintain positive interpersonal relationships, work related success, and favorable reputation undermining specifically, wherea s abusive supervision does not ascribe intentions to hostile behavior, supervisor undermining is a slow, enduring, and intentional attempt to harm a Supervisor aggression Like supervisor undermining, supervisor aggression is subsumed under a larger construct workplace aggression. Workplace aggression can or workers in a work 6). When abusive supervision and supervisor aggression. More specifically, supervisor
19 aggression includes physical hostility and requires harmful intentions on behalf of a p erpetrator, whereas abusive supervision does not. Aquino and Thau (2009) did not include aggression in their comparison of mistreatment constructs, though, if they had, supervisor aggression would differ from abusive supervision in its ability to thwart bo th psychological and physiological needs, rather than solely psychological needs. Victimization Compared to supervisor undermining and supervisor aggression, victimization has less in common with abusive supervision. Tepper (2007) noted that victimization differs from abusive supervision on three dimensions. Namely, victimization is not exclusively directed downward from supervisor to subordinate, it can include physically hostile acts, and its definition references intended outcomes. Victimization also di dimensions, as it can thwart both physical and psychological needs, and perpetrators are not confined supervisory roles. Conclusion After comparing the key features of mistreatment construc ts using existing frameworks, it is clear they share much in common. This is especially true for the abusive supervision, supervisor undermining, and supervisor aggression constructs. Additionally, it is worth noting that items typically used to measure va rious mistreatment constructs overlap considerably. Thus, consistent with Tepper (2007), I include research examining abusive supervision, supervisor undermining, and supervisor aggression in my review of the literature. Also in line with Tepper (2007), I draw from the workplace victimization literature in my discussion of the antecedents to abusive supervision, as there is a dearth of research in this area.
20 Review of Empirical Findings Antecedents Characteristics of targets, perpetrators, and the work envi ronment have each been examined as antecedents of abusive supervision and like constructs. I discuss findings involving each of these categories in the sections to follow. Target characteristics With few exceptions (e.g., Martinko, Harvey, Sikora, & Dougl abusive supervision have been explained using victim precipitation theory. Originating in the criminology literature (Amir, 1967; Curtis, 1974; Wolfgang, 1967), and introduced t o the management literature in a series of studies conducted by Aquino and his colleagues (e.g., Aquino, 2000; Aquino & Bradfield, 2000; Aquino, Grover, Bradfield, & Allen, 1999), victim precipitation theory posits that victims may knowingly or unknowingly contribute to their becoming targets of mistreatment. That is, individuals can become attractive targets for abuse because they possess characteristics or behave in ways that make them appear vulnerable to or deserving of mistreatment. In pursuit of learn mistreatment in the workplace, researchers have linked a number of characteristics to abusive supervision, supervisor undermining, and, most frequently, to victimization. Among the factors positivel y related to these behaviors are negative affectivity (Aquino & Bradfield, 2000; Aquino et al., 1999; Tepper et al., 2006), aggressiveness (Aquino & Bradfield, 2000), voicing discont ent (Kim, Rosen, & Lee, 2009), avoiding, obliging, and integrating conflic t management styles (Aquino, 2000), and external (hostile) attributional style (Martinko et al., 2009).
21 Additionally, several moderators of these relationships have been discovered. One For example, Aquino (2000) found that individuals occupying lower status jobs were indirectly (i.e., covertly) victimized more often than individuals occupying higher status jobs when they managed conflict with low regard for their own interests and high interests (i.e., when they possessed an obliging conflict management style). Conversely, having an integrating conflict management style, characterized by a high imization only for individuals who possessed high status jobs. Also examining the moderating role of job status, Aquino et al., (1999) found that negative affectivity was related to indirect (covert) victimization solely for individuals who occupied lower status jobs. In sum, although the findings are not particularly straightforward, the results of these studies collectively suggest that job status is important to consider when investigating mistreatment relationships with other variables. Two additional studies have also examined moderators. Kim et al., (2009) found that managers were more likely to engage in social undermining when cynical (rather than trusting) employees voiced discontent with company policy. This relationship was mediated by the motiv es supervisors attributed to employees. Specifically, supervisors concerns. In a final study examining moderation, Martinko et al., (2009) found that a external attributions on abusive supervision. Moreover, the effects of external
22 attributions on ab usive supervision were mediated by poor leader member exchange relationships. At this point, it is worth noting that each of the antecedents discussed thus far has been positively linked to abusive supervision and/or other closely related constructs. Some researchers, however, have identified factors that shield individuals from mistreatment. One such factor is core self evaluations (CSE; Wu & Hu, 2009), or general self concept (Judge, Locke, & Durham, 1997). In line with self consistency theory (Lecky, 194 5), Wu and Hu (2009) argued that high CSE individuals process social information in a manner consistent with their positive self concept. Thus, compared to low CSE individuals, high CSE individuals are more attentive to positive stimuli and less attentive to negative stimuli. As a consequence of allocating attentional resources in this fashion, high CSE individuals perceive less mistreatment than their low CSE counterparts. Like CSE, self determination, defined as the experience of a perceived locus of caus ality (Deci & Ryan, 1985), has also been negatively linked to victimization (Aquino et al., 1999). Aquino et al. (1999) argued that self determined individuals are empowered to control when, how, and with whom they interact, and thus, to avoid exchanges wi th suspected perpetrators. Additionally, Aquino et al. (1999) reasoned that the sense of control afforded to self determined individuals better empowers them to defend themselves and, therefore, detracts from their appearing vulnerable or helpless. Another factor examined for its ability to protect individuals from victimization is job status (Aquino 2000; Aquino & Bradfield, 2000; Aquino et al., 1999; Aquino & Bommer, 2003). Typically, researchers have proposed that the formal power characteristic of high
23 status positions better equips individuals occupying these positions to retaliate against aggressors. Fear of retaliation, in turn, deters less powerful individuals from directing hostility toward those possessing high status jobs. Interestingly, despite t he seemingly intuitive nature of this argument, findings regarding job status have been mixed; only one (Aquino, 2000) of four studies ( Aquino 2000; Aquino & Bradfield, 2000; Aquino et al., 1999; Aquino & Bommer, 2003) examining the job status victimizat ion relationship has found job status to be negatively related to victimization, and this relationship held only for mor e overt forms of victimization. In a final study examining factors that detract from victimization, Aquino and Bommer (2003) found subor related to perceived victimization. They argued that, by engaging in citizenship behavior, employees create positive reciprocity norms, thereby obligating others to favorable treatment. Race and job status also moderated the organizational citizenship behavior victimization relationship, such that it was stronger for White than for African status indicators like race and job st atus facilitate the creation of stronger reciprocity norms, however, possessing low job status was actually found to amplify the negative effects of organizational citizenship behavior on victimization. To explain this finding, Aquino (2000) suggested that high status employees are typically perceived as so cially attractive (Georgesen & Harris 1998), and thus, should be viewed positively regardless of whether or not they engage in citizenship behavior. This positive image, in turn, should make high status individuals less susceptible to victimization irrespective of whether or not they are good citizens.
24 In addition to factors that either increase or decrease victimization, researchers have also identified one trait that is curvilinearly related to victimiz at ion. Specifically, Aquino and B y r on (2002) found that males possessing high and low levels of dominance were victimized more often than males who possessed moderate levels of dominance One explanation of this finding is that individuals who lack appropr iate levels of dominance come across as weak or vulnerable, whereas overly dominant individuals provoke conflict and appear hostile; instilling these unfavorable impressions upon others can result in increased victimization. Supervisor and work environment characteristics Relative to research examining characteristics of targets, few er studies have focused on identifying hostility eliciting characteristics of perpetrators and the work environment. Researchers who have focused on these characteristics have tended to develop integrative models depicting supervisor characteristics as moderators or mediators of the effects of work environment features on abusive supervision. Quite frequently, the relationships contained with in these models are explained as mani festations of displaced aggression (e.g., Aryee et al., 2007; Hoobler & Brass, 2006; Wayne, Hoobler, Marinova, & Johnson, 2008; Tepper et al., 2006). According to this perspective, dissatisfying features of the work environment produce negative emotions th at leave supervisors ready and willing powerful individuals makes less influential subordinates opportune targets against which to aggress. Consistent with this explanation, Wayne et al. (2008) found that supervisors who were abused by their manage rs became cynical which, in turn, resulted in their abusing
25 subordinates. In another study, Hoobler and Brass (2006) found that supervisor s who experienced psychological contract vio lations were more likely to abuse subordinates, though this relationship was stronger for supervisors who possessed a tendency to infer hostile attribution bias ) Several studies have also examined supervisor onal justice as an antecedent of positively related to abusive supervision among a samp le of National Guard employees. stronger when subordinates were high in negative affectivity, perhaps because subordinates possessing high levels of negative affect are more submissive or Contra ry to the findings of Tepper et al. (2006) however, a more recent study by Rafferty, Restobug, and Jimmieson (2010) found that supervisors levels of interactional but not procedural justice predict ed subordinate s perceptions of abusive supervision. This effect was amplified when supervisors experienced higher levels of psychological distress. A final study investigating supervis Aryee et a l., 2007) found that supervisors who possessed authoritarian leadership style s were more likely to abuse their subordinates. Furthermore, authoritarian leadership style moderated the relationship between supervisor s perceptions of interactional justice an d abusive supervision such that interactional justice was positively related to abusive supervision only for auth oritarian leaders. Aryee et al. (2007) reasoned that abuse satisfied authoritarian a uthoritarian lea ders were less
26 able to manage their emotions when experiencing interactional injustice resulting in their lashing out at subordinates In sum, finding s involving procedural justice offer mixed support for its role as an antecedent to abusive supervision, whereas findings involving interaction al justice are more consistent ly supportive. I t does appear important however, to consider moderators when assessing these relationships. T he most recent study to investigate antecedents to abusive supervision linked (Kiazad, Restobug, Zage nczyk, Kiewitz, & Tang, 2010) using samples of Australian and Pilipino employees. In both samples, the effect of Machiavellianism on perceived abusive supervision was fully mediated by authoritarian leadership. Finally, based self esteem evaluation s of their meaningfulness and worthiness within the organization ( Pierce, Gardner, Cummings, & Dunham, 1989) moderated the relations hip between authoritarian leadership and abusive supervision, such that employees who possessed lower levels of organization based self esteem were more like ly to perceive authoritar ian leadership as abusive Kiazad et al. (2010) reasoned that their findi ngs supported dispositional rather than displaced aggression explanations for abusive supervision. Evaluative summary of antecedents Overall, there is a dearth of research examining antecedents to abusive supervision when the scope of interest is li mited strictly to is willing to venture into closely related literatures (i.e., victimizatio n) the research is more plentiful. B y linking certain personality traits and behaviors to mistr eatment, findings from both literatures cumulatively offer support for the victim precipitation
27 model. The traits and behaviors examined as antecedents to mistreatment, however, do remain somewhat limited. Consequently, research investigating how other tar get characteristics relate to mistreatment would be a welcome addition to the abusive supervision literature. Compared to the literature examining characteristics of the targets of abuse, less research has focused on the perpetrator and work environment ch aracteristics that elicit mistreatment. Existing findings seem consistent with a displa ced aggression explanation supervisors subjected to unfavorable working conditions react by abusing subordinates b seems to support supervisor trait based explanations (e.g. Kiaz ad et al., 2010). M ore research is needed however, to draw firm conclusions in both area s Additionally, the presence of several moderators in research examining all categories of mistreatment antecedents implies that it is especially important to consider boundary conditions when investigating antecedents to abusive supervision. Consequences The literature examining consequences of abusive supervision is quite voluminous. The most common ly stu died outcomes include job satisfaction and related attitudes (i.e., job involvement) ; stress and strain ; withdrawal cognitions attitudes, and behaviors; and extra role performance (i.e., counterproductive and citizenship behaviors) In the sections to fol low, I review consequences representing each of these categories, among others less studied. Abusive supervision has been linked to decreased job satisfaction (Breaux, Perrewe, Hall, Frink, & Hochwater, 2008; Duffy, Ganster, Shaw, Johnson, & Pagon, 2006; Schat et al., 2006; Tepper, 2 000; Tepper, Duffy, Hoobler, &
28 Ensley, 2004), job involvement (Duffy et al., 2006) and life satisfaction (Tepper, 2000). vision was also negatively related to thesis project satisfaction, though only for those students who did not receive social support from their colleagues (Hobman, Restubog, Bordia, & Tang, 2009). In seeking to explain the abusive supervision job satisfa ction relationship, Tepper (2000) examined the role of organizational justice and found evidence for full mediation. Also in support of a fairness related explanation, Duffy et al. ed the negative relationships between abusive supervision and job satisfaction and job predictions, Duffy et al. (2006) argued that singled out subordinates were more likely hold their supervisors accountable for abusive behavior, and thus, to react more negatively. A number of withdrawal related constructs, including commitment (Aryee et al., 2007; Duffy et al., 20 02; Schat et al., 2006; Tepper, 2000; Tepp er et al., 2004 ; Tepper, Henle Lambert, Giacalone, & Duffy, 2008), turnover intentions ( Burris Detert, & Chiaburu, 2008 ; Duffy Ganster, Shaw, Johnson, & Pagon, 2006; Harvey, Stoner, Hochwater, & Kacmar, 2007; Sc hat et al., 2006 ; Wayne et al., 2008 ) and voluntary turnover ( Tepper, 2000 ) have been linked to abusive supervision though not all findings have been consistent. For instance, contrary to some findings (e.g., Aryee, 2007; Tepper, 2000; Tepper et al., 200 4; Tepper et al., 2008), Tepper et al. (2004) did not find abusive supervision to be linked to affec tive commitment among a random sample of Midwestern U.S. employees, nor did
29 Burris et al (2008) in a sample of restaurant employees As with job satisfacti on, researchers have found the effects of abusive supervision on organizational commitment to be mediated by organizational justice (Aryee et al, 2007; Tepper, 2000) and, beyond that, to be exacerbated when abusive supervisors simultaneously exhibit suppor tive behaviors (Duffy et al., 2002 ). Likewise, the abusive supervision voluntary turnover relationship has also been found to be mediated by organizational justice (Tepper, 2000) and to be exacerbated when subordinates feel they are singled out for abuse (Harvey et al., 2007). One of the most studied outcome s of abusive Indeed, a busive supervision has been positively associated with these outcomes in many forms, including anxiety ( Gant Nagda, Brabson, Jayaratne, Chess, & Singh, 1993; Hobman et al., 2009; Tepper, 2000 ; Tepper, Moss, Lockhart, & Carr, 2007 ) emotional exhaustion ( Aryee, Chen, & Debrah, 2008; Breaux et al., 2008; Grandy & Kern, 2004; Gr andey et al., 2007; Tepper et al. 2007; Harvey et al 2007 ; Tepper, 2000; Wu & Hu, 2009 ), ego depletion (T hau & Mitchell, 2010), job tension (Breaux et al., 2008; Harvey et al., 2007), somatic complaints (Duffy et al., 2000 ; Gant et al., 1993), depression (Duffy et al., 2006 ; Gant et al., 1993; Tepper, 2000), work family conflict (Tepper, 2000), psychological and physi cal health, (Schat et al., 2006), psychological distress ( Rafferty et al., 2010; Yagil, 2006), job strain ( Harris et al., 2005), irritability (Gant et al., 1993), depersonalization (Gant et al., 1993), intrusive thoughts (Thau & Mitchell, 2010) n egative affec t (Yagil, Ben Zur, &
30 Tamir, 2011 ) and psychological well being ( Hobman et al., 2009) Researchers have also identified a number of factors that serve to either exacerbate or mitigate stress. Among the stress exacerbating factors are superviso r suppor t (Duffy et al., 2002 ; Hobman et al., 2009 ), susceptibility to emotional contagion (Wu & Hu, 2009), being singled out for abuse (Duffy et al., 2006), self esteem (Rafferty et al. 2010) and regulative maintenance strategies (i.e., stretching the t ruth to avoid problems; Tepper et al., 2007). Conversely, variables that have been shown to buffer stress include job mobil ity (Tepper, 2000), subordinate s power (Grandey & Kern, 2004), direct maintenance communication strategies (e.g., communication effo rts designed to convey relational expectations, and openly discuss problems with supervisors; Tepper et al., 2007), and 2007), though, the effects of ingratiation have been found to be more effective for individuals who possess higher levels of dispositional positive affectivity (Harvey et al, 2007). Coworker support has also been examined as a potential buffer to stress, but findings have been mixed. That is, coworker support has sometimes been found to b uffer stress (Duffy et al., 2002 ; Hobman et al., 2009) and has other times been found to exacerbate it (Wu & Hu, 2009). Collectively, the abovementioned findings suggest that many additional factors can affect the abusive supervision stress relationship. However, despite the multitude of research examining the moderators of this relationship, little is known regarding the mechanisms through which abusive supervision impacts stress; to date, organizational justice (Tepper, 2000) and support seeking and avo idant coping strategies (Yagil et al., 2011 ) remain the only known mediator s. It should be noted that Yagil et al. (2011 ) also
31 examined whether three other coping s trategies directly communicating supervisor, ingrat iat ion, and reframing mediated the abusive supervis ion negat ive affect relationship. Althoug h abusive supervision positively predicted ingratiation and reframing, none of the three coping strategies emerge d as significant mediators. e Counterproductive behaviors include volitional acts that harm or are intended to harm organizations or individuals in organizations (Spector & Fox, 2005), whereas deviance similarly defines volitional behavior that violates organizational norms and harm s the organization or its counterproductive behaviors have primarily been studied as outcomes to abusive supervision based on the premise that subordinates seek to retaliate u sing such behaviors when they are the victims of abuse. However, w hen subordinates fear retaliating against a perpetrator or when the perpetrator is unavailable, they may direct their hostility to less threatening targets, including fellow coworkers. Where as more retaliatory explanations are consi stent with the concept of reciprocity (Blau, 1964), aggressing against less powerful others who are not perpetrators is consistent with the concept of displaced aggression. Supporting these perspectives, abusive su pervision has been positively linked to deviance and/or counterproductive behavior directed toward supervisors (Dupre, Inness, Connelly, Barling, & Hoption, 2006; Inness, Barling, & Turner, 2005 ; M itchell & Ambrose, 2007; Tepper, Carr, Breaux, Geider, Hu, & Hua, 2009; Thau, Bennett, Mitchell, & Marrs, 2009 ; Thau & Mitchell, 2010 ), coworkers (Mitchell & Ambrose, 2007; Schat et a l., 2006 ; Schaubhut, Adams, & Jex, 2004; Thau et al., 2009), and the org anization (Duffy et al., 2002; Duffy et al., 2006; Mitchell &
32 Ambrose, 2007; Tepper et a l., 2009; Thau & Mitchell, 2006; Schaubhut et al., 2004; Tepper et al. 2008; Thau et al., 2009), as well as to counterproductive behavior at the un it leve l (Detert Trevi o, Burris, & Andiappan, 2007). As with stress and strain several moderators of these relationships have been identified. For example, Mitchell and Ambrose (2007) found that negative reciprocity beliefs beliefs that negative acts should be reciprocated (Cropanzano & Mitchell, 2005) amplified the relationship be tween abusive supervision and supervisor directed deviance (but not coworker directed or organizational directed deviance). Likewise, Tepper et al (2009) found that abusive supervision was more strongly related to organization when an employee intended to leave the organization. The authors reasoned that power differences often characteristic of supervisor subordinate dyads were reduced for individuals who intended to leave the organization because allocation of the organizational resources would not remain under the discretion of the current supervisor. Other factors also amplify the abusive supervision counterpro ductive behavior relationship. Th au et al. (2009) found that the e ffects of abusive supervision on workplace deviance were exacerbated under conditions of high uncertainty (i.e., when a supervisor had a less authoritative management style or when perceptions of management actions were seen as uncertain). Consistent wit h uncertainty management theory (Lind & Van den Bos, 2002), the authors argued that when environmental uncertainty is high, employees pay more attention to fairness related information such as treatment they receive from authorities in order to help make s ense
33 of the environment Accordingly, abusive supervision should become more salient in uncertain environments, and thus, should result in more negative reactions. Working largely for financial reasons (Dupre et al., 2006), supervisor support (Duffy et al 2002), and being singled out (Duffy et al., 2006) for abuse also amplify the abusive supervision organizational deviance relationship. One factor working primarily for reasons having to do with personal fulfillment has actually been shown to buffer abu sive supervis i directed deviance (i.e., aggression; Dupre et al., 2006), most likely because individuals working primarily to obtain personal fulfillment are less dependent on their employers compared to employees who must remain e mployed for financial reasons. As a result, the former group should have more freedom as to whether to endure an abusive environment or to exit (Dupre et al., 2006) In addition to the more situational moderators discussed above, two individual differences Scha ubhut et al. (2004) found that individuals with high self esteem were more likely to engage in interpersonal and organizational deviance when faced with abusive supervisors presuma bly because abusive supervision is more thre atening to individuals who are used to maintaining a positive self concept than for those who already possess low self esteem. Similarly, Thau and Mitchell (2006) found that individuals possessing more of a dispo sitional need to maintain their self worth (i.e., trait validation seeking) were more likely than those who did not possess this need to engage in organizational deviance in response to an abusive supervisor. In an attempt to better understand the mechanis ms und erlying the abusive supervision deviance relationship, Tepper et al. (2008) examined the potential
