Dyadic Invariance of the Adult Self-Report and Adult Behavior Checklist: Evidence From Young Adult Romantic Couples

The current paper investigated the invariance of the behavior problem scales of two measures from the Achenbach System of Empirically Based Assessment, the Adult Self Report (ASR) and the Adult Behavior Checklist (ABCL), across members of young adult romantic couples. Data were derived from three studies examining romantic relationships in young adulthood (total N = 672). Using a series of confirmatory factor analysis models, we tested levels of invariance of the behavior problem subscales of the ASR/ABCL across self-and partner-reports. All behavior problem subscales demonstrated invariant factor structures across self-and partner-reports. However, only the attention problems, aggressive behavior, and rule-breaking subscales achieved full metric invariance for both men and women. The intrusiveness (for men) and somatic complaints (for women) subscales also met full metric invariance. The remaining subscales only met partial metric invariance for both men (withdrawn, anxious/depressed, somatic complaints) and women (withdrawn, anxious/depressed, intrusiveness). All subscales only met partial scalar invariance. The current study extends the literature on cross-informant assessments by moving beyond studies of mean differences or correlations between informant reports. Moreover, our findings indicate that understanding agreement between informants should include multiple levels of measurement, such as factor structure, factor loadings, and item means. We recommend careful consideration of these issues when using the ASR/ABCL in a cross-informant framework. Public Significance Statement This study extends the psychometric properties of the behavior problem scales from the Adult Self Report (ASR) and Adult Behavior Checklist (ABCL) in a sample of young adult couples. Our findings indicate attention to multiple levels of measurement (error) is essential when using the ASR/ABCL in a cross-informant framework.