KMAN Publication Inc. (KMANPUB) تست 3060-6713 4 Serial Number 13 2026 03 01 A LightGBM-Based Analysis of Academic Resilience Among Gifted Students with Twice-Exceptional Profiles: Contributions of Metacognitive Awareness, Emotional Regulation, and Perceived Teacher Support A LightGBM-Based Analysis of Academic Resilience Among Gifted Students with Twice-Exceptional Profiles: Contributions of Metacognitive Awareness, Emotional Regulation, and Perceived Teacher Support 1 13 10.61838/kman.prien.5678 EN Journal Article 2025 11 14 <p>The present study aimed to examine the predictive contributions of metacognitive awareness, emotional regulation, and perceived teacher support to academic resilience among gifted students with twice-exceptional profiles using a Light Gradient Boosting Machine (LightGBM) model. This cross-sectional predictive study was conducted among 428 gifted students with documented twice-exceptional profiles recruited from educational institutions across Canada. Participants completed standardized measures assessing academic resilience, metacognitive awareness, emotional regulation, and perceived teacher support. Preliminary statistical analyses included descriptive statistics and Pearson correlation coefficients. Subsequently, a LightGBM machine learning model was developed to predict academic resilience. The dataset was divided into training (80%) and testing (20%) subsets, and model performance was evaluated using the coefficient of determination (R²), root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE). Feature importance and SHAP (Shapley Additive Explanations) analyses were performed to determine the relative contributions of the predictor variables. Inferential analyses revealed significant positive correlations between academic resilience and metacognitive awareness (r = .71, p &lt; .01), perceived teacher support (r = .63, p &lt; .01), and emotional regulation (r = .58, p &lt; .01). The LightGBM model demonstrated excellent predictive performance, explaining 84.7% of the variance in academic resilience within the testing dataset (R² = .847), with low prediction error values (RMSE = 6.84, MAE = 5.18, MAPE = 5.03%). Feature importance analysis indicated that metacognitive awareness was the strongest predictor of academic resilience (43.82%), followed by perceived teacher support (33.76%) and emotional regulation (22.42%). SHAP analyses confirmed these findings, revealing that metacognitive awareness exerted the greatest influence on resilience predictions, followed by teacher support and emotional regulation. Higher levels of all three predictors were consistently associated with increased predicted resilience scores. The findings demonstrate that academic resilience among gifted students with twice-exceptional profiles is strongly influenced by a combination of cognitive, emotional, and contextual factors. Metacognitive awareness emerged as the most powerful predictor, highlighting the importance of self-regulated learning processes in facilitating adaptive academic functioning. Perceived teacher support and emotional regulation also made substantial contributions, underscoring the critical roles of supportive educational relationships and emotional competencies. The results suggest that interventions designed to strengthen metacognitive skills, enhance emotional regulation, and foster supportive teacher-student interactions may effectively promote resilience and educational success among twice-exceptional learners.</p> https://journals.kmanpub.com/index.php/prien/article/download/5678/10705