We also created an unconstrained m in which we included paths between anxious attachment and discomfort with caregiving, anxious attachment and victim empathy, avoidant attachment and group empathy, and a direct path between anxious attachment and wrongness ratings (see Fig. 2). This model would suggest that anxiously and avoidantly attached individuals reacted similarly to the dilemmas. As depicted in the figure, these added paths were not significant, suggesting that anxiously and avoidantly attached individuals reached utilitarian judgments for different reasons. Although the unconstrained model fit the data well, RMSEA-0.02,90% CI [0.0, 0.06], .996 a Likelihood Ratio test of the two models suggested that the fit of the un- constrained model was not superior to that of the constrained model, x2(4) 5.08. p 0.28. Thus, the constrained model (with the additional, non-significant paths removed) is preferred for its parsimony.