Jennifer L Gay1, Jessalyn Smith. 1. University of Texas School of Public Health, Brownsville, TX 78520, USA. Jennifer.L.Gay@uth.tmc.edu
Abstract
OBJECTIVES: To examine the construct validity and reliability of existing self-report items measuring perceptions of the built environment for physical activity. METHODS: Exploratory factor analysis, internal consistency, and nonparametric scaling were assessed using data from a cross-sectional sample of adults (n = 479). RESULTS: A new 5-factor model was identified. Internal consistency and scaling reliabilities were moderate to strong, providing support for the new structure. CONCLUSIONS: Further studies are needed to confirm these results. Subscales should be treated as indices and use sum scores in analysis to detect subtle variations and provide more precise power calculations.
OBJECTIVES: To examine the construct validity and reliability of existing self-report items measuring perceptions of the built environment for physical activity. METHODS: Exploratory factor analysis, internal consistency, and nonparametric scaling were assessed using data from a cross-sectional sample of adults (n = 479). RESULTS: A new 5-factor model was identified. Internal consistency and scaling reliabilities were moderate to strong, providing support for the new structure. CONCLUSIONS: Further studies are needed to confirm these results. Subscales should be treated as indices and use sum scores in analysis to detect subtle variations and provide more precise power calculations.