Determinants of artificial intelligence self-efficacy in physiotherapists: The role of attitudes, AI anxiety and sociodemographic factors

Abstract

Determinants of artificial intelligence self-efficacy in physiotherapists: The role of attitudes, AI anxiety and sociodemographic factorsABSTRACTPurpose: This study examined the role of attitudes toward artificial intelligence (AI), AI-related anxiety, and sociodemographic factors in physiotherapists’ AI self-efficacy.Material and Methods: In this cross-sectional online survey, 205 physiotherapists completed the Artificial Intelligence Self-Efficacy Scale, the General Attitudes toward Artificial Intelligence Scale, and the Artificial Intelligence Anxiety Scale. Sociodemographic variables (gender, age, education level, professional experience, employment sector, academic position, prior AI training) were recorded. Multiple linear regression was used to identify factors associated with self-efficacy.Results: The multiple regression model showed that positive attitudes toward AI were the strongest predictor of AI self-efficacy (β = .637, p < .001). Negative attitudes were also significantly positively associated with self-efficacy (β = .145, p = .006). AI-related anxiety demonstrated a negative but statistically non-significant association with self-efficacy (p = .07). Socio-demographic variables did not significantly affect self-efficacy (p > .05). Overall, the model explained 52.6% of the variance in AI self-efficacy (R² = .526, adjusted R² = .489).Conclusion: AI self-efficacy in physiotherapists appears to be shaped mainly by attitudinal rather than sociodemographic factors. Education and professional development should therefore focus on strengthening positive and critically reflective attitudes toward AI to support its integration into clinical practice.Keywords: Anxiety, Artificial intelligence, Attitudes, Physiotherapy, Self-efficacy.