Abstract
Aims: This study compared the psychometric properties of two self-efficacy instruments related to physical activity. Factorial validity, cross-group and longitudinal invariance, and composite reliability were examined.Methods: Secondary analysis was conducted on data from a group randomized controlled trial investigating the effect of a 17-week intervention on increasing moderate to vigorous physical activity among 5th–8th grade girls (N = 1,012). Participants completed a 6-item Physical Activity Self-Efficacy Scale (PASE) and a 7-item Self-Efficacy for Exercise Behaviors Scale (SEEB) at baseline and post-intervention. Confirmatory factor analyses for intervention and control groups were conducted with Mplus Version 7.4 using robust weighted least squares estimation. Model fit was evaluated with the chi-square index, comparative fit index, and root mean square error of approximation. Composite reliability for latent factors with ordinal indicators was computed from Mplus output using SAS 9.3.Results: Mean age of the girls was 12.2 years (SD = 0.96). One-third of the girls were obese. Girls represented a diverse sample with over 50% indicating black race and an additional 19% identifying as mixed or other race. Both instruments demonstrated configural invariance for simultaneous analysis of cross-group and longitudinal invariance based on alternative fit indices. However, simultaneous metric invariance was not met for the PASE or the SEEB instruments. Partial metric invariance for the simultaneous analysis was achieved for the PASE with one factor loading identified as non-invariant. Partial metric invariance was not met for the SEEB. Longitudinal scalar invariance was achieved for both instruments in the control group but not the intervention group. Composite reliability for the PASE ranged from 0.772 to 0.842. Reliability for the SEEB ranged from 0.719 to 0.800 indicating higher reliability for the PASE. Reliability was more stable over time in the control group for both instruments.Conclusions: Results suggest that the intervention influenced how girls responded to indicator items. Neither of the instruments achieved simultaneous metric invariance making it difficult to assess mean differences in PA self-efficacy between groups.
Highlights
Despite the benefits of physical activity (PA), less than 25% of adolescents meet recommended guidelines (Fakhouri et al, 2014; Kann et al, 2016) calling for 60 min or more per day of at least moderate-intensity PA (U.S Department of Health and Human Services, 2008)
This study aims to fill this gap by comparing the factorial validity, measurement invariance, and reliability of two PA self-efficacy instruments used in a large-scale study to test a PA intervention with urban, adolescent girls
We ran item-level analyses for each indicator of the Physical Activity Self-Efficacy Scale (PASE) and Self-Efficacy for Exercise Behaviors (SEEB) instruments at baseline and post-intervention for the full sample and separately in intervention and control groups
Summary
While many psychosocial factors have been theorized to increase PA among adolescents, self-efficacy is an important correlate and determinant of PA (Craggs et al, 2011; Bauman et al, 2012) and mediator of PA intervention effects (Lubans et al, 2008). Some researchers have reported contradictory findings regarding the relationship between self-efficacy and adolescent PA suggesting that inadequate and varied measurement of the concept may explain the inconsistencies (Dewar et al, 2013; Plotnikoff et al, 2013). This study aims to fill this gap by comparing the factorial validity, measurement invariance, and reliability of two PA self-efficacy instruments used in a large-scale study to test a PA intervention with urban, adolescent girls. The results may contribute to better understanding of PA self-efficacy and its role in fostering PA among girls
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
