Examining the measurement equivalence of the Conditional Reasoning Test for Aggression across U.S. and Croatian samples

The Conditional Reasoning Test for Aggression (CRT-A; James et al., 2005) is based on the ideas that aggressive individuals use motive-based cognitive biases to see their behavior as reasonable and that those biases can be measured with specially designed inductive reasoning tasks. The test has shown promising psychometric characteristics for U. S. samples but has not been validated in other cultural contexts. In our study, we examined whether the items from the CRT-A were invariant across culture by testing whether these items displayed differential item functioning (DIF) across Croatian (N=530) and U.S. (N=1479) samples. The Lord's Chi Square (Lord, 1980), the Raju UA index (Raju, 1988), the Mantel-Haenszel procedure (Mantel & Haenszel, 1959), and the logistic regression procedures (Swaminathan & Rogers, 1990) revealed that DIF was pervasive. Although an implicit measure of personality, the CRT-A seems susceptible to differential item functioning in another culture