Validity and reliability of measures assessing social-cognitive determinants of physical activity in low-active Australian adults

This cross-sectional study of 504 community-dwelling Australian adults (328 females, 176 males, mean age 50.8 ± 13.0 years) sought to examine the reliability and validity of measurement scales for physical activity (PA) self-efficacy and outcome expectations. Participants completed demographic and anthropometric measurements, and a 23-item psychosocial questionnaire pertinent to an intervention target of 10,000 steps per day. Exploratory (n = 252) and confirmatory (n = 252) factor analyses were conducted to determine psychometric properties of the measures. Based on theory and goodness-of-fit indices, six factors were extracted from the questionnaire: PA self-efficacy; PA barriers self-efficacy (including general, personal, and conflict); and physical and mental outcome expectations. From confirmatory factor analysis, the model demonstrated good data fit in four out of five indices: CFI = 0.99; TLI = 0.99; SRMR = 0.03; RMSEA = 0.03, 90%CI = 0.01–0.05, χ² = 113.14 (88), p = 0.04; including good fit by sex, age, weight status, education, and birth country. PA interventions can employ our psychometrically sound social cognitive measures.