Application of Item Response Theory (IRT) to the GHQ-12 in Spanish University Students

AbstractBackground: The 12-item General Health Questionnaire (GHQ-12) is widely used to screen psychological distress, yet its factorial structure and item-level properties remain to be fully explored in university populations, where psychological distress is highly prevalent, and interventions are urgently needed.Methods: A cross-sectional survey was conducted with Spanish undergraduate students (N = 798). Factorial structure was examined through confirmatory factor analyses. Reliability and discriminant validity were assessed using satisfaction with life, sense of coherence, and self-esteem. Item-level properties were analysed using Item Response Theory (Graded Response Model), focusing on discrimination, threshold parameters, information curves, and differential item functioning (DIF) by gender.Results: Analyses supported an essentially unidimensional structure, despite slightly superior fit of multidimensional models, given high interfactor correlations. Internal consistency was robust (α = .88, ω = .89), and discriminant validity was confirmed. Item Response Theory (IRT) results indicated optimal precision in the mild-to-moderate distress range (θ ≈ 0–2). Items such as “Difficulty concentrating” and “Loss of sleep” emerged as early-warning signals, while “Couldn’t overcome difficulties” and “Feeling unhappy” and depressed functioned as critical markers of moderate-to-high risk. More severe items, including “Thinking of self as worthless”, acted as indicators for immediate referral. DIF analyses revealed minimal gender bias.Conclusions: The GHQ-12 demonstrated strong reliability, validity, and nuanced item functioning in this population. Its layered item structure supports its use not only as a global indicator of distress but also as a tool to inform stepped-care strategies in university mental health programmes, from universal prevention to targeted interventions.Keywords: GHQ-12; psychological distress; Item Response Theory; reliability; validity; university students