VARIANCE OF LEISURE-TIME PHYSICAL ACTIVITY EXPLAINED BY THREE OPERATIONAL VARIABLES

Abstract

The health benefits of moderate and vigorous leisure-time physical activity (LTPA) are well recognized. The lack of uniformity in operationalizing physical activity creates difficulties when researching or comparing studies in the literature. PURPOSE The purpose of this study was to examine whether different determinants exist for time spent in LTPA, average intensity of all activities, and work performed during LTPA. METHODS The theory of planned behavior (TPB) was chosen to investigate the behavioral aspects of LTPA. Rating scales were used to assess perceived control (PC) and intention to engage in LTPA (Int) among 129 high school girls. Using a compendium of physical activities with MET values, we assessed LTPA with a 21-day physical activity diary. LTPA was operationalized as duration in minutes (PA), work done expressed as total METs × minutes (PA-MM), and the average intensity per minute derived from PA-MM + PA (PA-AMM). RESULTS Hierarchical multiple regression analyses indicated that Int and PC explicated 9.8%, 18.0%, and 13.1% of the total variance of PA, PA-MM, and PA-AMM, respectively. This variance was due to fluctuations in PC and the Int × PC interaction term, and not to the Int term alone. PC by itself had the greatest impact on the predictive ability of LTPA when it was operationalized as average intensity (R2 change = 6.0%), and less impact when it was expressed as duration (R2 change = 1.4%) or work (R2 change = 4.2%). CONCLUSION Different determinants of LTPA produce substantially different amounts of variance explained and, certainly, the strength of one's argument would differ based on whether 9.8% or 18% of the total variance could be explained. However, regarding the TPB, all three operational definitions of LTPA would have elicited a similar conclusion supporting the basic tenets of the TPB.