Can current physical activity act as a reasonable proxy measure of future physical activity? Evaluating cross-sectional and passive prospective designs with the use of social cognition models

Abstract

Background The standard methodological approach for evaluating social cognitive theories when predicting physical activity behavior is the passive prospective/longitudinal survey design. Although this design is logical, a cross-sectional design may be a cost-effective alternative if the relationships between social cognitive constructs and physical activity are relatively stable. The purpose of this study was to evaluate the utility of a concurrent measure of physical activity used in a cross-sectional design in comparison to the standard prospective measure. Methods This study included two 6-month prediction time-periods, between 1997 and 1998, for the purpose of analysis replication, and the theory of planned behavior, the transtheoretical model, protection motivation theory, and social cognitive theory as the models of interest in a population sample ( N = 703). Results Results showed trivial (69% of tests; q < 0.10) to small (31% of tests; q = 0.11–0.18) differences in the correlations between social cognitive constructs and vigorous physical activity occur when using a cross-sectional or prospective design. The cross-sectional design estimated slightly larger coefficients than the prospective design. Conclusions It appears that a measure of concurrent physical activity included in a cross-sectional design can act as a reasonable proxy measure of future behavior measured in a passive prospective/longitudinal design. These findings support the use of cross-sectional designs when researchers seek a standard correlational investigation of physical activity and social cognitive constructs with the possibility that coefficients may be slightly biased upwards.