Validity of social-cognitive measures for physical activity in middle-school girls.

OBJECTIVE The factorial validity and measurement equivalence/invariance of scales used to measure social-cognitive correlates of physical activity among adolescent girls were examined. METHODS Confirmatory factor analysis was applied to questionnaire responses obtained from a multi-ethnic sample (N = 4885) of middle-school girls from six regions of the United States. A cohort of 1893 girls completed the scales in both sixth and eighth grades, allowing longitudinal analysis. RESULTS Theoretically and statistically sound models were developed for each scale, supporting the factorial validity of the scales in all groups. Multi-group and longitudinal invariance was confirmed across race/ethnicity groups, age within grade, BMI categories, and the 2-year period between grades. CONCLUSIONS The scores from the scales provide valid assessments of social-cognitive variables that are putative mediators or moderators of change in physical activity. The revised scales can be used in observational studies of change or interventions designed to increase physical activity among girls during early adolescence.