34 mediating role of affective commitment These authors argued that affective commitment would mediate abusive s effects on organizational deviance because abusive supervisors are unlikely to make subordinates feel as though they are valued members of the organization. As a result, these subordinates are unlikely to develop a sense of attachmen t to the organization and should be less hesitant to engage in organi zational deviance. Consistent with this li ne of reasoning, results revealed that affective commitment mediated the effects of abusive supervision on organizational deviance, and that this effect was stronger when employees perceived the ir coworkers to be more approving of deviant behaviors and when coworkers themselves reported engaging in more deviant behaviors. The most recently published study (Thau & Mitchell, 201 0) involving the abusive supervision deviance rel ationship offers a f resh explanation for the relationship among the two variables. The authors argued that self regulation impairment might explain the relationship between abusive supervision and deviant behaviors because the experience of abuse often results in victims spe nding substantial amounts of time attempting to comprehend and interpret the causes and consequences of their resulting in lowered inhi bitions and, thus, a prone ness to counte r normative behavior such a s deviance. To test whether self regulation impairm ent or the more traditional self gain perspective which argues that subordinates retalia te because it balances the socia l exchange and might deter future harm better explains the abusive supervision deviance relationship Thau and Mitchell (2010) use d distributive justice as a moderator of the relationship between abusive supervision and various types of deviance (i.e.,
35 organizational deviance, supervisor directed devia nce, and antisocial behavior) among three samples I f distributive justice amplified the abusive supervision deviance relationships, they reasoned, then results would support the self regulation impairment pers pective because distributive justice is incons istent with abusive treatment and would require more resources for cognitive processing Thus, subordinates would have few resources available for maintain ing appropriate behavior. On the other hand, if distribu tive justice attenuated the relationship s between ab usive superv ision and deviance ed distributive justice a s a fair reward that compensated them for the costs of abuse. Across the three studies, results tended to be more supportive of the self regulation impa irment view, in that the effects were typically amplified when distributive justice was high. There also appeared to be a m oderated mediation effect whereby distributive justice amplified the extent to which ego depletion and intrusive thoughts mediated the abusive s upervision deviance relationships. Because there are often repercussions for engaging in counterproductive behavior regardless of who is targeted, subordinate s may sometimes retaliate against abusive supervisors by engaging in behaviors that are through which subordinates can convey nonconformity (Tepper, Duffy, & Shaw 2001 ). In investigating whether or not employees do, in fact, engage in such behaviors, Tepper of constructive resistance behaviors, such as requesting clarification on a n assignment d ysfunctional
36 passive aggressive resistance behaviors such as acting like one did not hear or is too busy to complete a request Additionally, Tepper et al. ( 2001 ) found that individuals who were less conscientious and less agreeable were more likely to respond to abusive supervision using these destructive resistance behaviors. Conversely, the positive relationship between abusive supervision and constructive resistance was s tronger among subordinates who were more conscientious. Bamberger and Bacharach (2006) found a similar moderating effect for the abusive supervision problem drinking (i.e., alcoholic drinking) relationship. Specifically, they found abusive supervision to be less strongly associated with problem drinking for individuals who were both more agreeable and more conscientious. It is interesting to note that, although problem drinking is often portrayed as a dysfunctional response to stress, Bamberger and Bachar ach (2006) found no evidence that somatic stress mediated the abusive supervision problem drinking relationship. Thus, Bamberger and Bacharach (2006) concluded that problem drinking better represented a izenship behaviors Organizational citizenship behaviors include discretionary behaviors that are beneficial to the organization and that are not explicitly recognized by the formal reward system (Smith, Organ, & Near, 1983). Several studies have found abusive supervision to be negatively related to various al., 2008; Burris et al., 2008; Hmiele ski & Ensley, 2007; Wayne et al., 2008; Zellars, Tepper, & Duffy, 2002). Moreover, consistent with findings identifying the mechanisms underlying many other abusive supervision outcome relationships, organizational
37 justice has been shown to mediate this effect (Aryee et al., 2007; Zellars et al., 2002). al., 2008), emotional exhaustion (Aryee et al., 2008), and psychological detachment (Burris et al., 2008). The abusive supervision citizenship relationship is exacerbated when the presumably because organic organizations possess decentralized control and authority, reducing the power imbala nces between supervisor and subordinate. This reduced power imbalance should, in turn, minimize the tendency for supervisors to become overbearing and, ultimately, abusive. Also amplifying the effects of abusive supervision aviors, along with the extent to which organizational justice mediates the abusive supervision citizenship behavior relationship, is the degree to which subordinates believe citizenship behavior is not within their set of formal job requirements (Zellars et al., 2002). As with resistance behaviors, withholding citizenship behavior is thought to be a safer method of retaliation than engaging in counterproductive behavior or decreasing task performance. Consequently, when subordinates believe citizenship be havior to be a job requirement, they are unlikely to purposely withhold it when mistreated. I n the only study that has examined the effects of abusive supervision at the team level, Hmieleski and Ensely (2007) found that environmental uncertainty attenuate d These authors argued that in uncertain or dynamic environments, abusive beha viors should be attributed to the stressful environmental conditions rather tha n to the disposition of the
38 le ader. By alleviating supervisors from blame for abusive behavior, subordinates should perceive abuse as less unfair, and thus, should react less negatively It should be noted that this finding is somewhat at odds with that of Thau et al. (2009), who argue d and found support for the notion that the effects of abuse would be amplified in uncertain environments. Findings involving wo studies have focused on performance at the firm or team level. Whereas Detert et al. (2007) found abusive supervision to be unrelated to firm performance, Hmieleski and Ensley (2007) found abusive supervision to be negatively related to firm performance As with citizenship behavior, environmental uncertainty attenuated this relationship. In the o nly study to examine individual level task performa nce as a substantive variable, Harris et al. (2007) found abusive supervision to be negatively related t o bot h supervisor ratings and formal appraisals of job performance among a sample of automotive workers However, this effect existed only for individuals wh o derived a high degree of meaning from their work Moreover, the moderating effect of the meaning also held for self ratings of job performance despite the absence of a main effect of abusive theory, Harris et al. (2007) argued that abusive supervision could require subord inates especially those who derive meaning from their work to invest valuable time and effort into managing their relationships with supervisors. These relationship maintenance efforts can tax subordinates resources, leaving fewer available to contribute t o adequate job performance. Harris et al (2007) also offered a so cial
39 exchange based explanation, whereby employees decrease their performance in order to retaliate for abuse In a final study, Hoobler and Brass (2006) found that abusive supervision was n egatively related to subordinates performance, though in this study, performance was relegate d to the status of control variable. concept A handful of studies have investigated abusive conce pts. Specifically, Duffy et al. (2002 ) that supervisor support exacerbated this relationship. Additionally, Hobman et al. (2009) and Burton and Hoobler (2006) both fo und abusive supervision to be negatively related esteem, though Burton and Hoobler (2006) found this effect to be self esteem is traditionally g arnered more from the opinions of others and less from achievements, whereas the opposite is true for males. Vulnerability to others opinions, esteem than for esteem. Trust a nd justice supervisor (Duffy & Ferrier, 2003; Duffy et al., 2006) as well as to procedural (Tepper, 2000; Zellars et al., 2002), distributive (Tepper, 2000), and interactional justice (Aryee et a l., 2007; Tepper, 2000). Typically, however, organizational justice has been conceptualized as a mediator in studies linking abusive supervision to other outcomes. Miscellaneous A small number of studies have examined outcomes that cannot be neatly classi fied into any of the previously mentioned categories. For example, Hoobler and Brass (2006) found that subordinates seemed to displace aggression
40 another study, Tepper et al. (2007) examined the relationship between abusive found that individuals with abusive supervisors were more likely than their non abused counterparts to use regulative main tenance strategies consisting of behaviors such as stretching the truth to avoid problems and talking superficially when communicating with supervisors, whereas the use of more direct communication strategies was not directly related to abusive supervision Abusive supervision has also been positively irrit ation (Schat et al., 2006), cynicism (Wayne et al., 2008) and insomnia (through psychological d istress; Rafferty et a l., 2010), but has been found to be unassociated with positive affect (Yagil et al., 2011 ). T wo studies have portrayed abusive supervision as a moderator. In a longitudinal study, Tepper et al. (2004) found that abusive supervision moderated the effects of months later, even after controlling for initial levels of commitment and job satisfaction. were negatively related to employee job satisfaction and were unrelated to commitment. citize nship behaviors were likely to be attributed to less altruistic motives when supervisors were abusive. Supporting this explanation, a follow up study showed that ed this interaction. In a
41 fin al longitudinal study Breaux et al. (2008) found that accountability predicted exhaustion, and job satisfaction only when supervisors were abusive. Evaluative summary of co nsequences After reviewing the literature examining the consequences of abusive supervision, it is clear that abusive supervision covaries with a myriad of undesirable states and situations. It is also clear that a number of factors moderate these relatio nships for some individuals, the effects of abusive supervision seem less detrimental than for others. Notwithstanding these findings, a number of areas deserve further investigation. Three such areas that might be well served by future research are non wo rk, group, and supervisor outcomes of abusive supervision. Less than a handful of studies work lives. Moreover, no studies have examined how abusive supervision affec ts supervisors who behave abusively. This seems like an interesting area of investigation. Given the negative effects of abusive supervision on subordinates, it would seem that supervisors would derive little instrumental benefit from acting abusively. Eve counterproductive behavior an d stress, would counteract any positive effects. In this
42 I now turn to discussing limitations of the existing literature that are not specific to the consequences of abusive supervision, but are better discussed as limitations of the abusive supervision literature in its entirety. Limitations of the Existing Literature Collectively, this review of the existing literature highlights several theoretical and empirical gaps in the examination of abusive supervision. Most importantly, models of abusive supervision have overwhelmingly conceptualized abusive supervision as an antecedent to detrimental consequences. Such models are overly simplistic not only because they do not consider antecedent s to abusive supervision, but also because they do not portray abusive supervision as part of a dynamic pattern of social interaction between supervisor and subordinate. In doing so, these models fail to consider how abuse contribute to or detract from initial and continued mistreatment. This problem becomes even more evident once one considers that, when viewed from a victim precipitation perspective, many of the proposed consequences of abusive supervision become pro bable antecedents. For example, whereas Aquino and Bommer (2003) cogently argue that the social attractiveness afforded to individuals who engage in citizenship behavior makes them unlikely to be targeted for mistreatment, Zellars et al. (2002) provide an equally convincing argument as to why subordinates withhold citizenship behavior as a means of retaliating for abuse. Using similar arguments, other variables, inclu ding counterproductive behavior can also easily be envisaged as both antecedents to and co nsequences of abusive supervision. Thus, reciprocal relationships seem likely. That nearly all research involving abusive supervision to date has either been cross sectional or has not controlled for alleged outcomes when longitudinally testing abusive sup ervision
43 outcome relationships, however, makes it impossible to assess whether most variables studied in relation to abusive supervision are truly antecedents, truly consequences, or both. Second, we have a limited understanding of the mechanisms through which abusive supervision might impact subordinates. Studies examining mediators of the effects of abusive supervision have primarily focused on organizational justice. And, although a handful of studies have examined other mechanisms, these efforts can h ardly be described as voluminous. Noticeably absent from the abusive supervision literature, for example, is research examining the potential mediating role of evidence implic ating emotions as fundamental determinants of human behavior, and therefore, most likely of reactions to abusive supervision. A final limitation involves the role of individual differences. Although a plethora of moderators have been linked to abusive supe rvision, in only a few instances have these concept related variables (i.e., self esteem, validation seeking) on the abusive supervision counterproductive behavior rela tionship as well as the buffering effects of conscientiousness and agreeableness on the abusive supervision resistance relationship, however, offers preliminary support for the notion that individuals who possess different personalities can react differe ntly to abuse. If certain individuals do indeed consistently react more dysfunctionally to abuse, it seems likely that these reactions can elicit further abuse, which in turn can elicit further dysfunctional reactions, and so on. Thus in addition to acting in ways that might elicit initial mistreatment as the
44 result in a self reinforcing pattern of mistreatment, whereby a vicious circle of counterproductive supervisor and subordinate interactions emerges.
45 CHAPTER 3 THEORY AND HYPOTHESE S In light of the numerous gaps in the abusive supervision literature, this dissertation intends to make several contributions to abusive supervision theory and research. In order to address the first limitation, I extend research on abusive supervision by longitudinally examining reciprocal relationships between abusive supervision and three types of subordinate behavior : supervisor directed deviance, supervisor directed avoidance, and super visor directed citizenship behavior. I examine reciprocal relationships based on the concept of reciprocal determinism (Bandura, 1978, 1983), people are both producers and (Bandura, 1997, p. 6). More specifi cally, I use the concept of reciprocal determinism as an overarching framework to integrate arguments supporting the notion that subordinate behavior can both cause and be caused by abusive supervision. Concerning the second limitation, I examine three e motion based mediators of anger, fear, and compassion. In doing so, I illuminate how emotional appraisals of abusive behavior as mediators of the abusive supervision subordinate behavior relationship is also consistent with the motivational relational theory of emotion. According to this process based theory, emotional reaction s to an event or stressor depend on whether the event is appraised as being personally relevant to an being (primary appraisal), and whether or not an individual believes he or she can change the conditions responsible for the e vent (secondary
46 appraisal). These emotional reactions, in turn, lead to associated coping efforts and behavioral responses. Lazarus (1991a, 1991b) further suggested that dispositions could influence reactions to stimuli, in part, through their influence on appraisals. Thus, in order to address the final limitation referenced in the literature review, I examine whether tand; if certain individuals are more likely than others to consistently respond to abuse in a manner that further provokes it, then these individuals are at greater risk for potentially setting in motion a vicious cycle of mistreatment, whereby their own reactions to abuse continually reinforce its occurrence. This proposition is also consistent with the central tenet of victim precipitation theory that individuals can knowingly or unknowingly contribute to their own mistreatment (Elias, 1986). The concept ual model guiding this research and integrating the three aforementioned contributions is depicted in Figure 3 1. The remainder of this chapter specifically, I begin by disc ussing the emotional pathways via which abusive supervision influence abusive supervision. Finally, I discuss the extent to which personality influences these processes. Emotion Mediated Pathways In this section, I argue that motiva tional relational
47 theory of emotions. This theory is useful because it not only provides general predictions about the emotion process by listing key variables and the manners in which they interrelate, but also because it offers specific propositions for each discrete emotion. In light of this organizational scheme, I will begin my discussion by first work on more discrete emotions, along with other relevant literature, to garner support for hypotheses involving specific emotions. The Emotion Process an encounter in the environment depend on their cognitive appraisals or their evaluations of the personal significance of what is happening in the encounter. There are two parts to this appraisal process. During primary appraisal individuals determine whether or not they have a stake in an encounter. That is, they determine whether or not an encounter is relevant to thei to, among other things, achieving career related success, or maintaining social relationships or self esteem. If no personal goal is at stake, emotion will not be generated If, however, a personal goal is at stake, individuals further evaluate whether or not the environmental encounter is harmful or beneficial to their goals. Encounters s goals generate positive emotions. The specific emotion experienced depends on the content of the goal that is at stake, and also, on the secondary appraisal process. specifically, during secondary appraisal, individuals determine whether to blame or credit themselves or another, whether anything can be done to alter the situation (if
48 negative) for the better, and, if so, what, and whether the situation can worsen or im prove over time. Each subjective emotional state corresponds to a certain action tendency or state of action readiness that drives a person to behave in a certain way. These behaviors are thought to be biologically driven and relatively automatic and rigid but can, in some instances be suppressed or transformed. Coping what Lazarus (1991a) deemed the can also follow subjective emotional states, and is more complex, deliberate, and planful than are acti on tendencies. Coping can also shape subsequent emotions In applying this framework to abusive supervision, I argue that abusive supervision anger, fear, and lack of co mpassion. The subjective experience of these emotions, in turn, generates certain action tendencies and coping responses. I now turn to discussing these relationships in detail. Anger ger. According to Lazarus (1991a, 1991b), anger is a consequence of blaming another for a esteem Other researchers have conceptualized anger in a similar manner (Allred, 199 9 ; Averill 1983; Gibson & Callister, 2010 ; Roseman, Wiest, & Swartz, 1994 ; Smith & Ellsworth, 1985) and empirical research corroborate s anger as resulting from an individual blaming another for an unjus tified offense (Averill, 1983). Abusive supervision can be inferred to meet the criteria for eliciting anger inducing appraisals because it has been shown to reduce self esteem and thus, to be ego
49 threatening ( Burton & Hoobler, 2006 ) and because it is often perceived as unjustified (e.g., Aryee et al., 20 07), thereby implying that subordinates blame supervisors for abusive behavior Additionally, because abusive supervision can be viewed as an abuse of power in that it violates norms of respect i n the workplace, abused subordinates are also likely to feel moral outrage or what F olger and Skarlicki (2005) term deontic anger. Thus, Hypothesis 1: Abusive s upervision positively predicts Lazarus (1991a, 1991b) argued that each emotion possesses its own innate action tendency. For anger, this action tendency consists of ameliorating the effects of a demeaning offense by attacking the perpetrator (Lazarus, 1991a, 1991b). Consistent with this view, Roseman et al. (1994), found that college students who experienced anger reported feeling like yel ling, hitting, saying something nasty, and wanting to hurt appraisals regarding coping with anger, like the action tendency, typically suggest attack. However, because ins tantly and overtly attacking a perpetrator is unlikely to be successful if it invites further retaliation or social disapproval, individuals often attempt to cope with anger by enacting revenge in a more planful manner. Recognizing this, scholars have ackn owledged that, although reactions to anger share a common motivational goal behavioral manifestations of this goal can be exceedingly diverse. In the words of Roseman et al. se that is too faint, the silent treatment, and air let out of automobile tires may have no physical properties in common, but they can all be recognized as actions with the common goal of getting back at someone toward whom
50 one is angry (p. 218) Thus, rather t predicts more extreme behaviors like physical violence I argue that anger is more likely to predict less intense behaviors, such as gossiping about o r making fun supervisor. It is worth noting that, although subordinates can displace their aggression by lashing out at nonsupervisory targets, reciprocity norms typically imply that, when possible, individuals seek revenge against the source of their anger. This is done, in part, to punish a nd deter offenders from future abuse. In this way, revenge can be viewed as a rational and legitimate response enacted to restore justice and to defend oneself from future mistreatment (Bies & Tripp, 2005). Therefore, in this study, I limit my focus to sub rather than toward coworkers or to the organization in general Hypothesis 2 : Subordinates supervisor directed deviance. The abovementioned arguments positively li nk abusive supervision to anger and anger to increased supervisor directed deviance. A busive supervision has also been empirically linked to increased supervisor directed deviance (Mitchell & Ambrose, 2007; Thau et al. 2009) Together, this theoretical an d empirical evidence suggests that one directed deviance is because abuse angers subordinates and drives them to seek revenge. Thus, I hypothesize that anger partially mediates t he relationships between abusive supervision a nd supervisor directed deviance Partial rather than full mediation is
51 expected because other potential mediators of th is relationship exist, inc luding ego depletion and intrusive th oughts (Thau & Mitchell, 201 0). Hypothesis 3 : abusive supervision and supervisor directed deviance. Fear Another emotio n likely to arise in response to abusive supervision is fear. Lazarus (1991b) defined f ear as a response to a concrete, sudden, and overwhelming physical danger. Although Lazarus (1991b) emphasizes fear as a response to sudden, physical danger, other researchers, including myself, take a more liberal approach and assume that fear can arise f rom less sudden, nonphysical danger (e.g., Kish Gephart, Detert, Trevino, & Edmonson, 2009; Smith, & Ellsworth, 1985). In any case, as a reaction to threat, fear induces a captivating sense of uncertainty regarding when danger will occur and whether or not it can be avoided. Very little research has been conducted on fear in organizations (Kish Gephart et al., 2009). However, logic suggests that abusive supervision can spark both the threat and uncertainty appraisals necessary to induce fear. More specifica lly, subordinates can come to view abusive supervision as threatening to their livelihoods if it causes them to believe they are devalued and thus unlikely to fair well during performance evaluation processes that dictate raises, promotions, and employment Moreover, because abusive blame for something one did not do), subordinates may feel uncertain as to whether future incidences of abuse can be avoided. Taken toge ther, the abovementioned arguments suggest that abusive supervision can invoke appraisals involving both threat and uncertainty, and therefore, has the potential to induce fear.
