Claxton, S. E., DeLuca, H. K., & van Dulmen, M. H. M.
Testing psychometric properties in dyadic data using confirmatory factor analysis: Current practices and recommendations
The current paper examines the challenges — and provides recommendations — related to testing the psychometric properties of measures using dyadic data (distinguishable cases). The paper provides a review of the procedures that researchers have been using when performing confirmatory factor analyses (CFAs) on data that is dyadic in nature. In addition to reviewing common methods currently used, we provide examples of how different methods of testing measurement invariance with dyadic data affect results. Specifically, we utilize three dyadic samples (containing 336 couples, N = 672) to examine and cross-validate techniques across two measures: the Adult Self Report (Achenbach & Rescorla, 2003) and the Relationship Assessment Scale (Hendrick, 1988). Finally, we present specific recommendations for examining measurement invariance using a CFA framework with dyadic data.
Testing, Psychometrics, Methodology in Applied Psychology, 2015, Vol. 22, No. 2, pp. 199-217, DOI: 10.4473/TPM22.2.2Back