We found that overall the consensus manipulation did not influence participants judgments (P<.01.F(2,215-41.p-66 Deontological 2.57, sd 1.32: Utilitarian 2.75, sd 1.60: M 2.75. sde M Control 1.28). How, if at all, did participants' responses to the consensus manipulation vary according to their attachment style? To conduct a regression analysis, we created dummy variables for the different experimental conditions. The control group was coded as the reference group. Next, attachment anxiety and attachment avoidance scores were mean centered and all interaction terms were created (West, Aiken. & Krull. 1996). In the first step of a hierarchical regression analysis, we regressed