52 Hypothesis 4: Abusive supervision positively predicts When experiencing fear, individuals possess an innate action tendency to escape or avoid its source (Frijda, 1986 ; Lazarus, 1991 b; Roseman et al., 1994) to move more planfu rather than actually physically running away from him or her, it seems likely that one will instead seek to minimize contact, reserving communication for times only when it is abso lutely necessary; I refer to this response as supervisor directed avoidance. Hypothesis 5: fear positively predicts supervisor directed avoidance Abusive supervision has been positively linked to both avoidant coping strategies (Yagil et al., 2011 ) and the use of regulative maintenance strategies (Tepper et al., 2007) attempts to maintain the supervisor subordinate relationship by distorting messages (e.g., avoiding asking for direction and stretching the truth to avoid problems ) and avoiding contact (Lee, 1998). Moreover, in a review of the incivility literature, Porath and Pearson (2010) found that 63% of individuals surveyed lost time avoiding a perpetrator. These findings suggest a positive relationship between abusive superv ision and supervisor directed avoidance. If, as I have argued, abusive supervision also induces fear, and, if fear results in supervisor directed avoidance, then it seems logical to assume that fear is one mechanism through which abusive supervision resul ts in supervisor directed avoidance. Therefore, I hypothesize that fear partially mediates the abusive supervision
5 3 supervisor directed avoidance relationship. As with anger, partial mediation is expected because factors other than fear, such as disliking Hypothesis 6: f ear partially mediates the relationship between abusive supervisor directed avoidance Compassion Lazarus (1991b) argued that compassion arises from being m suffering and, subsequently, wanting to help. Typically, for compassion to arise, neither the person who is suffering nor the observer can be appraised to be at fault for the distressing circumstance. If the suffering individual is to bla me, the observer is less likely to feel compassion and more likely to feel anger; if the observer is to blame, he or she is likely to feel guilty for causing distress (Lazarus, 1991b). Additionally, feelings of compassion are usually reserved for individua being candidates for subsequent cooperation and reciprocated altruism ( Goetz, Keltner, & Simon Thomas, 2010). As with fear, very little is kn own about the causes and consequences of compassion in the workplace. There is, however, reason to believe that abusive supervision negatively predicts compassion. A long history of research shows that, when making attributions for the behavior of others, individuals have a tendency to over value dispositional explanations and under value situational explanations for observed behaviors (Ross, 19 77). This implies that, rather than viewing supervisors as blameless victims of unprovoked suffering, subordinates are prone to holding supervisors r esponsible for engaging in abusive behavior. As a result compassion for their supervisor is likely to be replaced with more hostile emotions such as anger
54 (Lazarus, 1991b) Of course, if abusive supervision is appraised to be a manifestation of suffering, it is certainly possible for subordinates to feel compassion for abusive supervisors. Nonetheless, attribution research suggests that the typical response is far less altruistic especially when negative ev ents are involved (Malle, 2006). Hypothesis 7: Abusive supervision negatively compassion. Lazarus (1991b) and others (e.g., Goetz et al., 2010) have argued that c ompassion drives individuals to express sympathy and to help those who are suffering (Goetz et al., 2010; ). Consistent with this perspective, meta analyses have shown sympathy to be linked to increased helping beha vior ( Rudolph, Roesch, Greitemeyer, & Weiner, 2004 ). Thus, Hypothesis 8 : compassion positively pred super visor directed citizenship behavior I have already provided arguments for why abusive supervision might be negatively related to compassion and for why compassion should positively predict supervisor directed citizens hip behavior P rior research has also found abusive (Wayne et al. 2008 ) Cumulatively, these theoretical and empirical arguments suggest that one reason subordinate s withhold citizenship behavior f rom abusive supervisors is because they lack compassion for these individuals, and, therefore, the desire to offer help Thus, I hypothesize that compassion partially mediates the effects of abusive supervision on directed citizens hip behavior Partial, rather than full mediation is hypot hesized because other mediators of the abusive supervision citizenship
55 behavior relationship, such as organi zational justice have been identified (Aryee et al., 2007; Zellars et al., 2002) Hypothe sis 9 : c ompassion partially mediates the relationship between abusive supervision and directed citizenship behavior. Effects of Subordinate Behavior on Abusive Supervision Most research involving leader behavior, in cluding abusive supervision, investigates its influence on subordinates (Sims & Manz, 1984). Consistent with this research stream, in the previous section, I argued that abusive supervision influences ores the perspective of victim precipitation theorists that subordinates may play a role in knowingly or unknowingly contributing to their own mistreatment (Elias, 1986). Thus, in this section, I formulate an also influence supervisors. As noted previously, the idea that supervisor and subordinate behaviors can mutually influence one another is consistent with the concept of reciprocal determinism (Bandura, 1978, 1983), or the idea that person and environmen t determine one another. Supervisor D irected Deviance and Abusive Supervision For many of the same reasons that abusive supervision elicits supervisor directed deviance, supervisor directed deviance should also incite abusive supervision. Nam ely, therefore, can induce both anger and a subsequent desire to retaliate in supervisors. reasons. Firs t, organizational norms generally dictate that subordinates treat supervisors with courtesy and respect. Because engaging in supervisor directed
56 deviance is in strict violation of these norms, and because norm violations can be seen as unfair ( van den Bos Rombouts & Crone 2010) such behaviors are likely to be perceived as injustices. Second, and related, supervisor directed deviance may be viewed as insubordination. G iven that organizational policies typically require subordinates to submit to authority, supervisor directed deviance can be perceived as unfair rule violation, deserving of punishment. Third, evidence suggests that supervisors are often unaware of the ext ent to which their behavior aff ects subordinates (Kish Gephart et al 2009). This lack of awareness implies that even when supervisor directed deviance is enacted by subordinates for retaliatory reasons, supervisors are unlikely to recognize their role in warrant retaliation or punishment. Though some supervisors surely punish subordinates through the use of formal organizationa l procedures, others may resort to the more informal behaviors characteristic of abusive supervision (e.g., yelling). In addition to being appraised as unfair, supervisor directed deviance might also be perceived as ego threatening. More specifically, subo directed deviance can be viewed as a status challenge (Porath et al., 2008), and can signal to supervisors that they lack adequate control over subordinates. Because possessing a certain level of order and control in the workplace is t ypically an important job requirement for individuals occupying supervisory roles, such status challenges can imply that a supervisor is performing his or her job poorly and thus, can threaten his or her self esteem (Gardner, Van Dyne, & Pierce, 2004). In order to reestablish a sense control and preserve their self esteem, supervisors may engage in abusive behavior to
57 assert their power and also to deter others from acting defiantly (Bushman, Baumeister, Thomaes, Ryu, Begeer, & West, 2009). H ypothesis 10: S supervisor directed deviance positively predicts abusive supervision Directed Citizenship Behavior and Abusive Supervision Researchers have argued that interpersonally directed organizational citizenship behavior can prevent mistreatment by other s because it increases social attractiveness and forms bonds of mutual obligation and reciproci ty (Aquino & Bommer, 2003) These same explanations are likely to apply when discussing the relationship between supervisor directed citizenship behavior and abusive supervision. More specifically, because poor citizens tend to be less cooperative and often fail to exhibit appropriate levels of social sensitivity, they are viewed as less socially attractive than good citizens (Bolino, 1999). These findings, coupled with findings that negatively link social attractiveness to mistreatment (Furr & Funder, 1999; Parker & Asher, 1987) imply that supervisor directed citizenship behavior predicts abusive supervision. Another explanation for th e citizenship behavior mistreatment relationship proposes that, by engaging in supervisor directed citizenship behavior, subordinates can help establish a history of positive social interaction (Aquino & Bommer, 2003). This history cultivates a norm of p ositive reciprocity (Gouldner, 1960) whereby supervisors directing abusive supervision toward good citizens would violate this norm, supervisors should feel compelled to rest rain from mistreating those who treat them benevolently (Cialdini, 2001). Conversely, subordinates who abstain from engaging in supervisor directed citizenship are unlikely to found relationships with their supervisors that
58 possess positive reciprocity nor ms. In the absence of such norms, supervisors are less obligated to refrain from mistreating subordinates who may otherwise be viewed as supervisor directed citizenship behavior should be negatively related to abusive supervision. Hypothesis 11: s upervisor d irected citizenship negatively predicts abusive supervision. Directed Avoidance and Abusive Supervision Though one may well argue that av simply because the subordinate is less available to abuse, Tepper et al. (2007) proposed three convincing reasons as to why this effect is likely to be short lived. First, ole ambiguity, because it can cause one to lose access to information and resources necessary for adequate job performance. Second, and related, to the extent one invests effort in avoidance rather than in performing his or her work, avoidant behavior is p rone to interfere with job productivity. Decreased job performance, in turn, may signal to supervisors that the subordinate is an unmotivated employee further deserving of abuse. Finally, by showing that one is affected by abuse in such a way that he or sh e is unlikely to retaliate, he or she reinforces an image of vulnerability that can breed further victimization. Together, these reasons provide sound logic for why supervisor directed avoidance can, rather than prevent abuse, result in increased mistreatm ent. Hypothesis 12 : supervisor directed avoidance positively predicts abusive supervisio n.
59 Main and Moderating E ffects of Personality In the previous sections I have provided arguments not only for why abusive supervision influence s subordin behaviors These mutually reinforcing behaviors suggest multiple pathways through which vicious cycles of dysfunctional behavior can emerge, resulting in sustained patterns of perpetuate or put an end to such patterns. Thus, in an eff ort to partially address this gap, in the sections to follow, I draw from both cognitive motivational relational (Lazarus, 1991a, 1991b) and victim precipitation emotional a nd behavioral reactions to abusive supervision. More specifically, in line with the cognitive motivational relational theory of emotions (Lazarus, 1991a, 1991b), I argue that individuals with different personality traits can appraise abusive behaviors diff erently. These different appraisals should result in distinct emotional and subsequent behavioral reactions. Furthermore, consistent with the victim precipitation perspective (Elias, 1986), to the extent that these reactions result in increased supervisor directed deviance, supervisor directed citizenship behavior, supervisor directed avoidance, and/or other potentially maladaptive behaviors, they may bring about further abuse. This implies that, over time, individuals with certain personality traits can ex perience more overall mistreatment, in part, because of their reactions to previous incidences of it. Accordingly, I al so hypothesize that subordinate s emotions and supervisor directed
60 deviance, supervisor directed citizenship behavior, and supervisor dir ected avoidance As shown in Figure 3 1 I focus on three personality traits: volatility, withdrawal, and compassion. These traits are narr ow facets of the Big Five pe rsonality taxonomy (Costa & McCrae, 1992 ; Digman, 1990 ) derived by DeYoung, Quilty, and Peterson (2007). Based on work in the behavioral genetics field (Jang, Livesly, Angleitner, Riemann, & Verbon, 2002), DeYoung et al. (2007) proposed and found evidence to suggest that each of the Big Five traits could be further subdivided into two lower order factors with distinct biological substrates. I focus on these traits rather than the broader Big Five dimensions because they allow for more specific theorizing re garding the describing these relationships in detail. Trait Volatility I n his work on schoolyard bullying, Olweus (1978) identified two types of victims, submissive vict ims and provocative victims. Submissive victims tend to be anxious, insecure, cautious, sensitive, quiet, and low in self esteem, and thus appear vulnerable to abuse, whereas provocative victims tend to engage in behaviors that elicit retaliatory responses from others. Aquino and Byron (2002) and others (e.g., Tepper et al., 2006) have argued that submissive and provocative victims are also likely to exist in the workplace, and thus, have sought to recognize factors that could identify such victims. As a re affectivity predicted the extent to which they experienced abusive supervision. However, Tepper et al. (2006) offered both submissive and provocative victim explanations for thei r results. This dual explanation implies that using broad personality traits like
61 negative affectivity does not enable one to distinguish between provocative and submissive victims. One possible solution to this problem may be to examine more narrow person ality traits as predictors of abuse and of the mediating mechanisms that lead to it. Here, the facets of n euroticism proposed by DeYoung et al. (2007) volatility and withdrawal are likely to be especially useful. On the one hand, v olatility describes indiv iduals who possess problems of dis i nihibition, leading to the outward expression of negative affect Volatile individuals are prone to ability, irritability or anger, and difficulty controlling emotional impulses 2007, p. 885 ). Withdrawal, on the other hand, describes individuals who possess problems of inhibition, causing negative affect to be directed inward. Such individuals are prone to feeling vulnerable, afraid, worried, and self conscious. At first glance, it appears th at trait volatility may characterize more provocative victims, whereas trait withdrawal may typify more submissive victims. Thus, I test whether volatility can be used to identify more provocative victims by examining whether er and deviant r eactions to abusive supervision responses theoretically characteristic of provocative victims. Likewise, I examine whether withdrawal can be used to identify more submissive victims by examining voidance reactions to abusive supervision responses theoretically characteristic of submissive victims. I also examine whether volatility predicts abusive supervision through its effects on s upervisor directed deviance and whether w ithdrawal predicts
62 In discussing these relationships, I first focus on describing how trait volatility can that volatile individuals may be prone to anger is that they tend to appraise events in such a way motivational their cognitive appraisals and, thus, their emotional reactions. In support of this general appraisal styles (c.f., Gardner, Rozell, & Walumbwa, 200 4), and cognitive appraisal Aquino, Douglas, & Martinko, 2004 ). Though I am aware of no studies that have linked trait volatility to as examined how trait anger influences cognitive appraisals (c.f., Tafrate, Kassinove, & Dundin, 2002). In one study, Hazebroek, Howells, and Day (2001) asked a group of university students to rate two social interaction scenarios, both of which had negati ve consequences. The researchers also varied the intentions of the antagonists and the cognitive load of participants. They found that high trait anger individuals tended to assign more blame to antagonists, to evaluate the identified situations as being p ersonally relevant, and to respond more angrily to the situations than low trait anger individuals. Moreover, these appraisal biases were exacerbated for high trait anger individuals when there was some ambiguity as to the whether or not the provoking even t was deliberate. Because trait anger reflects a narrower component of trait volatility, it seems likely that volatile individuals will exhibit similar appraisal patterns, and thus, increased state anger reactions to events. In further extending this logic to the
63 abusive supervision anger relationship, it follows that volatility will amplify the effects more likely than less volatile individuals to appraise superviso anger eliciting manner. H ypothesis 13: Trait volatility moderates the effect of abusive supervision on anger such that the positive relationship between abusive supervision rdinates who possess high levels of trait volatility. If volatile individuals are more likely to react angrily to mistreatment, and if, as I have previously argued, anger is positively related to supervisor di rected deviance then, it follows that volatili ty should also amplify the effects of abusive supervision on superviso r directed deviance H ypothesis 14 : Trait volatility moderates the effect of abusive supervision on supervisor directed deviance such that the positive relationship between directed deviance is amplified for subordinates who possess high levels of trait volatility. Continuing with this line of thought, if trait volatility a effects on superviso r directed deviance and if these behaviors in turn, predict further abusive supervision then volatility should also positively predict abusive supervision. More specifically, to the extent that volatile incidences of abuse invite retaliation in the form of further mistreatment, a negative reciprocity norm can be established. The generation of this norm can facilitate an escalating pattern of mistreatment betwee n supervisor and subordinate (Andersson &
64 Pearson, 1999), and therefore, increase overall levels of abusive supervision. This line of reasoning su super visor directed deviance partially mediate the effects of trait volati lity on abusive supervision. Partial mediation is hypothesized because other factors are likely to influence the extent to which volatile subordinates perceive abusive supervision. For example, volatile individuals may provoke conflicts with coworkers if t hey attribute harmful motives to accidental behaviors. Increased conflict, in turn, may decrease productivity in the workgroup, resulting in frustrated supervisors who are prone to lashing out at subordinates. Similarly, volatile subordinates may be more l ikely to perceive Martinko et al., 2009 ) and thus, to report abusive supervision even when it does not objectively exist. Hypothesis 15a : Trait v olatility is positively related to abusive supervision. Hypothesi s 15b : Anger and supervisor directed deviance partially mediate the Trait Withdrawal appraisals of abuse and therefore, their emotional reactions. Because no studies have linked trait withdrawal to cognitive appraisals, however, it is useful to draw from the literature involving similar narrow traits. One such trait is trait anxiety. In a review of the lit erature involving personality traits and cognitive appraisals, Winter and Kuiper (1997) concluded that trait anxiety was positively related to the tendency to appraise environments as stressful and threatening. Additionally, i n a later study, Roesch and Ro wley (2005) found trait anxiety to be negatively relat ed to challenge appraisals (i.e., believing that one can positive ly attack stress) and positively related to threat
65 appraisals, appraisals that one lack s resource s necessary to cope with stress, and app raisals that stres s is centra negatively impact it. Extending these findings to trait withdrawal implies that individuals possessing high levels of withdrawal will appraise abusive supervision as more threatening and as less managea ble, and, thus, as something to be feared. H ypothesis 16 : Trait withdrawal moderates the effect of abusive supervision on es who possess high levels of trait withdrawal. I have previously argued that fear is positively related to supervisor directed avoidance. Therefore, if high trait withdrawal amplifies the abusive supervision fear relationship, then, it seems logical to assume that trait withdrawal also amplifies the effects of abusive supervision on supervisor directed avoidance. Thus, H ypothesis 17 : Trait w ithdrawal modera tes the effect of abusive supervision on supervisor directed avoidance such that the positive relationship between directed avoidance is amplified for subordinates who possess high levels of trait withdrawal. Withdrawal should be linked to increased levels of abusive supervision for several reasons. One such reason is that the tendency for individuals who possess high levels of trait withdrawal to appraise situations as more stressful might cause them to seek unusually high levels of social support from others. By reaching out to others for help in coping with seemingly trivial stressors, high withdrawal individuals can come across as
66 degree, as deserving of abuse. Additionally, avoiding potentially challengi ng work assignments because they are appraised as threatening or difficult and dedicating job performance. Poor performance, in turn, is likely to reinforce the notion t hat the subordinate is deserving of abuse. A final explanation for the trait withdrawal abusive supervision relationship involves partial mediation through supervisor directed fear and avoidance. More specifically, if individuals who possess high levels of trait withdrawal tend to be more fearful and avoidant in response to mistreatment, they are also prone to come across as interaction patterns suggest that high trait wit hdrawal individuals can find themselves trapped in a vicious cycle of dysfunctional behavior, whereby their own behavior results in initial victimization and reinforces its use. Hypothesis 18a : Trait withdrawal is positively related to abusive supervision. Hypothesis 1 8b : Fear and supervisor directed avoidance partially mediate the effects of trait withdrawal on abusive supervision. Trait Compassion DeYoung et al. (2007) defined trait compassion a facet of agreeableness as a compassionate emotional affiliat ion with others. Individuals possessing high levels of trait compassion tend to exhibit warmth, sympathy, and tenderness, and to affiliate with others emotionally. Rather than helping to distinguish between provocative and submissive victims, I argue that trait compassion can instead protect individuals from abusive supervision. This shielding effect likely occurs because trait compassion can
67 cause individuals to appraise or reappraise abusive supervision in such a way that it is responded to with state com passion. Unfortunately, virtually no studies have examined the appraisal styles of individuals possessing trait compassion, trait sympathy, or other related narrow facets of agreeableness. However, research has linked agreeableness to positive reappraisal (Vickers, Kolar, & Hervig, 1989; Watson & Hubbard, 1996) and to forgiveness (McCullough, 2001) behavior toward and thoughts and feelings about an offender ( Aquino Grover, Goldman, & Folger, 2003). A dditionally, given that, by definition, compassionate individuals tend to have warm feelings toward others, it follows that compassionate subordinates are also prone to liking others. Liking, in turn, has been associated with a tendency to both forgive and avoid blaming transgressors (Bradfield & Aquino, 1999) Together, these findings suggest that subordinates who possess high levels of trait are more forgiving of transgres sors. As a result, subordinates possessing high levels of trait compassion are more likely to feel state compassion in response to mistreatment compared to their less compassionate counterparts. Hypothesis 1 9 : Trait compassion moderates the effect of abusi ve supervision on possess high levels of trait compassion. Provided trait compassion mo derates the relationship between abusive supervision and state compassion, it should also moderate the relationships between
68 abusive supervision and supervisor directed citizenship behavior More specifically, if, as previously argued, state compassion is positively related to supervisor directed citizenship behavior, and if trait compassion reverses the negative abusive supervision state compassion relationship, then, logically, a busive supervision should also reverse the abusive supervision supervisor directed citizenship behavi or relationship so that it become s positive H ypothesis 20 : Trait compassion moderates the effect of abusive supervision on supervisor directed citizenship behavior, such that the negative relationship between abus directed citizenship behavior becomes positive for subordinates who possess high levels of trait compassion As with trait volatility and trait withdrawal, the effects of trait compassion should be partially med possessing high levels of trait compassion appraise abuse in such a way that they can pro blems, then they are likely to feel increased compassion for their super visors. Consequently, they are more likely to help or offer support. In this way, compassionate spiraling effect of dysfunctional behavior that can elicit a more sustained pattern of abusive supervision. Also as with volatility and withdrawal, I hypothesize partial, rather than full mediation because other potential mechanisms explain why trait compa ssion negatively predicts abusive supervision. For instance, compassionate individuals may lend a
69 to facilitate a climate of cohesiveness that increases job performance. To the extent the superv isor recognizes establishment of a positive reciprocity norm that obligates the supervisor to treat the subordinate favorably. Hypothesis 21a : Trait compassion negati vely predicts abusive supervision. Hypothesis 21b : State compassio n and supervisor directed citizenship behavior partially mediate the effects of trait compassion on abusive supervision.
