Postgraduate education is often associated with high psychological demands that can significantly affect students’ mental health and academic performance. Among the most common issues are depression, anxiety, and stress, which may negatively influence students’ self-efficacy—the belief in one's ability to succeed in specific situations or accomplish tasks. This study aims to compare levels of self-efficacy, depression, anxiety, and stress among postgraduate students and identify multivariate outliers using Mahalanobis Distance, a statistical method that accounts for correlations between variables.

A cross-sectional survey design was employed, involving a sample of 250 postgraduate students from various academic disciplines. Participants completed standardized instruments: the General Self-Efficacy Scale (GSE) and the Depression Anxiety Stress Scales (DASS-21). Descriptive statistics, correlation analysis, and multivariate analysis were conducted. Mahalanobis Distance was used to detect individuals whose psychological profiles significantly deviated from the group norm.

Results revealed significant negative correlations between self-efficacy and each of the mental health indicators (depression, anxiety, and stress), indicating that higher levels of psychological distress are associated with lower self-efficacy. The Mahalanobis Distance analysis identified a subset of students as multivariate outliers—those exhibiting extreme combinations of low self-efficacy and high psychological distress.