Internal consistency and validity of measures of automatic exercise associations

Abstract

Researchers increasingly recognize that exercise and physical activity may be determined by interacting explicit and implicit processes. While this trend has led to a substantial increase in the use of measures of implicit processes within exercise psychology, such measures are typically selected without a supporting rationale. To facilitate the refinement of theoretical models and the testing of interventions, investigating the internal consistency and validity of measures of implicit processes, such as automatic exercise associations, is crucial. Objectives: To assess the internal consistency and validity of nine measures of automatic exercise associations. Method: Participants (N = 95) completed an exercise session at the intensity of the ventilatory threshold, intended to generate heterogeneous affective responses. One week later, they also completed nine randomly ordered measures of automatic exercise associations. The slope of affect ratings during exercise, affect ratings at the end of exercise, recalled affect, explicit affective attitude, self-reported exercise, and situated decisions to exercise served as validation criteria. Results: Three of the nine measures exhibited acceptable to good internal consistency. Only the Approach-Avoidance Task was significantly and meaningfully related to any of the validation criteria (i.e., self-reported exercise and situated decisions to exercise). Conclusions: Most measures of automatic exercise associations exhibited unsatisfactory internal consistency and were unrelated to the validity criteria. For research on the role of implicit processes in exercise and physical activity to advance, further psychometric evaluation and refinement of measures is needed.