70 Figure 3 1. Conceptual model Deviance Fear Citizenship Avoidance Anger Abusive Supervision Compassion Moderators Volatili ty Withdrawal Compassion
71 CHAPTER 4 METHODS Sample and Procedure The initial sample for this study consisted of 159 full time employees who represented a variety of industries including education, medicine, social work, finance, accou n ting, insu rance, information technology, travel engineering, law, transporta tion, and communications To recruit participants, I placed announcements on several popular classified advertising websites. The announcements instructed individuals who were interested in participating in a university study about in the to email me so that I could send them a link to a brief online registration form. In order to qualify for the study, participants were required to be U.S. citizens, to work full time and t o have an immediate sup ervisor. Additionally, during the registration process, each pa rticipant was asked to provide the name and email address of one coworker who could be contact ed to complete a single survey at the start of the study. To capture the dynamic, multilevel nature of the relationships proposed in this study participants were asked to complete a series of seven surveys over the course of a five month period (six waves). All surveys were conducted online and were distributed through email. The first survey contained measures of potential time invariant control variables (e.g., demographics) as well as a measure of perceived abusive supervision over the previous six months 1 and was available to complete for a one week period On 1 I compared both self and coworker (2000) measur e, participants were asked to rate how often their immediate supervisors engaged in each The same items were instructions were modified to refer to the focal participant. Self and coworker reports were highly correlated (r = .58, p < .01), providing some evidence that abusive supervision is more than merely a perceptual phenomenon.
72 the Tuesday morning following the init ial survey, participants began receiving a monthly survey containing measures of abusive supervisio n, emotions, and behaviors. Monthly surveys were available to participants from the first Tuesday morning through the first Thursday evening of each month, b eginning in September 2010 and ending in February complete a single survey available initial surveys. The coworker s urvey assessed foca and experienced levels of abusive supervision during the six month period prior to the initial survey. A $90 honorarium was awarded to those individuals who entirely complete d the study. O f the 159 indi viduals who registered for the study, 148 completed the initial survey, and 143 completed at least one monthly survey. Twelve coworkers did not complete surveys, further reducing the number of participants who provided usable data to 128 A small number participants (n = 6 ) either changed jobs or were assigned a different supervisor during the course of the study. In these instances, only data obtained prior to when the supervisor or job change occurred were included in the final set o f analyses. In sum, after listwise deletion, usable da ta were available for 119 focal participants, and the number of data point s yielded from th e monthly surveys range d from 484 to 489 for lagged an alyses and from 621 to 625 for non lagged analyses. Parti cipants included in the final set of analyses worked at their current organization for an a verage of 5.63 years ( SD = 5.71) and with their current supervisor for an average of 3.09 years ( SD = 2.83) The sample was predominantly female and
73 Caucasian (78.3% ) and the average age of participants was 37.92 ( SD = 11.47) Most participants (85.8%) had obtained at l east a b and 60% of the sample was married or residing with a romantic partner Measures Personality Coworkers assessed focal partici rait volatility, trait with draw al, and trait compassion using 30 items from DeYou ng et al. (2007). Sample items designed to measure trait volatility include d see my coworker as someone who get s and see my coworker as someone w ho get s ups Trait withdrawal sample items include d see my coworker as someone who feel s Finally, trait compassion sample items include d see my coworker as s omeone who sympathize s (reverse coded). Respondents use d a rating scale ranging from 1 = Very Inaccurate to 5 = Very Accurate alpha for thes e measures were: Volatility, = .88 ; Withdrawal, = .86 ; Compassion, = .87. Abusive Supervision ) 15 item scale. Respondents were asked to rate the frequency with which their immediate supervisor, e down in front Items use a response scale ranging from 1 = Never to 5 = Very often over the time intervals, was = 94.
74 Emotions Because emotions tend to be target specific ( c.f., Weiss & Cropanzano, 1996) participants w ere with or thinking about their immediate supervisor Specific ally, particip ants respond ed to the following stem: different feelings and emotions. Ple ase indicate to what extent you felt this way during the past month, when thinking about or interacting with your immediat e supervisor. Please click the option that accurately reflects your response to each question using the response scale provided. During the past month, when I thought about or interacted Responses were given on a 1 = Very Slightly to Not at A ll to 5 = Very much scale. Anger was assessed with a combination of three items from Rodell and Judge (2009) and Crossley (2009). Rodell and Judge (2009) measured anger using two items from the Positi ve and Negative Affect Schedule Expanded Form (PANAS X; Watson & Clark, 1994) (2009) added the item, X (Watson & Clark, 1994) were used to assess r to measure compassion, I use d six items identified by Goetz et al. (2010) as representing the subjective experience of compassion. These items were drawn from a number of n, Fultz, Vanderplas, each of these measures pooled over the time intervals w as : Anger = .92 ; Fear = .92 ; Compassion, = .96.
75 Behaviors assessed using a frequency scale ranging from 1 = Never to 5 = Very Often Items were The questions on this page refer to behavi ors both positive and negative that you may or may not have en gaged in during the past month. For each of the items, please indicate how often you engaged in the behavior during the past month using the response scale below. Please be open and honest in yo ur responding and remember that your responses are completely confidential. During the past month, I: Supervisor directed deviance Super visor directed deviance was measured using seven items from Mitchell and Ambrose (2007). Items include d Made fun of my supervisor Publicly embar for thi s scale, pooled over the time intervals, was = .84. Supervisor directed avoidance Supe rvisor directed avoidance was measured using seven items. Four items were ada pted from Wade (1989) to reference the supervisor as the target of avoidance. These (adapted) items include d A voided my L K ept as much distance as possible between my su pervisor and me adapted from Moss, Valenzi, and Taggart (2003). Their measure was originally designed to measure actions taken to can be easily modified to mea sure more general supervisor directed avoidance behaviors Modified items used in this study include d W ent the other way when I saw P retended to be sick and stay ed home in order to avoid my
76 T T ried to avoid eye contact with my supervisor so that (s)he didn conversation with me Cronbach was = .88. Supervisor directed citizenship Supervi sor directed citizenship was measured using five items f rom Malatesta (1995). Items include d Accepted added responsibil ity w with To ok a personal interest in Passed along work related information to my = 89. Control Vari ables I contro l l ed for a number of variables that can potentially influence the relationships examined within this study. First, I control led for subordinate supervisor tenure, given its potential influence on perceptions of social interaction at work ( Wayne, Shore, Bommer, & Tetrick, 2002 ). Second I contro l l ed ob mobility because it provides workers with a sense of personal control that can influence coping resources and strain reactions (Frazier, Mortensen, & Steward, 2005; Perrewe & Ganster, 1989) and because it can potentially affect the manner in which supervisors act toward subordinates To do so, I use d two it ems from Tepper et al. (2007): I would ha find anot Third I contro l l ed for demographic variables (gender and race ) and job status because the s e variables may influence individu levels of social attractiveness (Aquino & Bommer, 2003), and therefore, their social interaction patterns. Fourth I controlled for extraversion, trait volatility, trait withdrawal,
77 and trait compassion, given prior research findings that show persona lity to be linked to emotion, behavior, and social interaction ( Judge, Klinger, Simon, & Yang, 2008) Finally, because I lag the data to test for reciprocal relationships, for all lagged analyses relating I c ontrol led for the prior levels of the dependent variable
78 CHAPTER 5 RESULTS Analyses Overview Because monthly measures of a busive supervision, and behaviors were nested within individuals, I estimated hierarc hical linear models (Bryk & Raudenbush, 1992; Snijders & Bosker, 1999) using HLM 6 (Raudenbush, Bryk, & Congdon, 2005) to test my hypotheses Prior to testing the hypothesized models, however, I estimated null models consisting only o f intercepts and no pr edictors in order to examine the extent to which variance in each of the Level s ) was partitioned within and between individuals. If the Level 1 variables do not e xhibit su bstantial within individual fluctuation, then hierarchical linear modeling is unnecessary because there ceases to be within individual variation to explain. Between and within individual variance ons and behaviors are displayed in Table 5 1. As shown in the table 24.27 % of the variance in abusive supe rvision is within individual, 34.77 % of the variance in anger is wit hin in dividual, 39.29 % of the variance in fear is within individual, 29.04 % of th e variance in com passion is within individual, 28.88 % of the variance in supervisor directed d eviance is within individual, 22.30 % of the variance in supervisor directed avoida nce is within individual, and 27.64 % of the variance in supervisor directed cit i zenship is within individual. Moreover, significant between individual variance existed fo r each L evel 1 variable. Taken together the null model results sugge st that there is sufficient within individual variance to explain and thus, that hierarchical li near modeling is appropriate for testing the study hypotheses.
79 Table 5 2 displays means, standard deviations, and correlations among the study variables. W ithin individual correlations calculated by standardizing regression coefficients derived from HLM mo dels with a single predictor are above the diagonal. Between individual correlations are below the diagonal and were calculated using aggregated Level 1 scores (and therefore do not properly reflect multilevel relationships) Tests of Hypotheses Main Effec ts of Abusive Supervision on Emotions To test hypotheses linking abusive supervision to emotions, I specified a series of HLM regression e quations in which each emotion was regressed on abusive supervision at Level 1. Data were lagged such that Level 1 ind ependent variables were measured one month prior to each dependent variable Moreover, i n all analyses, extraversion, trait volatility, trait withdrawal, trait compassion, gender, race, supervisor tenure, job mobility, and job status were ent ered as Level 2 control variables and supervisor subordinate interaction time was entered as a Level 1 control Level 1 variables were also uncentered in all analyses. As shown in Table 5 3 abusive supervision positively predicted anger ( = .48, p < .01 ) and fear ( = .29, p < .01 ) and abusive supervision negatively predicted compassion ( = .30, p < .01 ). Thus, Hypotheses 1, 4, and 7 were supported. Emotion M ediate d Pathways To test the mediational hypotheses, I used Baron and Kenny (1986) causal steps approach According to Baron and Kenny (1986) mediation occurs when : (1) an initial independent variable predicts the outcome variable; (2) the independent variable predi cts the mediatin g variable; (3) the mediating variable predicts the outcome variable
80 when the effect of the independent variable on the outcome variable is also modeled; and (4) the effect of the independent variable on the outcome variable becomes non significant when th e mediating variable is included as a predictor of the outcome variable The previously discussed set of analyses linking abusive supervision to offer s support for the second causal step. To examine whether results supported Step 1 each behavior was regressed on abusive supervision and the control variables. Additionally, prior month s levels of the dependent variable were entered at Level 1 so that results would capture change in the dependent variable from one time period to the ne xt. As with the former set of analyses a one month lag between the indepe ndent and dependent variables was modeled R esults are displayed in Table 5 4 Although abusive supervision positively predicted change in supervisor directed deviance ( = .12, p < .05 ) and supervisor directed avoidance ( = .22, p < .01 ) the effect of abusive supervision on supervisor directed citizenship behavior ( = .16, n.s. ) was non significant. Thus criteria for Step 1 of the mediational process were met for supervisor directed deviance and supervisor directed avoidance outcomes, but not for supervisor directed citizenship behavior. R esults for Step 3 are displayed in Table 5 5. For these analyses a series of eq uations was estimated in which each potential emotion mediator, measured concurrently with each corresponding dependent variable, was added to the equations estimated f or Step 1. To allow the model to be estimated, Level 1 slopes between each mediator and outcome variable were fixed (i.e., they did not vary between individuals). All other Level 1 slopes were allowed to vary between individuals. Results supported
81 Hypotheses 2, 5, and 8 in that anger positively predicted supervisor directed deviance ( = .09 p < .01 ) fear positively predicted supervisor directed avoidance ( = .23, p < .01 ) and compassion positively predicted supervisor directed citizenship behavior ( = .20, p < .01 ) Thus, cri each dependent variable. To examine whether the data met criteria for Step 4 of the mediational process which requires that the effect of the independent variable on the dependent variable b ecome non significant when the media tors are entered into the model I compared the coefficients for the effect of abusive supervision on supervisor directed deviance and supervisor directed avoidance in Table 5 4 with those displayed in Table 5 5 As shown in Table 5 5 the effect of abusive supervision on sup ervisor directed deviance becomes non significant when anger is entered as a potent ial mediator ( = .06, n.s. ) Thus, Hypothesis 3 appears to be supported in that the criteria for each causal step are met. Comparing Tables 5 4 and 5 5 also reveals that the effect of abusive supervision on supervisor directed avoidance is reduced, but remains significant ( = .16, p < .05 ) when fear is entered as a potent ial me diator In support of Hypothesis 6 this result, together with results for Steps 1 through 3 suggests partial mediation of the abusive supervision supervisor directed avoidance relationship via fear. Partial mediation occurs when the effects of the inde pendent variable are reduced, but remain significant, when a significant mediating variable is entered into the model According to the Baron and Kenny (1986) logic, because criteria for Step 1 were not fulfilled for the abusive supervision citizenship rel ationship, there is no effect to be mediated. However, it remains possible that an indirect effect of abusive supervision on
82 citizenship behavior exists even in the absence of a direct effect. Such a situation might occur when an overall direct effect is c oncealed by conflicting indirect effects ( MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). To ascertain the significance of an indirect effect, I conducted a So bel (1982) test. Results of the Sobel test reveal a sig nificant indirect effect of abusive s upervision on citizenship behavior through compassion (Z = 3.32, p < .0 1) Thus, Hyp othesis 9 which argues for at least partial mediation of the effect of abusive supervision on supervisor directed citizenship behavior was supported. Sobel tests for the indirect effects of abusive supervision on supervisor directed deviance through anger (Z = 2.88, p < .01) and abusive supervision on supervisor directed avoidance through fear (Z = 2.35, p < .05) were also significant, further supporting Hypotheses 3 and 6 Effects of Subordinate Behavior on Abusive Supervision simu ltaneously entered supervisor directed deviance, supervisor directed avoidance, and supervisor directed citizenship b ehavior as Level 1 predictor s of abusive supervision. Like the analyses examining abusiv e supervision, I lag ged the data so that the independent variables were measured one month prior to the dependent variable Addit ionally, along with controls, I also entered abusive supervision measured c oncurrently with the independent variables as a Level 1 predictor To allow the model to be estimated, Level 1 slopes between supervisor directed deviance, supervisor directed avoid ance, supervisor directed citizenship behavior and abusive supervision were fixed (i.e., they did not vary between individuals). The supervisor interaction abusive supervision slope was allowed to vary between individuals. Results for these ana lyses are displayed in Table 5 6 As shown in
83 the table, supervisor directed deviance ( = .06, n.s. ) and supervisor directed avoidance ( = .15, p < .01) positively predicted abusive supervision, whereas supervisor directed citizenship ( = .01, n.s. ) negatively predicted abusive supervision. However, only the effect of supervisor directed avoidance on abusive supervision was significant. Thus, Hypothesis 12 was supported, but results failed to support Hypotheses 10 and 11 Moderating Effects of Personality on Abusive Supervision and Emotions To test hypotheses involving the moderating role of personality traits on the abusive supervision emotion relationships I entered relevant personality traits, ce ntered at their respective sample means, as Level 2 predictors of within individual slopes of these relationships. There was a one month lag between the independent and dependent variables. Results of these analyses are displayed in Table 5 7 Contr ary to Hypotheses 13 and 16 trait volatility did not moderate the abusive supervision anger relationship ( = .12 n.s.) and trait withdrawal did not moderate the abusive supervision fear relationship ( = 11 n.s.) Trait compassion did moderate the relationship between abusive supervision and state compassion ( = .23, p < .05) However, Figure 5 1 reveals the form of this interaction was n ot as hypothesized Instead, the relationship between abusive supervision and state compassion tended to be more strongly negative (m = .54, p < .01 ) for individuals who possessed higher levels of trait compassion than for individuals who possessed lower levels of trait compassion (m = .23, p < .05 ) Thus, Hypothesis 19 was not supported.
