KMAN Publication Inc. (KMANPUB) Applied Family Therapy Journal (AFTJ) 3041-8798 6 Serial Number 30 2025 11 01 Explainable AI Prediction of Parenting Self-Efficacy from Socio-Emotional Family Dynamics Explainable AI Prediction of Parenting Self-Efficacy from Socio-Emotional Family Dynamics 1 10 10.61838/kman.aftj.4953 EN Journal Article 2024 05 25 <p><strong>Objective:</strong> The objective of this study was to develop and interpret an explainable artificial intelligence model for predicting parenting self-efficacy from socio-emotional family dynamics among parents in Greece.</p> <p><strong>Methods and Materials: </strong>This cross-sectional study was conducted with 438 parents of children aged 6–16 years recruited from multiple urban and semi-urban regions of Greece using multi-stage cluster sampling. Participants completed standardized measures assessing parenting self-efficacy, parenting stress, emotion regulation, family emotional cohesion, parent–child attachment, coparenting quality, family conflict, family expressiveness, and perceived social support. Data preprocessing included normalization, multiple imputation, and feature selection. Predictive modeling was performed using an ensemble learning architecture integrating gradient boosting, random forest, and elastic net regression. Model training employed nested cross-validation with an 80/20 train–test split. Model interpretability was ensured through the application of explainable AI techniques, including SHAP and LIME.</p> <p><strong>Findings: </strong>The ensemble model demonstrated strong predictive performance (R² = 0.79, RMSE = 0.24, MAE = 0.18). Parenting stress and emotion regulation difficulties emerged as the most influential negative predictors of parenting self-efficacy, while family emotional cohesion, parent–child attachment security, and coparenting quality were the strongest positive predictors. Significant interaction effects were observed, particularly between parenting stress and emotion regulation (ΔR² = 0.061), indicating that effective emotional regulation substantially buffers the adverse impact of stress on parental confidence.</p> <p><strong>Conclusion: </strong>The findings highlight the central role of emotional and relational processes in shaping parenting self-efficacy and demonstrate the value of explainable artificial intelligence for modeling complex family systems with both high predictive accuracy and theoretical transparency.</p> https://journals.kmanpub.com/index.php/aftj/article/download/4953/8916