Validating the couple relationship skills inventory

Objective We tested the validity of the factor structure and reliability of a new research-informed comprehensive inventory of key relationship skills predictive of couple quality, the Couple Relationship Skills Inventory (CRSI). Background The CRSI is based on the National Extension Relationship and Marriage Education Model, an evidence-derived framework developed as a guide for couple relationship education content. For internal consistency in assessing the effectiveness of programming for couples and for general use in practice and research with couples, an important next step is the design and validation of a comprehensive measure of these core behavioral/attitudinal skills. Method The analytic (or "training") sample of ethnically and economically diverse adults included 824 (independent) men and women and two cross-validation samples (n = 763 and n = 470). Results Bayesian confirmatory factor analysis of individual measurement models informed refinement to a 32-item, nine-factor (seven-subscale) measure and indicated excellent fit of the model to the data. Reliabilities for the full scale and the subscales were good to excellent. Cross-validation study fit statistics and reliabilities were similar, and measurement invariance across samples was validated. Further, support for internal discriminant validity was implied by small to moderate covariances among the factors and concurrent and predictive validity was evidenced (i.e., significant associations among CRSI scores and measures of relationship quality and family harmony). Implications This measure provides an efficient assessment of core relational skills critical for healthy couple quality and may prove useful in practice and for future studies of couple relationships and couple relationship education.