84 Moderating E ffects of Personality on Abusive Supervision and Behaviors Table 5 8 contains the results of analyses conducted to examine whether personality moderated the relationships between monthly levels of abusive superv ision To conduct these analyses, I added relevant personality predictors of the Level 1 abusive supervision subordinate behavior slopes to the equations modeling the main effects of abusive supervision on subordi na ors. Contrary to Hypothesis 14 trait volatility did not emerge as a significant moderator of the abusive supervision supervisor directed deviance relationship ( = .06 n.s.) Likewise, trait compassion did not moderate the relation ship between abusive supervision and supervisor dire cted citizenship behavior ( = .01 n.s.) Thus, Hypothesis 20 was not supported. A final hypothesis inv olving moderation Hypothesis 17 suggested that trait withdrawal would amplify the positive relationship between abusi ve supervision and supervisor directed avoidance. Table 5 8 shows the abusive supervision trait withdrawal interaction term to be significan t ( =.13 p < .01). A plot of the interaction (see Fi gure 5 2 ) reveals the pattern of relation ships to be consistent with those proposed in Hypothesis 17 For individuals who possessed high levels of trait withdrawal, the abusive supervision supervisor directed avoidance simple slope was more strongly posi tive (m = .28, p < .01) than for individuals who possessed low levels of trait withdrawal (m = .08, n.s.) Main Effects of Personality on Abusive Supervision To test whether personality predicted overall levels of abusive supervision, of trait volatility, trait withdrawal, and trait compassion were simultaneously entered as predictors of abusive supervision. Results of these analyses
85 are displayed in Table 5 9 In support of Hypothesis 18a trait withdrawal positively predicted abusive supervision ( = .21, p < .05) Neither trait volatility ( = .09, n.s.) nor trait compassion ( = .11, n.s.) however, significantly predicted abusive supervision. Therefore, Hypotheses 15a a nd 21 a were not supported. Mediators of the Relationship between Personality and Abusive Supervision between their personality traits and overall levels of abusive supervision, I again used ausal steps approach. Results for Step 1 are described in the previous section. Tables 5 10 and 5 11 contain results for analyses involving Step 2. As shown in the tables, lly unrelated to their emotions and behaviors. More specifically, trait volatility was unrelated to both anger ( = .13, n.s.) and supervisor directed deviance ( = .01, n.s.) Likewise, trait compassion was neith er related to state compassion ( = .08 n.s.) n or supervisor directed citizenship behavior ( = .21 n.s.) Finally, although trait withdrawal was positively and significantly related to fear ( = .20 p < .05 .) it was unrelated to supervisor directed avoidan ce ( = .19 n.s.) Results involving Step 3 of the mediational an al yses are displayed in Table 5 12 In examining potential mediators of the trait volatility abusiv e supervision relationship, analyses revealed that both anger ( = .35, p < .01) and deviance ( = .28, p < .05) positively predicted abusive supervision Among the potential mediators of the trait withdrawal abu sive supervisi on relationship, a voidance ( = .60, p < .01 ) but not fear ( = .15, n.s. ) significantly predicted abusive supervision. None of the proposed
86 mediators of the trait compassion abusive s upervision rel ationship were significant (compassion: = .09, n.s. ; supervisor directed citizenship: = .02, n.s. ) To assess whether the data met the criteria for Step 4 I compared coefficients for the effects of relevant personality traits on abusive s up ervision displayed in Table 5 9 with those displayed in Table 5 12 In all cases, effect sizes were reduced when potential mediators were included as predictors. However, because results of Step 1 only revealed a significan t overall effect of withdrawal on abusive supervision, and because results tended to be mixed for Steps 2 and 3, all four causal steps were not met for any of the proposed mediated relationships. Sobel tests confirmed these findings in that none of the tes ts revealed the existence of significant indirect effects. More specifically, Sobel statistics for each proposed mediator were as follows: Anger, Z = .96; Supervisor directed deviance, Z = .15; Fear, Z = 1.45; Avoidance, Z = 1.39; Compassion, Z = .64; Supe rvisor directed citizenship behavior Z = .20. Thus, Hypotheses 15b, 1 8b, and 21b were not supported.
87 Table 5 1. Variance decomposition of Level 1 variables HLM variance estimate Variance decomposition Between Within Between Within Abusive s upervis ion 0. 25534 0. 50531 75.73 % 24.27 % Anger 0.59036 0.31474 65.23% 34.77% Fear 0.20719 0.13407 60.71% 39.29% Compassion 0.97315 0.39834 70.96% 29.04% Deviance 0.19426 0.07890 71.12% 28.88% Avoidance 0.34205 0.09818 77.70% 22.30% Citizenship 0.77400 0.29563 72.36% 27.64% Note: Variance decompositions are computed by dividing the between or within individual variance components by the sum of the two (between and within) variance estimates.
88 Table 5 2. Means, standard deviations and intercorrela tions among study variables M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1. Abusive supervision (L1) 1.32 0.58 1.00 47 ** .3 7 ** .1 8 ** .3 4 ** .5 9 ** .01 .02 2. Anger (L1 ) 1.54 0.94 .78 ** 1.00 .4 4 ** 10 21 ** 34 ** .03 .0 5 3. Fear (L1 ) 1.25 0.56 .68 ** .65 ** 1.00 .0 7 20 ** .3 8 ** .01 .03 4. Compassion (L1 ) 2.22 1.19 .21 .20 .07 1.00 .02 .1 6 ** .24 ** .1 6 ** 5. Deviance (L1 ) 1.28 0.53 .67 ** .62 ** .58 ** .09 1.00 53 .06 .01 6. Avoidance (L1) 1.37 0.65 .81 ** .75 ** .67 ** .29 ** .66 ** 1.00 .0 8 .10 7. Citizenship (L1) 3.38 1.04 .08 .05 .09 .55 ** .13 .27 ** 1.00 .1 5 ** 8. Supervisor interaction (L1 ) 30.94 24.78 .01 .02 .03 .35 ** .17 .12 .41 ** 1.00 9. Volatility (L2) 2.16 0.81 .39 ** .32 ** .27 ** .14 .28 ** .28 ** .00 .08 1.00 10. Withdrawal (L2) 2.18 0.78 .40 ** .33 ** .32 ** .11 .33 ** .31 ** .06 .11 .69 ** 1.00 11. Compassion (L2) 4.23 0.68 .22 .17 .23 .16 .27 ** .26 ** .15 .15 .39 ** .31 ** 1.00 12. Extraversion (L2) 3.68 0.96 .04 .06 .05 .23 .02 .01 .19 .00 .13 .31 ** .33 ** 1.00 13. Supervisor tenure (L2 ) 3.09 2.83 .02 .03 .01 .12 .12 .03 .14 .14 .04 .01 .07 .08 1.00 14. Gender (L2 ) 0.25 0.43 .05 .04 .05 .08 .03 .01 .04 .10 .01 .07 .09 .05 .04 1.00 15. Race (L2 ) 0.79 0 .41 .11 .15 .00 .02 .15 .10 .03 .15 .00 .01 .05 .04 .05 .11 1.00 16. Job m obility (L2) 3.11 1.17 .01 .07 .06 .09 .06 .01 .14 .08 .08 .11 .09 .10 .06 .03 .14 1.00 17. Job status (L2 ) 0.08 0.28 .17 .18 .24 ** .11 .20 .08 .02 .02 .14 .16 .03 .0 2 .21 .18 .01 .04 1.00 Note s : N = 119. L1 = Level 1 (within individual) variables; L2 = Level 2 (between individual) variables. Between individual correlations are below the diagonal. Within individual corre lations are above the diagonal. p < .05, tw o tailed. ** p < .01, two tailed.
89 Table 5 3 Anger Fear Compassion Parameter B SE T value B SE T value B SE T value Intercept, B 00 0.844 0.161 5.24 ** 0.8 46 0.074 11 37 ** 2.461 0.156 15 81 ** Gender, B 01 0.078 0.133 0.59 0.145 0.076 1.91 0.132 0.174 0 76 Race B 02 0.395 0.205 1 .9 3 0.040 0.064 0.62 0.118 0.227 0 52 Job status, B 03 0.264 0.213 1 24 0.050 0.102 0 49 0.007 0.361 0 .0 2 Supervis or tenure B 04 0.003 0.011 0 24 0.005 0.007 0. 63 0.046 0.029 1 59 Job mobility, B 0 5 0.007 0.038 0.18 0.005 0.018 0.27 0.043 0.081 0.54 Extraversion, B 0 6 0.042 0.067 0.63 0.003 0.031 0.10 0.229 0.096 2.39 Volatility, B 0 7 0.055 0.104 0.53 0 .032 0.049 0.66 0.195 0.135 1.44 Withdrawal, B 0 8 0.093 0.106 0.88 0.101 0.045 2.23 0.138 0.137 1.01 Compassion, B 0 9 0.051 0.109 0.47 0.001 0.049 0.01 0.018 0.127 0.14 Supervisor interaction (t 1), B 10 0.001 0.002 0.92 0.000 0.001 0.26 0 .003 0.002 1.34 Abusive supervision (t 1), B 20 0.477 0.110 4.33 ** 0.292 0.068 4.30 ** 0.298 0.069 4.29 ** Pseudo R 2 0.34 0.69 .09 Notes: B = unstandardized regression coefficient. SE = standard error. T 1 indicates variable wa s measured one month prior to the dependent variable. p < .05, two tailed. ** p < .01, two tailed.
90 Table 5 4 Supervisor directed d eviance Supervisor directed a voidance Supervisor directed c itizenship Parameter B SE T value B SE T value B SE T value Intercept, B 00 0.413 0.047 8.74 ** 0.243 0.057 4 29 ** 2.429 0.248 9 79 ** Gender B 01 0.014 0.026 0.54 0.008 0.032 0.24 0.065 0.107 0 61 Race B 02 0.057 0.047 1 22 0.109 0.051 2.12 0.063 0.141 0 45 Job s tatus, B 03 0.152 0.081 1 89 0.051 0.045 1 13 0.029 0.226 0 13 Supervisor t enure B 04 0.001 0.004 0 26 0.000 0.005 0. 04 0.030 0.021 1 43 Job m obility, B 0 5 0.015 0.011 1.39 0.024 0.012 1. 94 0.128 0.050 2.55 Extraversion, B 0 6 0.017 0.019 0.89 0.001 0.017 0.08 0.126 0.063 2.00 Volatility, B 0 7 0.013 0.029 0.44 0.004 0.028 0.13 0.054 0.089 0.60 Withdrawal, B 0 8 0.023 0.034 0.70 0.041 0.035 1.18 0.006 0.103 0.06 Compassion, B 0 9 0.030 0.025 1.19 0.040 0.029 1.37 0.124 0.080 1.56 Supervisor i nteraction (t 1) B 10 0.002 0.001 1.78 0.000 0.001 0.48 0.006 0.002 2.93 ** Dependent variable (t 1) B 20 0.512 0.064 8.03 ** 0.597 0.066 9.03 ** 0.283 0.054 5.28 ** Abusive s uperv ision (t 1) B 3 0 0.115 0.057 2.03 0.224 0.068 3.33 ** 0.155 0.101 1.53 Pseudo R 2 0.58 0.71 0.31 Notes: B = unstandardized regression coefficient. SE = standard error. T 1 indicates variable was measured one month p rior to the dependent variable. p < .05, two tailed. ** p < .01, two tailed.
91 Table 5 5 HLM regression results depicting emotions as mediators of the relationships between abusive supervision and Supervisor directed d eviance Supervis or directed a voidance Supervisor directed c itizenship Parameter B SE T value B SE T value B SE T value Intercept, B 00 0.395 0.044 8 .93 ** 0.094 0.052 1 81 1. 772 0.2 42 7 33 ** Gender B 01 0.024 0.025 0.97 0.021 0.031 0.66 0.097 0.09 1 1 06 Race B 02 0.034 0.04 2 0 82 0.106 0.052 2.04 0.0 5 3 0.13 3 0. 40 Job s tatus, B 03 0.144 0.08 1 1 79 0.017 0.045 0 39 0.0 50 0.173 0.29 Supervisor t enure B 04 0.00 0 0.004 0 1 1 0.000 0.005 0. 05 0.012 0.019 0.63 Job m obility, B 0 5 0.013 0.011 1.2 1 0.028 0.012 2.22 0.11 6 0.04 2 2. 79 Extraversion, B 0 6 0.02 4 0.018 1.4 0 0.003 0.015 0.19 0.065 0.0 49 1 32 Volatility, B 0 7 0.01 4 0.028 0.5 2 0.006 0.028 0.23 0.09 2 0.07 5 1 2 3 Withdrawal, B 0 8 0.02 0 0.03 3 0.6 0 0.014 0.035 0.41 0.058 0.09 0 0. 6 5 Compas sion, B 0 9 0.03 5 0.024 1.4 6 0.052 0.027 1.96 0. 0 96 0.06 7 1. 44 Supervisor i nteraction (t 1) B 10 0.00 1 0.001 1.6 3 0.000 0.001 0.31 0.005 0.002 2. 54 Dependent variable (t 1) B 20 0.490 0.064 7.70 ** 0.572 0.069 8.23 ** 0.316 0.05 5 5. 74 ** Propo sed emotion mediator, B 30 0.087 0.022 3.95 ** 0.226 0.080 2.81 ** 0.199 0.038 5.25 ** Abusive supervision (t 1), B 4 0 0.057 0.055 1.03 0.155 0.078 2.00 0.043 0.10 0 0.4 3 Pseudo R 2 0. 59 0.72 0.40 Notes: B = unstandardized regress ion coefficient. SE = standard error. T 1 indicates variable was measured one month prior to the dependent variable. Anger is the proposed emotion me diator of the abusive supervision supervisor directed deviance relationship. Fear is the proposed emoti on mediator of the abusive supervision supervisor directed avoidance relationship. Compassion is the proposed emotion med iator of the abusive supervision supervisor directed citizenship relationship. p < .05, two tailed. ** p < .01, two tailed.
92 Ta ble 5 6 supervision Abusive s upervision Parameter B SE T value Intercept, B 00 0 392 0. 087 4.52 ** Gender B 01 0.0 29 0.038 0 76 Race B 02 0. 0 63 0. 040 1 59 Job s tatus B 03 0.01 4 0. 061 0. 23 Supervisor t enure B 04 0. 002 0. 004 0. 52 Job m obility B 05 0. 016 0.0 12 1.36 Extraversion B 06 0.0 15 0.0 12 1 24 Volatility B 07 0. 026 0. 023 1.51 Withdrawal B 08 0.0 45 0. 027 1 68 Compassion, B 09 0.016 0.021 0.74 Supervisor i nteract ion (t 1) B 10 0.001 0.001 1.27 Abusive s up ervision (t 1) B 20 0. 490 0. 052 9.41 ** Deviance (t 1) B 30 0.059 0.046 1.27 Avoidance (t 1) B 40 0.150 0.051 2.94 ** Citizenship (t 1) B 50 0.014 0.017 0.81 Pseudo R 2 0. 43 Notes: B = unstan dardized regression coefficient. SE = standard error. T 1 indicates variable was measured one month prior to the dependent variable. p < .05, two tailed. ** p < .01, two tailed.
93 Table 5 7 HLM regression results depicting personality traits as modera tors of the relationship between abusive supervision Anger Fear C ompassion Parameter B SE T value B SE T value B SE T value Intercept, B 00 0.873 0.159 5.48 ** 0.885 0.074 12 03 2. 557 0. 162 15 .7 7 ** Gender B 01 0.084 0.13 2 0.64 0.138 0.076 1.83 0. 166 0.1 65 1 01 Race B 02 0.390 0.204 1 91 0.038 0.065 0.58 0. 106 0. 207 0. 51 Job status, B 03 0.267 0.207 1 29 0.042 0.102 0 41 0.0 54 0. 319 0. 17 Supervisor tenure B 04 0.003 0.011 0 26 0.005 0.007 0. 72 0. 048 0 .02 8 1. 71 Job mobility, B 0 5 0.005 0.038 0.14 0.005 0.018 0.25 0. 053 0.0 75 0 70 Extraversion, B 0 6 0.046 0.067 0.68 0.004 0.031 0.14 0. 228 0.0 94 2. 43 Volatility, B 0 7 0.095 0.149 0.64 0.032 0.049 0.66 0.178 0. 127 1 40 Withdrawal, B 0 8 0.08 8 0.104 0.84 0.021 0.085 0.24 0.1 6 3 0.1 28 1 27 Compassion, B 0 9 0.053 0.108 0.49 0.008 0.050 0.16 0. 372 0. 217 1. 71 Supervisor interaction (t 1) B 10 0.001 0.002 0.93 0.000 0.001 0.23 0.00 3 0.002 1 47 Abusive supervision (t 1) B 20 0.447 0 .110 4.06 ** 0.256 0.066 3.88 ** 0.387 0.065 6.00 Abusive supervision (t 1) Personality trait, B 21 0. 121 0. 098 1.24 0.105 0.068 1.56 0.228 0.105 2.18 Pseudo R 2 0. 36 0.70 0.10 Notes: B = unstandardized regression coefficie nt. SE = standard error. T 1 indicates variable was measured one month prior to the dependent variable. Trait volatility is the proposed moderator of the abusive supervision anger relationship. Trait withdrawal is the proposed moderator of the abusive supervision fear relationship. Trait compassion is the proposed moderator of the abusive supervi sion compassion relationship. p < .05, two tailed. ** p < .01, two tailed.
