Factorial Validity and Invariance of Questionnaires Measuring Social-Cognitive Determinants of Physical Activity among Adolescent Girls

Background. There are few theoretically derived questionnaires of physical activity determinants among youth, and the existing questionnaires have not been subjected to tests of factorial validity and invariance. The present study employed confirmatory factor analysis (CFA) to test the factorial validity and invariance of questionnaires designed to be unidimensional measures of attitudes, subjective norms, perceived behavioral control, and self-efficacy about physical activity. Methods. Adolescent girls in eighth grade from two cohorts (N = 955 and 1,797) completed the questionnaires at baseline; participants from cohort 1 (N = 845) also completed the questionnaires in ninth grade (i.e., 1-year follow-up). Factorial validity and invariance were tested using CFA with full-information maximum likelihood estimation in AMOS 4.0. Initially, baseline data from cohort 1 were employed to test the fit and, when necessary, to modify the unidimensional models. The models were cross-validated using a multigroup analysis of factorial invariance on baseline data from cohorts 1 and 2. The models then were subjected to a longitudinal analysis of factorial invariance using baseline and follow-up data from cohort 1. Results. The CFAs supported the fit of unidimensional models to the four questionnaires, and the models were cross-validated, as indicated by evidence of multigroup factorial invariance. The models also possessed evidence of longitudinal factorial invariance. Conclusions. Evidence was provided for the factorial validity and the invariance of the questionnaires designed to be unidimensional measures of attitudes, subjective norms, perceived behavioral control, and self-efficacy about physical activity among adolescent girls.