Measurement Invariance Across Sexes in Intimate Partner Abuse Research

Objective: To examine the assumption of measurement invariance across sexes from one representative instrument that uses subscale scores to represent the latent construct intimate partner abuse (IPA). Method: Extant data were used to investigate measurement invariance across sexes with one screening measure of IPA, the Relationship Behavior Rating Scale. This particular measure was used in the current study for the purpose of illustrating the need to investigate measurement equivalence as a standard procedure for all instruments measuring IPA when comparisons across sexes are made. IPA was measured in a sample of 1,654 divorcing individuals mandated to mediation for custody and parenting-time disputes. Results: Although a multiple-group confirmatory factor analysis provided evidence that the same 5-factor latent model of IPA (coercive controlling behaviors, psychological abuse, physical abuse, threats and escalated violence, and sexual abuse) is measured in men and women, the assumption that an analogous pattern of observed variables contribute to each of the 5 subcategories of IPA was not supported. Exploratory analyses suggest that items measuring threatened and escalated violence and sexual abuse may be a source of such variance across sexes. Conclusions: Measurement invariance is an important step in advancing the empirical investigation of IPA. Future research should explore sources of variance to advance the understanding of similarities and differences between sexes in reporting of IPA.