Beyond Scale Correspondence: A Comparison of Continuous Open Scaling and Fixed Graded Scaling When Using Social Cognitive Constructs in the Exercise Domain

Abstract

The standard response format for self-reported exercise-behavior measurement is the continuous open scale, but popular social cognitive theories use fixed graded scales, a noncorrespondent format. Benefits of using continuous open scales for social cognitive constructs include scale correspondence with the behavior measure, the potential of increased variability from an open response format, and potentially more accurate measurement from using a ratio-level scale. The purpose of this study was to create continuous open measures of theory of planned behavior (TPB) constructs and compare these to standard fixed graded measures. Participants were 420 undergraduate students who completed measures of fixed graded and continuous open TPB scales and a 2-week follow-up of exercise behavior. Psychometric properties were acceptable for both types of scaling, but the correlations of TPB constructs with exercise intention and behavior were larger for continuous open scaling than fixed graded scaling. This result was due to the additional variability afforded by continuous open scaling and not from scale correspondence; further, the 2 types of scales had convergent measurement domains. Finally, in an applied analysis, regression equations derived from the continuous open measures showed that higher levels of social cognition toward a behavioral frequency than toward the desired behavioral outcome are needed to produce that desired behavioral outcome. Overall, continuous open scaling may be a superior form of scaling for social cognitive constructs in the exercise domain in comparison to standard fixed graded measures.