Measurement of Physical Activity Self-Efficacy in Adults With Obesity: A Latent Variable Approach to Explore Dimensionality, Temporal Invariance, and External Validity.

Abstract

The objective of this study was to improve the measurement of physical activity self-efficacy (PASE) in adults with obesity. To accomplish this objective, a latent variable approach was used to explore dimensionality, temporal invariance, and external validity of responses to a newly developed battery of PASE scales. Data (Nbaseline = 461 and N30 days postbaseline = 427) from the Well-Being and Physical Activity Study (ClinicalTrials.gov, identifier: NCT03194854), which deployed the Fun For Wellness intervention, were analyzed. A two-dimensional factor structure explained responses to each PASE scale at baseline. There was strong evidence for at least partial temporal measurement invariance for this two-dimensional structure in each PASE scale. There was mixed evidence that the effectiveness of the Fun For Wellness intervention exerted a direct effect on latent PASE in adults with obesity at 30 days postbaseline (i.e., external validity) of this two-dimensional structure.