Individual and situational antecedents of workplace victimization

Purpose - The purpose of this paper is to simultaneously examine individual‐ and team‐level predictors of workplace victimization by applying two‐level modeling. Previous workplace victimization research has primarily been conducted on the individual level of analysis, which may be insufficient when assessing organizational‐ and team‐level predictors of workplace victimization. The authors examined the relationships between target personality (Big Five personality factors), perceived stressors, work climate, and perceived workplace victimization. Design/methodology/approach - The research was conducted in organizations within the restaurant sector; the data were obtained using questionnaires completed by employees and supervisors (n=207) in 70 restaurants. Findings - A model that included all variables fitted the data well. However, individual‐level perceived stressors was the only variable that was significantly related to workplace victimization. The facts that our model included team‐level climate factors and individual‐level personality traits, and both individual‐ and team‐level victimization showed good fit to the data, with only one specific variable in the model showing any significant relationship with bullying, may indicate that bullying is more a consequence of the total risk factors than related to specific factors, on an individual or team level. Hence, it is the additive effect of the factors that matters more than each individual factor. Practical implications - Of practical relevance is that the paper provides evidence that both individual and team‐level factors are related to workplace victimization, indicating that bullying cannot be prevented by focusing on clearly defined risk factors either on team or individual levels. As a manager, one must work on all aspects of the social working environment, including the total vulnerability of one's employees and their respective levels of work stress. Originality/value - From a methodological viewpoint, the paper demonstrates that an analysis of clustered individual‐level data, without the application of proper multilevel analysis, may lead to biased results.