Sherry E. Moss

Predictors of Abusive Supervision: Supervisor Perceptions of Deep-Level Dissimilarity, Relationship Conflict, and Subordinate Performance

The moral exclusion literature identifies three previously unexamined predictors of abusive supervision: supervisor perceptions of deep-level dissimilarity, relationship conflict, and subordinate performance. Invoking theory and research on workplace diversity, relationship conflict, and victim precipitation, we model the three predictors as associated with abusive supervision. Path-analytic tests using data collected from supervisor-subordinate dyads at two time points suggest that supervisor perceptions of relationship conflict and subordinate performance mediate the relationship between perceived deep-level dissimilarity and abusive supervision and that relationship conflict mediates that between perceived deep-level dissimilarity and abusive supervision when supervisors perceive subordinates as having low performance. Estimates suggest that more than 13 percent of working people in the United States become targets of abusive supervision, or nonphysical hostility perpetrated by employees’ immediate superiors (Schat, Frone, & Kelloway, 2006). Examples of behaviors that fall within the abusive supervision content domain include undermining, public denigration, and explosive outbursts (Tepper, 2007). Sustained exposure to abusive supervision is associated with serious negative outcomes for victims and employers, including psychological distress (Tepper, 2000), problem drinking (Bamberger & Bacharach, 2006), and aggression directed against a

The Mediating Role of Feedback Avoidance Behavior in the LMX—Performance Relationship:

The literature on leader—member exchange theory (LMX) has consistently demonstrated the positive relationship between member perceptions of the quality of their relationship with the leader and member performance. The process through which relationship quality influences member performance, however, is still not fully understood.The present study provides an explanatory mechanism for this process. Specifically, feedback avoiding behavior, a feedback management strategy used by poor performers to minimize exposure to negative feedback from their leaders, fully mediates the relationship between LMX and member performance. Implications for theory and practice are discussed.

When and How Subordinate Performance Leads to Abusive Supervision: A Social Dominance Perspective:

While we would typically expect poor performers to elicit abusive responses from their supervisors, we theorize that high performers may also be victims of abusive supervision. Specifically, we draw on social dominance theory to hypothesize and demonstrate that subordinate performance can have a positive, indirect effect on abusive supervision through the mediator of perceived threat to hierarchy. And this positive indirect effect prevails when the supervisor’s social dominance orientation is high. We found support for our theoretical model using data collected from supervisor–subordinate dyads.

From Synchronizing to Harmonizing: The Process of Authenticating Multiple Work Identities:

To understand how people cultivate and sustain authenticity in multiple, often shifting, work roles, we analyze qualitative data gathered over five years from a sample of 48 plural careerists—people who choose to simultaneously hold and identify with multiple jobs. We find that people with multiple work identities struggle with being, feeling, and seeming authentic both to their contextualized work roles and to their broader work selves. Further, practices developed to cope with these struggles change over time, suggesting a two-phase emergent process of authentication in which people first synchronize their individual work role identities and then progress toward harmonizing a more general work self. This study challenges the notion that consistency is the core of authenticity, demonstrating that for people with multiple valued identities, authenticity is not about being true to one identity across time and contexts, but instead involves creating and holding cognitive and social space for several true versions of oneself that may change over time. It suggests that authentication is the emergent, socially constructed process of both determining who one is and helping others see who one is.

An exploration of the role of subordinate affect in leader evaluations.

Leadership research has been encumbered by a proliferation of constructs and measures, despite little evidence that each is sufficiently conceptually and operationally distinct from the others. We draw from research on subordinates’ implicit theories of leader behavior, behaviorally anchored rating scales, and decision making to argue that leader affect (i.e., the degree to which subordinates have positive and negative feelings about their supervisors) underlies the common variance shared by many leadership measures. To explore this possibility, we developed and validated measures of positive and negative leader affect (i.e., the Leader Affect Questionnaires; LAQs). We conducted 10 studies to develop the five-item positive and negative LAQs and to examine their convergent, discriminant, predictive, and criterion-related validity. We conclude that a) the LAQs provide highly reliable and valid tools for assessing subordinates’ evaluations of their leaders; b) there is significant overlap between existing leadership measures, and a large proportion of this overlap is a function of the affect captured by the LAQs; c) when the LAQs are used as control variables, in most cases, they reduce the strength of relationships between leadership measures and other variables; d) the LAQs account for significant variance in outcomes beyond that explained by other leadership measures; and e) there is a considerable amount of unexplained variance between leadership measures that the LAQs do not capture. Research suggestions are provided and the implications of our results are discussed.