Angry but not Deviant: Employees’ Prior-Day Deviant Behavior Toward the Family Buffers Their Reactions to Abusive Supervisory Behavior

We controlled for the effect of daily abusive supervisory behavior on deviant behavior toward the family in our research.

We conducted multi-group invariance analyses to determine whether the deviant behavior toward the family measure was invariant across groups with (a) different spousal status (married/partnered vs. not) and (b) child status (having a child or not). Specifically, we examined two forms of invariance: (a) configural invariance, which pertains to the same factor structure being measured in different groups, and (b) metric invariance, which assesses whether the measures are calibrated to the construct in the same way in different groups (Chan 1998; Ployhart and Vandenberg 2010; Vandenberg and Lance 2000; Vandenberg and Morelli 2016). We examined whether change of chi-square between the configural model and the metric model was significant, with a non-significant chi-square difference indicating metric invariance (Vandenberg and Lance 2000). We used deviant behavior toward the family from the first survey for the multi-group invariance analyses. First, we examined whether the deviant behavior toward the family measure was invariant as a function of spousal status. The difference between the configural model (χ² = 38.38, df = 10) and the metric model (χ² = 45.68, df = 14) was not statistically significant (Δχ² = 7.30, Δdf = 4, p = .12). Second, we examined whether the deviant behavior toward the family measure was invariant as a function of child status. The difference between the configural model (χ² = 29.38, df = 10) and the metric model (χ² = 37.37, df = 14) was also not statistically significant (Δχ² = 7.99, Δdf = 4, p = .09).

We thank one anonymous reviewer for suggesting this control variable.

We thank one anonymous reviewer for suggesting this analysis.