Rethinking Mental Health Assessment: A Network-Based Approach to Understanding University Students’ Well-Being with Exploratory Graph Analysis

Abstract

Mental health (MH) in university students is often studied through isolated variables. However, a dynamic systems perspective suggests that psychological well-being results from interactions among multiple dimensions such as personality, mood, resilience, selfesteem, and psychological distress. A total of 928 university students (M = 21.01 ± 1.95) completed validated questionnaires: Big Five Inventory (BFI-44) for personality, Profile of Mood States (POMS), Connor-Davidson Resilience Scale (CD-RISC 25), Rosenberg SelfEsteem Scale, and Depression Anxiety Stress Scale (DASS-21). Exploratory Graph Analysis (EGA) using the EGAnet package in RStudio (v. 2025.09.01) was employed to identify latent dimensions and their interconnections. EGA revealed five stable and interconnected dimensions with good fit indices (TEFI = −9.00; ≥0.70): (a) Personality as socio-emotional regulation, (b) Mood as a generalized affective continuum, (c) Resilience as a unified coping process, (d) Self-esteem based on competence and self-worth, and (e) Psychological distress integrating depression, anxiety, and stress. MH appears as a complex and dynamic network of interrelated psychological components. This network-based approach provides a more integrative understanding of well-being in students and supports the development of interventions that target multiple dimensions simultaneously, enhancing effectiveness in academic settings.