94 Table 5 8 HLM regression results depicting personality traits as moderators of the relationship between abusive Supervisor directed d eviance Supervisor directed a voidance Supervisor directed c itizenship Parameter B SE T value B SE T value B SE T value Intercept, B 00 0.4 38 0.0 59 7. 46 0.332 0.077 4 31 ** 2. 454 ** 0. 2 5 7 9 56 ** Gender B 01 0.01 1 0.02 6 0.4 5 0.000 0.033 0.00 0. 0 63 0.1 1 0 0 57 Race B 02 0.05 9 0.04 8 1 2 5 0.126 0.054 2.34 0. 061 0. 1 43 0. 42 Job status, B 03 0.16 1 0.0 80 2 0 1 0.007 0.046 0 15 0.0 32 0. 232 0. 1 4 Supervisor tenure B 04 0.001 0.004 0 3 2 0.002 0.005 0. 42 0. 030 0.02 1 1 42 Job mobility, B 0 5 0.014 0.011 1.3 3 0.028 0.013 2.22 0. 12 7 0.0 51 2 50 Extraversion, B 0 6 0.017 0.019 0.87 0.005 0.017 0.32 0. 1 25 0.0 6 3 1 98 Volatility, B 0 7 0.08 6 0.05 8 1. 50 0.004 0.029 0.14 0.054 0. 090 0 60 Withdrawal, B 0 8 0.02 1 0.033 0.6 4 0.117 0.060 1.96 0.0 04 0.1 0 4 0 0 3 Compassion, B 0 9 0.0 32 0.025 1 30 0.057 0.030 1.89 0. 131 0. 162 0 81 Supervisor interaction, B 10 0.002 0.001 1.7 4 0.000 0.001 0.33 0.00 6 0.002 2 9 0 ** Dependent variable (t 1), B 20 0.51 4 0.06 4 8.09 ** 0.563 0.068 8.24 ** 0.2 7 7 0.05 4 5 18 ** Abusive supervision (t 1) B 30 0.0 90 0.062 1.4 6 0.180 0.077 2.32 0.1 6 0 0.117 1 36 Abusive supervision (t 1) Personality trait B 31 0.060 0.041 1.47 0.126 0.035 3.56 ** 0. 005 0. 113 0 04 Pseudo R 2 0.58 0. 71 0.30 Notes: B = unstandardized regression coefficient. SE = standard error. T 1 indicates variable was measured one month prior to the dependent variable. Trait volatility is the proposed moderator of the abusive supervision supervisor directed deviance relationship. Trait withdrawal is the proposed moderator of the abusive supervision supervisor directed avoidance relationship. Trait compassion is the proposed moderator of the abusive supervision supervisor dir ected citizenship relationship. p < .05, two tailed. ** p < .01, two tailed.
95 Table 5 9 abusive supervision A busive s upervision Parameter B SE T value Intercept, B 00 1 366 0. 062 22.18 ** Gender B 01 0.0 96 0.129 0 75 Race B 02 0.222 0. 104 2 12 Job status B 03 0. 253 0. 219 1 16 Supervisor tenure B 04 0. 006 0. 016 0. 41 Job mobility B 05 0. 029 0.0 33 0.90 Extr aversion B 06 0.0 99 0.0 44 2 27 ** Volatility B 07 0. 089 0. 071 1.26 Withdrawal B 08 0. 205 0. 094 2 18 Compassion, B 09 0.106 0.074 1.44 Supervisor interaction, B 10 0.001 0.001 0.45 Pseudo R 2 0.15 Notes: B = unstandardized regression coefficien t. SE = standard error. p < .05, two tailed. ** p < .01, two tailed.
96 Table 5 10 HLM regression results for the effects of personality traits on emotions Anger Fear Compassion Parameter B SE T value B SE T value B SE T value Inter cept, B 00 1.509 0.087 17.39 ** 1.245 0.053 23.53 ** 1.918 ** 0.096 20 00 ** Gender B 01 0.109 0.175 0.63 0.030 0.117 0.26 0.050 0.185 0.27 Race B 02 0.321 0.184 1 75 0.013 0.083 0.16 0.042 0.203 0.21 Job status, B 03 0.399 0.322 1 24 0.362 0.270 1.34 0.334 0.287 1.17 Supervisor tenure B 04 0.015 0.021 0 70 0.000 0.018 0.00 0.015 0.029 0.51 Job mobility, B 0 5 0.007 0.046 0.15 0.005 0.024 0.22 0.068 0.078 0.87 Extraversion, B 0 6 0.034 0.087 0.39 0.103 0.058 1.78 0.209 0.091 2.31 Vol atility, B 0 7 0.129 0.132 0.97 0.018 0.081 0.22 0.187 0.135 1.39 Withdrawal, B 0 8 0.224 0.153 1.46 0.199 0.089 2.24 0.108 0.141 0.76 Compassion, B 0 9 0.055 0.131 0.42 0.142 0.079 1.80 0.081 0.120 0.68 Supervisor interaction, B 10 0.001 0.002 0.77 0.000 0.001 0.39 0.009 0.002 3.73 ** Pseudo R 2 0.08 0.09 0.09 Notes: B = unstandardized regression coefficient. SE = standard error. p < .05, two tailed. ** p < .01, two tailed.
97 Table 5 11 HLM regression results for the e Deviance Avoidance Citizenship Parameter B SE T value B SE T value B SE T value Intercept, B 00 1.304 0.058 22.60 ** 1.473 0.075 19.57 ** 3.200 ** 0.103 31.08 ** Gender B 01 0.034 0.118 0.29 0.106 0.142 0.75 0.030 0.157 0.19 Race B 02 0.161 0.099 1.62 0.225 0.113 1.99 0.049 0.174 0.28 Job s tatus, B 03 0.372 0.192 1.93 0.265 0.182 1.45 0.033 0.275 0.12 Supervisor t enure, B 04 0.033 0.022 1.52 0.023 0.024 0.93 0.043 0.023 1.85 Job m obility, B 05 0.013 0.030 0.44 0.047 0.033 1.41 0.130 0.064 2.02 Extraversion, B 06 0.087 0.041 2.13 0.076 0.044 1.71 0.114 0.082 1.40 Volatility, B 07 0.013 0.081 0.16 0.038 0.088 0.43 0.135 0.127 1.06 Withdrawal, B 08 0.190 0.122 1.56 0.187 0.133 1.41 0.088 0.134 0.66 Compassion, B 09 0.164 0.079 2.08 0.183 0.086 2.13 0.211 0.122 1.73 Supervisor i nteraction, B 10 0.001 0.001 0.28 0.003 0.001 2.64 0.007 0.002 3.74 ** Pseudo R 2 0. 13 0.10 0.09 Notes: B = unstandardized regression coefficient. SE = standard error. p < .05, two tailed. ** p < .01, two tailed.
98 Table 5 12 between personality and abusi ve supervision Volatility abusive supervision r elationship Withdrawal abusive s uperv ision r elationship Compassion abusive supervision r elationship Parameter B SE T value B SE T value B SE T value Intercept, B 00 1 369 0. 046 29.9 7 ** 1.323 0.0 4 3 30 93 ** 1 3 5 3 ** 0. 061 22 06 ** Gender B 01 0.0 67 0.0 71 0 94 0.0 87 0.0 67 1 30 0. 0 80 0. 127 0 63 Race B 02 0.05 7 0.0 88 0 65 0.097 0.071 1. 3 6 0 213 0. 105 2 02 Job status, B 03 0.013 0. 104 0.12 0.108 0.0 80 1 36 0. 228 0. 225 1 02 Supervisor tenu re B 04 0.00 6 0.00 8 0 75 0.002 0.00 5 0. 31 0. 0 11 0.0 15 0 71 Job Mobility, B 0 5 0.030 0.0 26 1.1 6 0.0 1 2 0.021 0. 59 0. 0 37 0.0 32 1 1 7 Extraversion, B 0 6 0.0 60 0.0 29 2 0 4 0.026 0.02 3 1 13 0. 120 0.0 46 2 63 Volatility, B 0 7 0.0 6 0 0.0 4 6 1 30 0.0 8 9 0.0 36 2.47 0.07 6 0. 0 69 1 1 0 Withdrawal, B 0 8 0.068 0.04 9 1.41 0.0 38 0.0 42 0.90 0. 2 14 0. 0 93 2 29 Compassion, B 0 9 0.0 36 0.0 50 0 71 0.032 0.0 36 0.88 0. 0 88 0. 0 7 5 1 1 7 Supervisor interaction, B 10 0.001 0.001 1 15 0.00 1 0.001 0. 79 0.00 0 0.001 0 24 Proposed e motion mediator B 2 0 0.348 0.071 4.88 0.145 0.108 1.3 5 0.093 0.050 1.87 Proposed b ehavior mediator B 3 0 0.281 0.122 2.30 0.598 0.095 6 31 ** 0.015 0.077 0.20 Pseudo R 2 0.51 0.61 0.16 Notes: B = unstandardized reg ression coefficient. SE = standard error. Anger is the proposed emotion mediator and supervisor directed deviance is the proposed behavior mediator of the trait volatility abusive supervision relationship. Fear is the proposed emotion mediator and superv isor directed avo idance is the proposed behavior mediator of the trait withdrawal abusive supervision relationship. Compassion is the proposed emotion mediator and supervisor directed citizenship behavior is the proposed behavior mediator of the trait co mpassion ab usive supervision relationship. p < .05, two tailed. ** p < .01, two tailed.
99 Table 5 13. Summary of study findings Hypothesis Finding H1 Abusive supervision positively predicts Supported H2 itively predicts directed deviance Supported H3 relationship between abusive supervision and directed deviance Supported H4 Abusive supervision positively pred icts Supported H5 fear positively predicts supervisor directed avoidance. Supported H6 relationship between abusive supervision and directe d avoidance. Supported H7 Abusive supervision negatively predicts Supported H8 compassion positively predicts directed citizenship behavior Supported H9 compassion partially mediates the relationship between abusive supervision and directed citizenship behavior. Supported H10 supervisor directed deviance positively predicts abusive supervision. Not supported H11 r directed citizenship negatively predicts abusive supervision. Not supported H12 supervisor directed avoidance positively predicts abusive supervision. Supported H13 Trait volatility moderates the effect of abusive supervision on subordina positive relationship between abusive supervision subordinates who possess high levels of trait volatility. Not supported
100 Table 5 13. Continued Hypothesis Finding H14 Trait volatility moderates the effect of abusive directed deviance, such that the positive relationship supervisor directed deviance is amplified for subordinates who possess high levels of trait volatility Not supported H15a Trait volatility is positively related to abusive supervision. Not supported H15b Anger and supervisor directed deviance partially on abusive supervision. Not supported H16 Trait withdrawal moderates the effect of abusive positive relationship between abusive supervision subordinates who possess high levels of trait withdra wal. Not supported H17 Trait withdrawal moderates the effect of abusive directed avoidance, such that the positive relationship supervisor directed avoidance is amplifie d for subordinates who possess high levels of trait withdrawal. Supported H18a Trait withdrawal is positively related to abusive supervision. Supported H18b Fear and supervisor directed avoidance partially mediate the effects of trait withdrawal on abusi ve supervision. Not supported H19 Trait compassion moderates the effect of abusive such that the negative relationship between compassion becomes positive for subor dinates who possess high levels of trait compassion. Not supported
101 Table 5 13. Continued Hypothesis Finding H20 Trait compassion moderates the effect of abusive directed citizenship behavior, such that the ne gative relationship between abusive supervision and directed citizenship behavior becomes positive for subordinates who possess high levels of trait compassion. Not supported H21a Trait compassion negatively predicts abusive super vision. Not supported H21b State compassi on and supervisor directed citizenship behavior partially mediate the effects of trait compassion on abusive supervision. Not supported
102 Figure 5 1. Plot of the moderating effect of trait compassion on the rel ationship between abusive supervision and state compassion Figure 5 2 Plot of the moderating effect of trait withdrawal on the relationship between abusive sup ervision and supervisor directed avoidance
103 CHAPTER 6 DISCUSSION Perhaps because mistreatmen t can influence us so powerfully a burgeo ning literature has examined its prevalence and correlates in the workplace. Beyond these findings, however, much remains to be discovered regarding the dynamic nature of destructive workplace relationships and nu merous scholars have called for suc h research (Aquino & Thau, 2009; Barling et al., 2009; Bowling & Beehr, 2006; Rodriguez Munoz, Baillien, De Witte, Moreno Jimenez, & Pastor, 2009). wave longitudinal design was, in part, an effort to respond to these calls and allowed a number of dynamic processes potentially involved in the development of abusive supervisor subordinate relationships to be brought to light One of the most interesting set s of findings demonstrated that the causal nature of the relationships varie d with the type of behavior examine d Abusive supervision emerged as a predictor, but not as a consequence of supervisor directed deviance; abusive supervision and su pervisor directed avoidance were reciprocally related; and abusive supervision and supervisor directed citizenship behavior were unrelated directly R esults also reveal ed that emotions played a substantial role in mediating the effects of abusive supervisi abusive supervision supervisor directed deviance relationship and fear partially mediated the abusive supervision supervisor directed avoidance relationship Furthermore, although abusive supervi sion had no impact alone on superviso r directed citizenship behavior individuals who felt less compassion for their supervisors after falling prey to mistre atment tended to withhold such citizenship behavior
104 A final set of findings demonstrated that pers onality influenced the amount of abuse employees experienced as well as their reactions to mistreatment. In particular, compared to low trait withdrawal individuals high trait withdrawal individuals were more likely to endure abuse and, also, to steer cle ar of their supervisors in response to it Interestingly, although trait withdrawal moderated the abusive supervision supervisor directed avoidance relationship, and although it was linked to higher general levels of fear, trait withdrawal d id not influe nce whether individuals resp ond ed to mistreatment with fear It was also unrelated to supervisor directed avoidance and its effects on abusive supervision were not mediated by fear or supervisor directed avoidance Taken together, t his pattern of findings reveals an interesting dynamic H igh trait withdrawal individuals are more likely to experience overall levels of abusive supervision not because they are more likely to be fearful or avoidant in general, but at least, in part, because they are prone to r eacting to abuse with avoidance ( which, in turn, tends to incite further abuse ) One additional p ersonality trait influenced how subordinates res ponded to abusive supervision. C ontrary to my predictions high trait compassion individuals were especially la cking in state compassion after experiencing mistreatment Although u nexpected, th is finding does have a viable explanation Cognitive affective personality system s theory ( CAPS; Mis chel & Shoda, 1995) argues that behavi ors in given situations vary reliabl y within individuals not necessarily that a given trait predicts similar behaviors across all situations Applying this t heory to the current set of findings involving compassion means that even though some individuals might possess a tend ency to respond compassionately across several situations (typifying the high trait
105 compassion individual) in certain instances these individuals can feel reliably less compassionate than their l ow trait compassion counterparts. For instance, because of their u nderlying prosocial values, high trait compassion individuals might respond with heightened compassion to ward s someone who has been harmed but especially without compassion toward the perpetrator responsible for the harm T hese seemingly disparate responses are bo th consistent with the fundamental prosocial motivation often thought to cha racterize agreeable individuals, and may explain why hig h trait compassion individuals are especially lacking in compassion for abusive supervisors Theoretical Implications By she d ding light on the dynamic nature of certain supervisor subord inate social interactions, this stud y contribute s to our understanding of how s ubordinate behaviors can encourage and reinforce abusive supervision and thus how discrete episodes of mistreat ment can develop into dysfunctional relationship s over time Of each of the potential behavioral reactions to abusive supervision examined, supervisor directed avoidance appeared to be the most destructive, in that it was the most likely to and lead to more, not less, mistreatment. This reciprocal relationship suggests that, in some instanc es, abusive relationships are the product of self fulfilling prophecies (Merton, 1948) beliefs that either directly or indirectly cause themselves to be true a s a result of positive feedback between the belief and behavior. In th is instance, a subordinate who experience s an episode of mistreatment might come to believe that his or her supervisor is prone to treating employees poorly and thus, should be avoid ed. The supervisor, i n turn might mis interpret this reaction as indicative of a subor a characteristic perhaps perceived as deserving of further mistreatment S ubsequent abuse can reinforce the subordinate s initial belief
106 resultin g in additional avoidant behavior that further serves to reinforce the supervisor s belief and so on. Such interactions set the stage for a vicious cy cle o f dysfunctional behavior that can eventually transform a positiv e working relationship, punctuated by occasional unpleasant social interaction into a destructive one. Other behaviors, such as supervisor directed deviance and supervisor directed citizenship behavior did not appear to predict increases in abuse, and thu s, to be more effective coping mechanisms T he results of this study also have implications for victim precipitation theory, which proposes that victims may knowingly or unknowingly contribute to their becoming ta rgets of mistreatment (Amir, 1967; Curtis, 1974 ; Wolfgang, 1967 ). T hat the relationship between abusive supervision and avoidance was largely mediated by subordina fear, and that individuals who possessed a dispositional tendency to react to mistreatmen t with avoidance experienced higher levels of overall abuse lends credence to the notion that individuals who respond to abusive supervision with avoidance are lik According to Olweus (1978), submissive victims, who appea r to be anxious, sensitive, and low in self esteem are vulnerable to abuse because they are argets Contra r y to my expectations however, findings did not support the idea that individuals who engage in deviance or who are dispositionally aggressi ve (i.e., possessing high levels of trait volatility) become what Olweus (1978) described as provocative individuals prone to aggressive behavior that invite s retaliatory responses from others. Likewise, contra r y to other cross sectional findings ( e.g., Aquino & Bommer, 2003 ) I found no support for the proposition that acting fav orably toward
107 or that possessing high levels of trait compassion likel ihood of further victimization. With respect to deviance, p erhaps such behavior, acts as a deterrent for mistreatment in certain instances. This proposition is consistent with a growing stream of literature suggesting that anger and aggression can be functional in a number of circumstances (c.f., Averill, 1983). If true in certain instances for the supervisor directed deviance abusive supervision relationship, an y positive effect of supervisor directed deviance on abusive supervision might be suppressed. Thus, moderators of this relationship should be investigated to ga in a better understanding of the conditions under which supervisor directed deviance positively and negatively predicts abusive supervision. Regarding the supervisor directed citizenship abusive supervision relationship, o ne explanation for the failure o f supervisor directed citizenship behavior to predict abusive supervision might have to do with whether or not supervisors perceive d supervisor directed citizenshi p behavior to be of (i.e., n ot truly extra role ) More specifically, i f some s upervisors consider citizenship behavior to be within the realm of subor it seems unlikely that it would foster the positive reciprocity norms typically generated by behaviors viewed as more benevolent As a result, supervisors would not necessarily feel obligated to treat employees that engage in supervisor directed citizenship especially well. A final theoretical implication involves the role of emotions in adversarial interactions. The resul ts of this study suggest that being is a highly emotional experience E ven when controll ing for personality traits that are
108 especially associated with the tendency to experience emotion (i. e., extraversion, neuroticism) s ubordinates who perceived that their sup ervisor had mistreated them tended to feel angrier, more fearful, and less compassionate toward their supervisor. Furthermore, t hat fear partially mediated the relationship between abusive supervision and supervisor directed avoidance and that anger fully mediated the relationship between abusive supervision and supervisor directed deviance provides evidence that these e motions play a central role in fueling to abusive supervision. Limitations and Future Research The resul ts of this study should be interpreted with its limitations in mind One limitation is that the data collected to assess many of the relationships in this study w ere gathered from a single source raising the possibility that common method variance might h ave inflate d some of the effect sizes (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). However, many aspects of the study design should mitigate this concern. First, for several of the analyses, the independent and dependent measures were separated in time. Temporal separation diminishes the transfer of contextual ly relevant retrieval cues from one time point to another because previously recalled information is likely to (Podsakoff et al., 2003) By controlling for ext raversion and neuroticism (i.e., trait volatility and trait withdrawal), the possibility that trait affectivity is responsible for the significant relationships found in this study has also been reduced. Finally, the use of within person analyses help s to rule out individual d ifferences and stable environmental factors as possible explanations for the
109 A more substantive limitation stemming from the fact that abusive supervis ion and solely by subordinates, is that the results of this study do little to show exactly what the driving forces are behind supervisors to subo Although theory provides convincing ar gu ments for why supervisors respond to subordinates in certa in ways future research is necessary to bolster these arguments with empirical support. Relatedly, it would also be beneficial to obtain a more accurate view of how supervisor and subordinates p ersonalities might interact to inf luence patterns of mistreatment. To do so, researchers should collect longitudinal data from both supervis ors and subordinates Finally, although I have discussed several advantages of using longitudinal data to understand dynamic relational patterns, there are some potential drawbacks to measuring predictor and criterion variables separately One such drawback involves the appropriateness of the lag time between measurements If lag times are too long, they can result in i ncreased sample attrition and may also mask relationships that truly exist if the causal effect fades over time (Podsokoff et al 2003). Likewise, if lag times are too short, the process of interest will not have sufficient time to unfold. Thus, it is ess ential that the lag times incorporated into longitudinal study designs correspond to the process of interest as it naturally occurs However, as noted by Selig and Preacher (2009) often there is no theoretical or emp (p. 1 50 ) For this study the choice of a lag time was especially difficult. On the one hand, emotions tend to be fleeting and short lived, whereas relational patterns may take a comparatively longer time to establish. In light of trying to capture both emotion al and relational aspects of abusive supervision, a one month lag seemed appropriate. However, future
110 research may wish to implement other lag times to better understand how they influence findings. Practical Implications Unlike with sexual and physical ha rassment, in the United States, there is no legislation to guard employees from psychological harassment However, there are hints that the legal landscape may soon change. Laws already protect worker s in Australia, Canada, Sweden, and the United Kingdom and a t least 16 U.S. states have proposed legislation Thus, although it is important to better understand how abusive supervision and related phenomena (e.g., workplace bullying) emerge and develop over time for financial and psychological reasons, such a n understanding also seems especially important for legal reasons To this end it is crucial to identify the mechanisms that perpetuate abusive su pervision so that they c an be targeted for training and intervention Obtaining a n understanding of dispositi onal predictors of abusive supervision would also be useful in enabling managers to identify applicants and employees most likely to become victims and perpetrators of mistreatment This study helps to accomplish some of these objectives Knowledge that su n can help sustain mistreatment suggests that training programs and interventions targeting both subordinates a nd supervisors rather than only supervisor s would be useful in stopping or preventing abusive supervi sion. Such programs might provide organizational members with information about what abusive and avoidant behaviors entail as well as with information about how these behaviors can destructively reinforce one another. Training of this nature should enable supervisors and subordinates to better recognize when their own actions or
111 those of their coworkers are fueling a vicious cycle of behavior one that could ultimately mature into a full fledged, psychologically abusive relationship. In cases when it is impr actical to provide such train ing for all organizational members, it would be sensible to identify individuals who are most likely to be victimized for enrollment into these programs. T he finding that high trait withdrawal individuals are both more likely t o react to abuse in a manner that reinforce s it (i.e., with avoidance) and more likely to experience higher levels of abusive supervisi on suggests that selecting candidates based on levels of trait withdrawal would be one way to accomplish this goal. Of co urse, it would als o be useful to identify c haracteristic s of those who are prone to becoming abusive so that they too can enroll in training programs. Unfortunately, the data collected for this study do not allow me to identify any such characteristics and few studies have explored t hem Thus, this is another area in much need of future research. Conclusion For both instrumental and affective reasons, developing and maintaining positive social relationships is vital to our well being. Our social environm 2000, p.1026). They shape our identities, help us make sense of our surroundings, and fulfill our innate needs for belongingness ( Baumeister & Leary, 199 5 ; Chiaburu & Harrison, 2008). What is more, when social relationships become adversarial, we are often deeply affected ; research suggests that the damaging effects of negative social interaction can far outweigh the benefits derived from interactions more positive in nature (Rook, 1984).
