An experimental test of control theory-based interventions for physical activity.

Objectives: To provide an experimental test of control theory to promote physical activity. Design: Parallel groups, simple randomized design with an equal chance of allocation to any group. Methods: Participants not meeting recommended levels of physical activity but physically safe to do so (N = 124) were recruited on a UK university campus and randomized to goal-setting + self-monitoring + feedback (GS + SM + F, n = 40), goal-setting + self-monitoring (GS + SM, n = 40), or goal-setting only (GS, n = 44) conditions that differentially tapped the key features of control theory. Accelerometers assessed physical activity (primary outcome) as well as self-report over a 7-day period directly before/after the start of the intervention. Results: The participants in the GS + SM + F condition significantly outperformed those in the GS condition, d = 0.62, 95% CI d = 0.15–1.08, and marginally outperformed those in the GS + SM condition in terms of total physical activity at follow-up on the accelerometer measure, d = 0.33, 95% CI d = −0.13 to 0.78. The feedback manipulation (GS + SM + F vs. GS + SM and GS) was most effective when baseline intentions were weak. These patterns did not emerge on the self-report measure but, on the basis of this measure, the feedback manipulation increased the risk that participants coasted in relation to their goal in the first few days of the intervention period. Conclusions: Using behaviour change techniques consistent with control theory can lead to significant short-term improvements on objectively assessed physical activity. Further research is needed to examine the underlying theoretical principles of the model.