A review of the outcome expectancy construct in physical activity research

Background: Outcome expectancy is a central construct in social cognitive models of health behavior widely used as frameworks for physical activity research.Purpose: This article provides a review of the outcome expectancy construct and its application to research on physical activity.Methods: Theoretical articles describing definitions and placement of outcome expectancy within social cognitive models, as well as empirical research on outcome expectancy and physical activity, were reviewed.Results: Self-efficacy theory, the transtheoretical model, the theory of planned behavior, and protection motivation theory differ in their labeling and conceptualization of outcome expectancy but unanimously include expected outcomes of behavior. Preliminary empirical investigation of the role of outcome expectancy in understanding physical activity has yielded mixed results. Positive outcome expectancy appears to be more predictive of physical activity in older adults than in young to middle-aged adults, and personal barriers appear to be the most predictive subtype of negative outcome expectancy. In addition, a small number of studies indicate relations between outcome expectancy and other theoretical variables, including behavioral intention, stage of change, and self-efficacy.Conclusions: Further research on the role of outcome expectancy is necessary to design effective physical activity interventions. New directions in outcome expectancy research could involve (a) expanding the conceptualization of outcome expectancy to include expected outcomes of sedentary behavior and affective responses to physical activity, (b) further examination of potential moderators of the relation between outcome expectancy and physical activity (such as outcome value and outcome proximity), (c) distinguishing between the role of outcome expectancy in behavior onset versus behavior maintenance, (d) examining outcome expectancy as a mechanism of change in environmental intervention approaches, and (e) further analysis of interrelations between outcome expectancy and other social cognitive variables. (Ann Behav Med 2005,29(1):70-79) Support for this article was partly provided by Grant R01CA79469 from the National Cancer Institute to Richard A. Winett, principal investigator, at Virginia Tech. This research was conducted to complete requirements for David M. Williams's preliminary examination at Virginia Polytechnic Institute and State University. Richard Winett and Eileen Anderson served as co-chairs on the committee. Special thanks to other committee members Lee Cooper, Robert Stephens, and Janet Wojcik.