112 This dissertation provided insight into how such adversarial relationships particularly those among supervisors and subordinates come to be. F indings show that, at least in some instances, victims and perpetrators in abusi ve relationships engage in behaviors that are mutually reinforcing These findings are important because they can be used to arm supervisors and subordinates with the knowledge to understand when and why des tructive relational patterns are beginning to occ ur S uch knowledge I hope will empower individuals to alter their behavior and prevent from cascading into the we know to be so psychologically damaging
113 APPENDIX A LIST OF ITEMS USED I N INITIAL SURVEY (CO WORKER R EPORT) Trait v olatility I see my coworker as someone who... Get s angry easily. Rarely get s irritated. Get s upset easily. Keep s his/her emotions under control. Changes his/her mood a lot. Rarely lose s his/her composure. Is a person wh ose moods go up and d own easily. Is not easily annoyed. Get s easily agitated. Can be stirred up easily. Trait w ithdrawal I see my coworker as someone who... Seldom feel s blue. Is f illed with doubts about things. Feel s comfortable with his/her self. Feel s threatened easily. Rarely feel s depressed. Worries about things. Is easily discouraged. Is not embarrassed easily. Become s overwhelmed by events. Is afraid of many things. Trait c ompassion I see my coworker as someone who... Is not interes F eel s Inquire s being. be bothered with other s needs. Sympathize s Is indiffe rent to the feelings of others. Take s no time for others. Take s Do es soft side.
114 Like s to do things for others. Extraversion I see my coworker as someone who... Is the life of the party. Talks to a lot of different people at parties. Keeps in the background. Abusive s upervision During the past six m onths my coworker's immediate supervisor has : Ridiculed my coworker Told my coworker his/her thoughts and feelings were stupid. Gi ve n my coworker the silent treatment. Put my coworker down in front of others. Invaded my privacy. Reminded my c oworker of his/her past mistakes and failures. Not given my coworker credit for jobs requiring a lot of effort. Blamed my coworker to save him/herself embarrassment. Broken promises he/she made to my coworker. Expressed anger at my coworker when he/she was mad for another reason. Made negative comments about my coworker to others. Been rude to my coworker N ot allow ed my coworker to interact with other coworkers. Told my coworker he/she is incompetent. Lied to my coworker
115 APPENDIX B LIST OF ITEMS USED I N INITIAL SURVEY (SE LF REPORT) Demographics What is your gender? (Male, Female) What is your ethnicity? (Caucasian, African American, Hispanic, Native American, Asian/Pac Isle, Other) About how long have you been reporting to your current immediate superv isor? What best characterizes the nature of your present job position? (Entry level; Non managerial, experienced employee; Managerial, non professional; Non managerial, professional; Managerial, professional; CEO or to p executive) Job m obility I would ha ve no problem finding an acceptable job if I quit. If I were to quit my job, I could find ano ther one that is just as good. Abusive s upervision During the past six months my immediate supervisor has: Ridiculed me Told me my thoughts and feelings were stupid. Given me the silent treatment. Put me down in front of others. Invaded my privacy. Reminded me of my past mistakes and failures. Not given me credit for jobs requiring a lot of effort. Blamed me to save him/herself embarrassment. Broken promises he /she made. Expressed anger at me when he/she was mad for another reason. Made negative comments about me to others. Been rude to me N ot allow ed me to interact with my coworkers. incompetent. Lied to me
116 APPENDIX C LIST OF ITEMS USED I N MONTHL Y SURVEY Anger During the past month when thinking about or interacting with my immediate supervisor I have felt... Angry Hostile Enraged Fear During the past month when thinking about or interacting with my immediate supervisor I have felt... Afr aid Scared Frightened Nervous Jittery Shaky Compassion During the past month when thinking about or interacting with my immediate supervisor I have felt... Compassionate Sympathetic Moved Tender Warm Softhearted Supervisor directed deviance During t he past month I Made fun of my supervisor. Acted rudely toward my supervisor. Gossiped about my supervisor. Publicly embarrassed my supervisor. Refused to talk to my supervisor. Said something hurtful to my supervisor. Swore at my supervisor.
117 Supe rvisor directed avoidance During the past month I Avoided my supervisor. Lived as Kept as much distance as possible between my supervisor and me. Went the other way when I saw my supervisor coming. Pretended to be si ck and stay home in order to avoid my supervisor. Took vacation days in order to avoid any interaction with my supervisor. conversation with me. Supervisor directed citizenship Dur ing the past month I Accepted added responsibility to help my supervisor. Helped my supervisor when he/she had a heavy workload. Assisted my supervisor with his/her work when not asked. Took a pers onal interest in my supervisor. Passed along work r elated information to my supervisor. Abusive s upervision During the past month my immediate supervisor has... Ridiculed me Told me my thoughts and feelings were stupid. Given me the silent treatment. Put me down in front of others. Invaded my privacy. Reminded me of my past mistakes and failures. Not given me credit for jobs requiring a lot of effort. Blamed me to save him/herself embarrassment. Broken promises he/she made. Expressed anger at me when he/she was mad for another reason. Made negative com ments about me to others. Been rude to me N ot allow ed me to interact with my coworkers. incompetent. Lied to me Time spent interacting with supervisor
118 During the past month roughly what percentage of your time at work did you spend interac ting with your immediate supervisor ?
119 LIST OF REFERENCES Allred, K. G. (1999). Anger and retaliation: Toward an understanding of impassioned conflict in organizations. Research on Negotiation in Organizations, 7 27 58. Amir, M. (1967). Victim precipitat ed forcible rape. The Journal of Criminal Law, 58 493 502. Andersson, L. M., & Pearson, C. M. (1999). Tit for Tat?: The spiraling effect of incivility in the workplace. Academy of Management Review, 24 452 471. Aquino, K. (2000). Structural and individ ual determinants of workplace victimization: The effects of hierarchical status and conflict management style. Journal of Management, 26 171 193. Aquino, K., & Bommer, W. (2003). Preferential Mistreatment: How Victim Status Moderates the Relationship Bet ween Organizational Citizenship Behavior and Workplace Victimization. Organization Science 14 374 385. Aquino, K., & Bradfield, M. (2000). Perceived victimization in the workplace: The role of situational factors and victim characteristics. Organization Science, 11 525 537. Aquino, K., & Byron, K. (2002). Dominating interpersonal behavior and perceived victimization in groups: Evidence for a curvilinear relationship. Journal of Management, 28 69 87. Aquino, K., & Douglas, S. (2003). Identity threat a nd antisocial behavior in organizations: The moderating effects of individual differences, aggressive modeling, and hierarchical status. Organizational Behavior and Human Decision Processes, 90 195 208. Aquino, K., Douglas, S., & Martinko, M. (2004). Ove rt Anger in Response to Victimization: Attributional Style and Organizational Norms as Moderators. Journal of Occupational Health Psychology 9 152 164. Aquino, K., Grover, S. L., Bradfield, M., & Allen, D.G. (1999). The effects of negative affectivity, hierarchical status, and self determination on workplace victimization. Academy of Management Journal, 42 260 272. shove: Interpersonal forgiveness in workplace relati onships Journal of Management Inquiry, 12 209 216. perspective. Annual Review of Psychology, 60 717 741.
120 Aryee, S., Chen, Z. X., Sun, L., & Debrah, Y. A. (2007). Ante cedents and outcomes of abusive supervision: Test of a trickle down model. Journal of Applied Psychology, 92 191 201. Aryee, S., Chen, Z. X. G. C., & Debrah, Y. A. (2008). Abusive supervision and contextual performance: The mediating role of emotional ex haustion and the moderating role of work unit structure. Management and Organization Review, 4 393 411. Ashforth, B. E. (1994). Petty tyranny in organizations. Human Relations, 47 755 778. Averill, J. R. (1983). Studies on anger and aggression: Implica tions for theories of emotion. American Psychologist, 38 1145 1160. Bamberger, P. A., & Bacharach, S. B. (2006). Abusive supervision and subordinate problem drinking: Taking resistance, stress and subordinate personality into account. Human Relations, S9 723 752. Bandura, A. (1978). The self system in reciprocal determinism. American Psychologist, 33 344 358. Bandura, A. (1983). Temporal dynamics and decomposition of reciprocal determinism: A reply to Phillips and Orton. Psychological Review, 90 166 170. Bandura A. (1997). Self efficacy: The exercise of control New York: Freeman. Barling, J., Dupre, K. E., & Kelloway, E. K. (2009). Predicting workplace aggression and violence. Annual Review of Psychology, 60 671 6 92. Baron, R. M., & Kenny, D. A. (1986). The moderator mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51 1173 1182. Basch, J., & Fisher, C. D. (2000). Affective events emotion matrix: A classification of work events and associated emotions. In N. M. Ashkanasy, C. E. J. Artel, & W. J. Zerbe (Eds.), Emotions in the workplace: Research, theory, and practice (pp. 36 48). Batson, C. D., Fultz, J., & Schoenrade, P. A. (1987). Distress and empathy: Two qualitatively distinct vicarious emotions with different motivational consequences. Journal of Personality, 55, 19 39. Quin, K., Fultz, J., Vanderplas, M., & Isen, A. M. (1983). Influence of self reported distr ess and empathy on egoistic versus altruistic motivation to help. Journal of Personality and Social Psychology, 45, 706 718.
121 Batson, C. D., Sager, K., Garst, E., & Kang, M. (1997). Is empathy induced helping due to self other merging? Journal of Personal ity and Social Psychology, 73, 495 509. Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin 117, 497 529. Bies, R. J., & Tripp, T. M. (2005). The st udy of revenge in the workplace: Conceptual, ideological, and em pirical issues. In S. Fox & P. E. Spector (Eds.), Counterproductive work behavior: Investigations of actors and targets (pp. 13 40). Washington DC: American Psychological Association. Bjrkqv ist K, sterman K, Hjelt Bck M. (1994). Aggression among university employees. Aggressive Behavior, 20 173 83 Blau, P. M. (1964). Exchange and Power in Social Life New York: John Wiley Bolino, M. C. (1999). Citizenship and impression management: Goo d soldiers or good actors ? Academy of Management Review, 24 82 98. perspective: A theoretical model and meta analysis. Journal of Applied Psychology, 91 998 1012. Bradfield, M. & Aquino K. (1999). The effects of blame attribution and offender likableness on forgiveness and revenge in the workplace. Journal of Management, 25 607 631. Breaux, D., M., Perrewe, P. L., Hall, A. T., Frink, D. D., & Hochwater, W. A. (2008). Time t o try a little tenderness?: The detrimental effects of accountability when coupled with abusive supervision. Organizational Studies, 15 111 122. Bryk, A., & Raudenbush, S. W. (1992). Hierarchical linear models for social and behavioral research: Applicat ions and data analysis methods Newbury Park, CA: Sage. Burris, E. R., Detert, J. R., & Chiaburu, D. S. (2008). Quitting before leaving: The mediating effects of psychological attachment and detachment on voice. Journal of Applied Psychology, 93 912 922 Burton, J., & Hoobler, J. M. ( 2006 ) Subor dinate self esteem and abusive supervision. Journal of Managerial Issues, 18 340 355.
122 Bushman, B., Baumeister, R., Thomaes, S., Ryu, E., Begeer, S., & West, S. (2009). Looking again, and harder, for a link be tween low self esteem and aggression. Journal of Personality 77 427 446. Chiaburu, D. S., & Harrison, D. A. (2008). Do peers make the place? Conceptual synthesis and meta analysis of coworker effects on perceptions, attitudes, OCBs, and performance. Jou rnal of Applied Psychology, 93 1082 1103. Cialdini, R. B. (2001). Influence: Science and Practice Boston: Allyn and Bacon. Costa, P. T., Jr., & McCrae, R. R. (1992). Four ways five factors are basic. Personality and Individual Differences, 13 653 665. Crossley, C. D. (2009). Emotional and behavioral reactions to social undermining: A closer look at perceived offender motives. Organizational Behavior and Human Decision Processes, 108 14 24. Cropanzano, R., & Mitchell, M. S. (2005). Social exchange th eory: An interdisciplinary review. Journal of Management, 31 874 900. Curtis, L. A. (1974). Victim precipitation and violent crime. Social Problems 21 594 605. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self determination in human be haviour New York: Pienum. Detert, J. R., Trevino, L. K., Burris, E. R., & Andiappan, M. (2007). Managerial modes of influence and counterproductivity in organizations: A longitudinal business unit level investigation. Journal of Applied Psychology, 92 9 93 1005. DeYoung, C. G., Quilty, L. C., & Peterson, J. B. (2007). Between facets and domains: 10 aspects of the Big Five. Journal of Personality and Social Psychology, 5 880 896. Digman, J. M. (1990). Personality structure: Emergence of the five factor model. Annual Review of Psychology, 41, 417 440. Duffy, M. K., & Ferrier, W. J. ( 2003 ) Birds of a feather. ? : How supervisor subordinate dissimilarity moderates the influence of supervisor behaviors on workplace attitudes. Group and Organization Man agement, 28 217 248. Duffy, M. K., Ganster, D., & Pagon, M. (2002). Social undermining in the workplace. Academy of Management Journal 45 331 351.
