Exploring voluntary employee attrition prediction using personality traits and machine learning: An HR analytics and decision support study based on a public dataset

Voluntary employee attrition imposes substantial organizational costs through workforce instability and knowledge depletion. Despite the rapid growth of people analytics, most predictive voluntary attrition models rely on observable variables, overlooking the psychological determinants of voluntary attrition. This study addresses this gap by developing and evaluating machine learning models that integrate Big Five personality traits with HR data, through a precision-optimized evaluation framework aligned with HR intervention costs. Three algorithms were trained on a dataset of 1,129 employees - Logistic Regression, Support Vector Machine (SVM), and Random Forest (RF). Precision was prioritized over Recall to minimize wasted retention resources. The RF achieved the highest performance (AUC = 0.69, Precision = 0.74), outperforming Logistic Regression (AUC = 0.64) and SVM (AUC ≈ 0.66). All five Big Five traits ranked among the top seven predictors alongside tenure and age, suggesting their contribution to attrition risk. The findings support personality-informed models for evidence-based retention decisions and workforce planning.