Psychological and technological predictors of the physical activity intention-behavior gap: an explainable machine...

Area of Science:

• Behavioral Science
• Health Psychology
• Technology Acceptance

Background:

• Physical activity (PA) intentions often fail to translate into actual behavior, creating a significant intention-behavior (I-B) gap.
• Understanding the moderators of this gap is crucial for developing effective interventions.
• Existing research often uses dichotomous measures, limiting a nuanced understanding of the I-B gap's magnitude.

Purpose of the Study:

• To integrate psychological and technological frameworks to examine mechanisms moderating the physical activity intention-behavior gap.
• To develop a standardized, continuous measure of the PA I-B gap using baseline correction.
• To identify key predictors and their non-linear interactions influencing the PA I-B gap.

Main Methods:

• Employed a machine learning framework (XGBoost) on survey data from 1,334 Chinese adults.
• Combined the Health Belief Model and the Unified Theory of Acceptance and Use of Technology.
• Utilized baseline correction for a continuous measure of the PA I-B gap.

Main Results:

• The XGBoost model (R²=0.647) outperformed traditional regression, identifying perceived barriers, self-efficacy, intention to use smart tools, and social support as core predictors.
• Higher perceived barriers and late-night activity widened the gap; greater self-efficacy, perceived benefits, intention to use smart tools, social support, social influence, and innovation narrowed it.
• Psychological cognition was a stronger predictor than smart sports tools, which acted as auxiliary resources.

Conclusions:

• Psychological cognition and smart tools jointly influence the PA I-B gap.
• Self-efficacy and perceived barriers are critical psychological factors.
• Future research should utilize longitudinal designs and objective data for causal inference.