123 Duffy, M. K., Ganster, D. C., Shaw, J. D., Johnson, J. L., & Pagon, M. (2006). The social context of und ermining behavior at work. Organizational Behavior and Human Decision Processes, 101 105 126. Dupre, K. E., Inness, M., Connelly, C. E., Barling, J., & Hoption, C. (2006). Workplace aggression in teenage part time employees. Journal of Applied Psychology 91 987 997. Einarsen, S. (2000). Harassment and bullying at work: A review of the Scandanavian Approach. Aggression and Violent Behavior, 5 379 401. Elias, R. (1986). The politics of victimization: Victims, victimology, and human rights New York: Ox ford University Press. Folger, R., & Cropanzano, R. (2001). Fairness theory: Justice as accountability. In J. Greenberg & R. Cropanzano (Eds.), Advances in organizational justice (pp. 1 55). Stanford, CA: Stanford University Press. Folger, R., & Skarlick i, D. (2005). Beyond Counterproductive Work Behavior: Moral Emotions and Deontic Retaliation Versus Reconciliation. Counterproductive work behavior: Investigations of actors and targets (pp. 83 105). Washington, DC US: American Psychological Association. Frazier, P., Mortensen, H., & Steward, J. (2005). Coping Strategies as Mediators of the Relations Among Perceived Control and Distress in Sexual Assault Survivors. Journal of Counseling Psychology 52 267 278. Frijda, N. (1986). The emotions New York: C ambridge University Press. Furr, R. M., & Funder, D. C. (1999). A multi modal analysis of personality negativity. Personality and Social Psychology, 74 1580 1591. Gant, L. M., Nagda, B. A., Brabson, H. V., Jayaratne, S., Chess, W. A., & Singh, A. (1993) perceptions of coworker and supervisor relationships and psychological well being Social Work, 38 158 64. Gardner, W. L., Rozell, E. J., & Walumbwa, F. O. (2004). Positive and neg ative affect and explanatory style as predictors of work attitudes. In M. J. Martinko (Ed.), Attribution theory in the organizational sciences: Theoretical extensions and empirical contributions (pp. 49 81) Greenwich, CT: Information Age Publishing. Gard ner, D., Van Dyne, L., & Pierce, J. (2004). The effects of pay level on organization based self esteem and performance: A field study. Journal of Occupational and Organizational Psychology 77 307 322.
124 Georgesen, J., & Harris, M. (1998). Why's my boss al ways holding me down? A meta analysis of power effects on performance evaluations. Personality and Social Psychology Review, 2 184 195 Gersick, C. J. G., Bartuneck, J. M., & Dutton, J. E. (2000). Learning from academia: The importance of relationships i n professional life. The Academy of Management Journal, 43 1026 1044. Gibson, D., & Callister, R. (2010). Ang er in organizations: Review and integration. Journal of Management 36 66 93 Goetz, J. L., Keltner, D., & Simon Thomas, E. (2010). Compassion: An evolutionary analysis and empirical review. Psychological Bulletin, 136 351 374. Gordon, J. (1992). Work teams: How far have they come? Training, 29, 59 65. Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American Sociologi cal Review, 25 161 178. Grandey, A. A., & Kern, J. (2004). Biting the hand that serves them: When does customer aggression predict employee exhaustion? Unpublished manuscript, Penn State University, University Park. Grandey, A. A., Kern, J. H., & Frone, M. R. (2007). Verbal abuse from outsiders versus insiders: Comparing frequency, impact on emotional exhaustion, and the role of emotional labor. Journal of Occupational Health Psychology, 12 63 79. B., van den Bos W., Rombouts S. A. R. B., & Crone, E. A. (in press). Unfair? It depends: Neural correlates of fairness in social context. Social Cogn itive & Affective Neuroscience. Harris, K. J., Kacmar, K. M., & Zivnus ka, S. (2007). An investigation of abusive supervision as a predictor of performance and the meaning of work as a moderator of the relationship. The Leadership Quarterly, 18 252 263. Harvey, P., Stoner, J., Hochwater, W., & Kacmar, C. (2007). Coping with abusive supervision: The neutralizing effects of ingratiation and positive affect on negative employee outcomes. The Leadership Quarterly, 18 264 280. Hazebroeck, J. F., Howells, K., & Day, A. (2001). Cognitive appraisals associated with high trait ange r. Personality and Individual Differences, 30 31 45. Hmieleski, K., & Ensley, M. (2007). The effects of entrepreneur abusive supervision. Academy of Management Proceedings 1 6
125 Hobman, E. V., Restubog, S. L. D., Bordia, P., & Tang, R. L. (2009). Abusiv e supervision in advising relationships: Investigating the role of social support. Applied Psychology: An International Review, 58 233 256. Hoel, H., & Cooper, C. L. (2001). Origins of bullying: Theoretical frameworks for explaining workplace bullying. I n N. Tehrani (Ed.), Building a culture of respect: Managing bullying at work : pp. 3 19. London: Taylor & Francis. Hofball, S.E. (1989). Conservation of resources: A new attempt at conceptualizing stress. American Psychologist, 44 513 524. Hollinger, R. C. (1986). Acts against the workplace: Social bonding and employee deviance. Deviant Behavior, 7 53 75. Hoobler, J., & Brass, D. (2006). Abusive Supervision and Family Undermining as Displaced Aggression. Journal of Applied Psychology 91 1125 1133. In ness, M., Barling, J., & Turner, N. (2005). Understanding supervisor targeted aggression: A within person between jobs design. Journal of Applied Psychology, 90 731 739. Inness, M., Leblanc, M. M., & Barling, J. (2008). Psychosocial predictors of supervi sor peer subordinate and service provider targeted aggression. Journal of Applied Psychology, 93 1401 1411. Jang, K. L., Livesley, W. J., Angleitner, A., Reimann, R., & Vernon, P. A. (2002). Genetic and environmental influences on the covariance o f facets defining the domains of the five factor model of personality. Personality and Individual Differences, 33, 83 101. Judge, T. A., Klinger, R., Simon, L. S., & Yang, I. W. F. (2008). The contributions of personality to organizational behavior and ps ychology: Findings, criticisms, and future research directions. Social and Personality Psychology Compass, 2 1982 2000. Judge, T. A., Locke, E. A., & Durham, C. C. (1997). The dispositional causes of job satisfaction: A core evaluations approach. Researc h in Organizational Behavior, 19, 151 188. Keashly, L. (1998). Emotional abuse in the workplace: conceptual and e mpirical issues. Journal of Emotional Abuse, 1 85 117 Kiazad, K., Restobug, S. L. D., Zagenczyk, T. J., Kiewitz, C., & Tang, R. L. (2010). In pursuit of power: The role of authoritarian leadership in the relationship between supervisory behavior. Journal of Research in Personality, 44 512 519.
126 Kim, T., Rosen, B., & Lee, D (2009). South Korean managerial reactions to voicing discontent: The effects of employee attitude and employee communication styles. Journal of Organizational Behavior, 30 1001 1018. Kish Gephart, J. J., Detert, J. R., Trevino, L. K., & Edmonson, A. C. (2009). Silenced by fear: The nature, sources, and consequences of fear at work. Research in Organizational Behavior, 29 163 193. Kozlowski, S. W. J. (2009). Editorial. Journal of Applied Psychology, 94 ,1 4. Lazarus, R. S. (1991a) Cognition and motivat ion in emotion. American Psychologist, 46 352 367. Lazarus. R. S. (1991b). Emotion and adaptation New York: Oxford University Press. Lecky, P. (1945). Self consistency: A theory of personality New York: Island Press. Lee, J. (1998). Maintenance commu nication in superior subordinate work relationships: Southern Communication Journal, 63 144 159. Leymann, H. & Gustafsson, A. (1996). Mobbing at work and the development of post traumatic stress disorders. European Journal of Work and Organizational Psychology, 5 251 275. Lind, E. A., & Van den Bos, K. (2002). When fairness works: Toward a general theory of uncertainty management. In B. M. Staw & R. M. Kramer (Eds.), Research in organizational behavior (Vol. 24, pp. 181 222). Boston: JAI Press MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of methods to test mediation and other intervening variable effects. Psychological Method s, 7 83 104. Magee, J. C., & Galinsky, A. D. (2008). Social hierarchy: The self reinforcing nature of power and status. Academy of Management Annals, 2 351 398. Malatesta, R. M., (1995). Understanding the dynamics of organizational and supervisory comm itment using a social exchange framework. Unpublished doctoral dissertation, Wa yne State University, Detroit, Michigan. Malle, B. F. (2006). The actor observer asymmetry in attribution: A (surprising) meta analysis. Psychological Bulletin, 132 895 819. Martinko, M. J., Harvey, P., Sikora, D., & Douglas, S. C. (2009) Abusive supervision: Perception or reality? Academy of Management Proceedings, 1 7
127 McCullough, M. E. (2001). Forgiveness: Who does it and how do they do it? Current Directions in Psychologi cal Science, 10 194 197. Merton, R. K. (1948). The self fulfilling prophecy. The Antioch Review, 8 193 210. Mitchell, M. S., & Ambrose, M. L. (2007). Abusive supervision and workplace deviance and the moderating effects of negative reciprocity beliefs. Journal of Applied Psychology, 92, 1159 1168. Mischel, W., & Shoda, Y. (1995). A cognitive affective system theory of personality: Reconceptualizing situations, dispositions, dynamics, and invariance in personality structure. Psychological Review, 102 2 46 268. Moss, S., Valenzi, E., & Taggart, W. (2003). Are you hiding from your boss? The development of a taxonomy and instrument to assess the feedback management behaviors of good and bad performers. Journal of Management 29 487 510. Neuman, J. H., & Baron, R. A. (2005). Aggression in the workplace: A social psychologic al perspective. In S. Fox & P. E. Spector (Eds.), Counterproductive work behavior: Investigations of actors and targets (pp. 13 40). Washington DC: American Psychological Association. O lweus. D. (1978). Aggression in Schools: Bullies and Whipping Boys Washington, D.C.: Hemisphere. Parker, J. G., & Asher, S. R. (1987). Peer relations and later personal adjustment: Are low accepted children at risk? Psychological. Bulletin, 102 357 389. Pearson, C. M., & And ersson, L. M., & Porath, C. L. (2005). Workpla ce incivility. In S. Fox & P. E. Spector (Eds.), Counterproductive work behavior: Investigations of actors and targets (pp. 13 40). Washington DC: American Psychological Association. Pe rrewe, P., & Ganster, D. (1989). The impact of job demands and behavioral control on experienced job stress. Journal of Organizational Behavior 10 213 229. Pierce, J. L., Gardner, D. G., Cummings, L. L., Dunham, R. B. (1989). Organization based self es teem: Construct definition, measurement, and validation. Academy of Management Journal, 32 622 648. Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literatur e and recommended remedies. Journal of Applied Psychology, 88 879 903.
128 Porath, C. L., Overbeck, J. R., & Pearson, C. M. (2008). Picking up the gauntlet: How individuals respond to status challenges. Journal of Applied Social Psychology, 38 1945 1980. P orath, C. L., & Pearson, C. M. (2010). The cost of bad behavior. Organizational Dynamics, 39 64 71. Rafferty, A. E., Restobug, S. L. D., & Jimmieson, N. L. (2010). Losing sleep: Examining psychological health. Work & Stress, 24 36 55. Rath, T. (2006 ). Vital Friends : New York : The Gallup Press Raudenbush, S., Bryk, A., & Congdon, R. (2005). HLM 6.02 for Windows [Computer software] Lincolnwood, IL: Scientific Software International Raynor, C., & Hoel, H. (1997). A summary of literature relating to workplace bullying Journal of Community and Applied Social Psychology, 7 181 191. Robinson, S. L., & Bennett, R. J. (1995). A typol ogy of deviant workplace behaviors: A multidimensional scaling study. Academy of Management Journal, 38 555 572. Rodell, J. B., & Judge, T. A. (2009). Can "good" stressors spark "bad" behavior?: The mediating role of emotions in links of challenge and hi ndrance stressors with citizenship and counterproductive behaviors. Journal of Applied Psychology, 94 1438 1451 Rodriguez Munoz, A., Baillien, E., De Witte, H., Moreno Jimenez, B., & Pastor, J. C. (2009). Cross lagged relationships between workplace bul lying, job satisfaction and engagement: Two longitudinal studies. Work & Stress, 23 225 243. Roesch, S. C., & Rowley, A. A. (2005). Evaluating and developing a multidimensional, dispositional measure of appraisal. Journal of Personality Assessment, 85 1 88 196. Rook, K. S. (1984). The negative side of social interaction: Impact on psychological well being. Journal of Personality and Social Psychology 46 1097 1108. Roseman, I., Wiest, C., & Swartz, T. (1994). Phenomenology, behaviors, and goals differe ntiate discrete emotions. Journal of Personality and Social Psychology 67 206 221. Rospenda, K. M., Richman, J. A., Wislar, J. S., & Flaherty, J. A. (2000). Chronicity of sexual harassment and generalized work place abuse: Effects on drinking outcomes. Addiction, 95 1805 1820.
129 Ross, L. (1977). The intuitive psychologist and his shortcomings: Distortions in the attribution process. In L. Berkowitz (Ed.), Advances in experimental social psychology (vol. 10, pp. 173 220). New York: Academic Press. Rudolp h, U., Roesch, S. C., Greitemeyer, T., & Weiner, B. (2004). A meta analytic review of help giving and aggression from an attributional perspective: Contributions to a general theory of motivation. Cognition and emotion, 18 815 848. Schat, A. C. H., Desma rais, S., & Kelloway, E. K. (2006). Exposure to workplace aggression from multiple sources: Validation of a measure and test of a model Unpublished manuscript, McMaster University, Hamilton, Canada. Schaubhut, N., Adams, G. A., & Jex, S. M. (2004). Self esteem as a moderator of the relationships between abusive supervision and two forms of workplace deviance Paper presented at the annual meeting of the Society for Industrial Organizational Psychology, Chicago. Selig, J. P., & Preacher, K. J. (2009). Med iation models for longitudinal data in developmental research. Research in Human Development, 6 144 164. Sims, H. P., Jr., & Manz, C. C. (1984). Observing leader verbal behavior: Toward reciprocal determinism in leadership theory. Journal of Applied Psyc hology, 69 222 232. Snijders, T. A. B., & Bosker, R. J. (1999). Multilevel analysis: An introduction to basic and advanced multilevel modeling Newbury Park, CA: Sage. Smith, C., & Ellsworth, P. (1985). Patterns of cognitive appraisal in emotion. Journa l of Personality and Social Psychology 48 813 838. Smith, C. A., Organ, D. W., & Near, J. P. (1983). Organizational citizenship behavior: Its nature and antecedents. Journal of Applied Psychology, 68 653 663. Sobel, M. E. (1982). Asymptotic intervals for indirect effects in structural equations models. In S Leinhart (Ed.), Sociological Methodology (pp. 290 312). San Francisco: Jossey Bass Spector, P. E., & Fox, S. (2005). The stressor emotion model of counterprodu ctive behavior. In S. Fox & P. E. S pector (Eds.), Counterproductive work behavior: Investigations of actors and targets (pp. 13 40). Washington DC: American Psychological Association. Tafrate, R. C., Kassinove, H., & Dundin, L. (2002). Anger episodes in high and low trait anger community adults. Journal of Clinical Psychology, 5 8, 1573 1590.
130 Tepper, B. J. (2000). Consequences of abusive supervision. Academy of Management Journal, 43 178 190. Tepper, B. J. (2007). Abusive supervision in work organizations: Review, synthesis, and researc h agenda. Journal of Management, 33 261 289. Tepper, B. J., Carr, J., Breaux, D. M., Geider, S., Hu, C., & Hua, W. (2009). Abusive supervision, power/dependence analysis. Organizational Behavior an d Human Decision Processes, 109 156 167. Tepper, B. J., Duffy, M. K., & Shaw, J. D. ( 2001 ) Personality moderators of the Journal of Applied Psychology, 86 : 974 983 Tepper, B. J., Duffy, M. K., Henle, C. A., & Lambert, L. S. (2006). Procedural justice, victim precipitation, and abusive supervision. Personnel Psychology, 59 101 123. Tepper, B. J, Duffy, M. K., Hoobler, J. H., & Ensley, M. D. (2004). Moderators of the relationships Journal of Applied Psychology, 89 455 465. Tepper, B. J., Henle, C. A., Lambert, L. S., Giacalone, R. A., & Duffy, M. K. (2008). organizational deviance. Journal of Applied Psychology, 93 721 732. Tepper, B. J., Moss, S. E., Lockhart, D. E., & Carr, J. C. (2007). Abusive supervision, Academy of Management Journal, 50 1169 1180. Thau, S., Bennett, R. J., Mitchell, M. S., & Marrs, M. B. (2009). How management style moderates the relationship between abusive supervision and workplace deviance: An uncertainty management theory perspective. Organizational Beh avior and Human Decision Processes, 108, 79 92. Thau, S., & Mitchell, M. S. ( 2006 ) How self goals moderate the relationship between abusive supervision and retaliation: A model of self defeating behaviors Unpublished manuscript. Thau, S., & Mitchell, M S. (2010). Self gain or self regulation impairment? Tests of competing explanations of the supervisor abuse and employee deviance relationship through perceptions of distributive justice. Journal of Applied Psychology, 95 1009 1031.
131 Vickers, R. R., Jr. Kolar, D. W. & Hervig, L. K. (1989). Personality correlates of coping with military basic training (Rep. No. 89 3). San Diego, CA: Naval Health Research Center. Wade S. H. (1989). The development of a scale to measure forgiveness Unpublished doctoral dissertation, Fuller Graduate School of Psychology, Pasadena, CA Watson, D., & Clark, L. A. (1994). The PANAS X: Manual for the positive and negative affect schedule expanded form Unpublished manuscript, University of Iowa. Watson, D., & Hubbard, B. (1 996). Adaptational style and dispositional structure: Coping in the context of the five factor model. Journal of Personality 64 737 774. Wayne, S. J., Hoobler, J. M., Marinova, S., & Johnson, M. M. (2008). Abusive behavior: Trickle down effects beyond t he dyad. Academy of Management Proceedings 1 6. Wayne, S., Shore, L., Bommer, W., & Tetrick, L. (2002). The role of fair treatment and rewards in perceptions of organizational support and leader member exchange. Journal of Applied Psychology 87 590 598 Weiss, H. M. & Cropanzano, R. (1996). Affective events theory: A theoretical discussion of the structure, causes and consequences of affective experiences at work. In B.M. Staw and L.L. Cummings (Eds.), Research in Organizational Behavior (Vol. 19, pp. 1 74). Greenwich, CT: JAI Press. Winter, K. A., & Kuiper, N. A. (1997). Individual differences in the experience of emotions. Clinical Psychology Review, 17 791 821. Wolfgang, M. E. (1967). Victim precipitated criminal homicide. In M. E. Wolfgang (Ed.), Studies in homicide (p. 72 87). New York: Harper & Row. Wu, T., & Hu, C. (2009). Abusive supervision and employee emotional exhaustion: Dispositional antecedents and boundaries. Group & Organization Management, 34, 143 169. Yagil, D. (2006). The relatio nship of abusive and supportive workplace supervision to employee burnout and upward influence tactics. Journal of Emotional Abuse, 6 49 65. Yagil, D., Ben Zur, H., & Tamir, I. (2011). Do employees cope effectively with abusive supervision at work?: An e xploratory study. International Journal of Stress Management, 18 5 23.
132 Zapf D Knorz C, Kulla M. (1996). On the relationship between mobbing factors, and job content, social work environment, and health outcomes. European Journal of Work and Organiza tional. Psychology, 5, 215 37. Zellars, K. L., Tepper, B. J., & Duffy, M. K. (2002). Abusive supervision and Journal of Applied Psychology, 87 1068 1076.
133 BIOGRAPHICAL SKETCH Originally from Fort Laude rdale, Florida, Lauren graduated summa cum laude with a Bachelor of Science degree in psychology from the University of Florida and cum laude with a Bachelor of Science degree in business administration also from the University of Florida. There, s he cont inued her stud ies at the graduate level and in the spring of 2011, earned her doctorate degree in business administration with a concentration in organizational behavio r and human resource management Lauren has accepted a position as Assistant Professor at Portland State University.