Asian Journal of Probability and Statistics

This study aims to analyze the influence of stress, anxiety, and depression on the quality of life of University of Lampung students using a Structural Equation Modeling (SEM) approach with the Diagonally Weighted Least Squares (DWLS) estimation method, which is suitable for Likert-scale ordinal data. The study utilized primary data obtained via online questionnaires from 300 active University of Lampung students across eight faculties, selected using purposive sampling. Prior to SEM analysis, the instrument underwent validity and reliability testing. Test results showed that all indicators were valid, with calculated r-values exceeding critical r-table values, while a Cronbach's Alpha of 0.92 indicated excellent instrument reliability. Measurement model results revealed that all indicators had loading factors ranging from 0.505 to 0.857, thereby satisfying convergent validity requirements. In the initial structural model, stress, anxiety, and depression did not show a significant direct effect on quality of life (p > 0.05). However, the inclusion of emotion regulation as a mediating variable revealed a significant indirect effect, indicating that emotion regulation mediates the relationship between stress, anxiety, depression, and quality of life. The structural model indicated that emotion regulation exerts an indirect influence on quality of life, with an R² value of 0.028. This suggests that 2.8% of the variance in quality of life is explained by emotion regulation within the proposed model, while the remaining 97.2% is explained by factors outside the model. Goodness-of-fit evaluation yielded values of RMSEA = 0.078, NFI = 0.976, GFI = 0.982, AGFI = 0.973, IFI = 0.985, and CFI = 0.985, indicating a good model fit. Thus, the SEM-DWLS approach is capable of producing a suitable model for ordinal data and demonstrates that emotion regulation mediates the influence of stress, anxiety, and depression on university students' quality of life.