Measurement characteristics of activity-related psychosocial measures in 8- to 10-year-old African-American girls in the Girls health Enrichment Multisite Study (GEMS)

Background. This paper presents reliability and validity analyses of physical activity-related psychosocial questionnaires completed by 8-to 10-year-old African-American girls at baseline and follow-up assessments of pilot intervention studies in the Girls health Enrichment Multi-site Study (GEMS). Methods. Two hundred ten girls participated in the GEMS 12-week pilot studies and had their height and weight measured, wore an accelerometer for 3 days and completed a measure of their usual physical activity (PA) at baseline and after the 12-week intervention. Subgroups of girls also completed physical activity-related psychosocial measures at these two time points including: (a) self-concept; (b) self-efficacy; (c) outcome expectancies; and d) preferences. Principal components analysis was conducted on the psychosocial measures obtained at baseline. Cronbach's alpha and test-retest reliability were computed. Convergent validity was assessed by correlating the baseline psychosocial measures with baseline physical activity measures and body mass index (BMI). Results. The following sub-scales were derived: Activity Preference, Positive Expectancies and Negative Expectancies for physical activity. Physical Performance Self-Concept and Self-Efficacy for physical activity were kept as single dimensional scales. Sub-scales, derived from principal components analyses, were Activity Preference, Positive Expectancies, and Negative Expectancies for physical activity. Internal consistency estimates for the various scales were substantial to excellent (0.67-0.85), while test-retest reliability estimates were fair to moderate (0.22-0.56). Correlations between the PA psychosocial sub-scales and measured levels of activity measures showed evidence of convergent validity for the Activity Preference sub-scale, although social desirability may have influenced the significant associations observed. Conclusions. The Activity Preference was a fairly reliable and valid measure. Further studies are needed to examine the utility of activity related psychosocial measures in interventions to increase physical activity among preadolescent African-American girls. (C) 2004 The Institute For Cancer Prevention and Elsevier Inc. All rights reserved.