You don't know why you (don't) exercise: The relationship between automatic processes and physical activity (or sedentary behavior): A meta-analysis.

Objectives: This meta-analysis aimed to examine the relationship between automatic processes and behavior as well as their moderators.

Methods: There were 126 effect sizes (Fisher's Z) extracted from 55 independent research studies involving 10,432 participants. Meta-analyses were conducted using Comprehensive Meta-Analysis Version 3.3. Univariate meta-regression models were used to analyze underlying moderators.

Results: Meta-analysis after correction found that automatic processes could significantly and positively predict behavior, but the effect size was small (Fisher's Z = 0.057, p < 0.001, 95% CI [0.043, 0.070]).

Conclusions: Different components of automatic processes, task reliability, and study design moderated the relationship between automatic processes and behavior. Future research could explore automatic measures at a neurophysiological level or use population-matched stimuli for specific populations to improve measure validity. In addition, future research should accumulate evidence on how to reduce sedentary behavior.