almost significantp = .083
Logistic regression analyses using these dummy coded variables yielded an almost significant interaction effect between violation type and the dummy that contrasted equal compensation with no compensation (Wald = 3.01, p = .083), a significant interaction effect between violation type and the dummy contrasting overcompensation to no compensation (Wald = 5.46, p = .019), and a non-significant interaction effect between violation type and the dummy that contrasted overcompensation with equal compensation (Wald = 0.65, p = .421).