KMAN Publication Inc. (KMANPUB) Journal of Assessment and Research in Applied Counseling (JARAC) 8 Serial Number 29 2026 04 01 AI-Based Prediction of Social Anxiety: Interactions Among Intolerance of Uncertainty, Self-Focused Attention, Safety Behaviors, and Fear of Evaluation AI-Based Prediction of Social Anxiety: Interactions Among Intolerance of Uncertainty, Self-Focused Attention, Safety Behaviors, and Fear of Evaluation 1 9 10.61838/kman.jarac.5260 EN Journal Article 2025 12 13 <table> <tbody> <tr> <td> <p><strong>Objective:</strong> The present study aimed to develop and evaluate an artificial intelligence-based predictive model of social anxiety severity by examining the individual and interactive effects of intolerance of uncertainty, self-focused attention, safety behaviors, and fear of evaluation.</p> <p><strong>Methods and Materials: </strong>This cross-sectional predictive study was conducted on a sample of 412 young adults recruited from universities and community settings in Canada. Participants completed standardized self-report measures assessing social anxiety, intolerance of uncertainty, self-focused attention, safety behaviors, and fear of evaluation. Data were collected through an online survey platform and subjected to preprocessing procedures including missing data imputation, outlier management, and feature standardization. The dataset was divided into training and testing subsets, and multiple machine learning algorithms, including Random Forest, Support Vector Machine, and Extreme Gradient Boosting (XGBoost), were implemented to predict social anxiety severity. Model performance was evaluated using R², root mean square error (RMSE), and mean absolute error (MAE), while SHAP (Shapley Additive Explanations) analysis was employed to interpret feature importance and interaction effects among predictors.</p> <p><strong>Findings: </strong>The results indicated that all predictors were significantly associated with social anxiety, with fear of evaluation emerging as the strongest predictor, followed by safety behaviors, intolerance of uncertainty, and self-focused attention. Among the tested models, XGBoost demonstrated the highest predictive accuracy (R² = 0.74), outperforming Random Forest and Support Vector Machine. Interaction analyses revealed that fear of evaluation combined with safety behaviors, as well as intolerance of uncertainty combined with self-focused attention, exerted the strongest synergistic effects on social anxiety severity, indicating the presence of nonlinear and interaction-based relationships among variables.</p> </td> </tr> </tbody> </table> <p><strong>Conclusion: </strong>The findings underscore the importance of integrating cognitive and behavioral mechanisms in understanding social anxiety and highlight the utility of machine learning approaches in capturing complex interactions among psychological predictors. The study provides evidence for the central role of fear of evaluation and safety behaviors, particularly in interaction with intolerance of uncertainty and self-focused attention, and suggests that AI-based models can enhance predictive accuracy and inform personalized intervention strategies for social anxiety.</p> https://journals.kmanpub.com/index.php/jarac/article/download/5260/9564