Academic performance of students in Relation to Emotional Intelligence and mental health

Emotional competences and psychological functioning have been found to affect academic achievement, but their unique explanatory and predictive power is unclear. The current study aimed at analyzing the personality trait of emotional intelligence (EI), mental health related emotional well-being and academic achievement of secondary school students in an integrated statistical and machine-learning approach. Secondary cross-sectional data of 96 students aged 15–18 years was analyzed using Pearson and Spearman correlation, hierarchical regression (HC3 robust inference), sensitivity analysis, and Elastic Net regression (nested five-fold outer cross-validation and three-fold inner cross-validation). There was no significant bivariate correlation between achievement and either EI, r = 0.192, p = 0.121 (after Holm adjustment), or emotional well-being, r = 0.179, p = 0.121, but EI was a significant correlate of emotional well-being, r = 0.438, p < 0.001. Well-being did not add much in terms of explanatory power and the explanatory power of demographic controls increased by adding EI, from 0.1658 to 0.2275. Adding EI and well-being increased held-out R² from 0.0804 to 0.1345; the broader psychological model achieved R² = 0.2187, MAE = 0.8043, and RMSE = 1.0069. An eight-predictor EI-branch model (MAE = 0.8012) was found to be preferred by ablation; repeated validation showed partition sensitivity. Small sample sizes and the lack of external validation make claims of deployment impossible and psychological measures provide only modest incremental information.