Validity and factor structure of the General Health Questionnaire (GHQ-12) in university students

Abstract

Introduction: The aim of this study was to investigate psychometric properties of General Health Questionnaire of the students.Method: In this study, 500 students of medical universities in Tehran were selected in a relative sampling. Participants responded to the General Health Questionnaire 12 items form and interviewed by trained interviewers with the Composite International Diagnostic Interview Form (CIDI).Results: Exploratory factor analysis through principle component analysis showed the two-factor structure of this questionnaire. This questionnaire was made of two components, of which the first factor explained the greatest amount of variance. The best cut-off point of GHQ-12 questionnaire was 9, and the sensitivity, specificity, and overall classification error for the best cut-off point were81.5%, 77.1%, and 17.8%, respectively. The cut-off point with maximum sensitivity index (95.5%) was five and the maximum specificity score (95.8%) was 15. Construct validity by examining correlations between the subscales of the questionnaire together with the total score was calculated, which were strongly significantly correlated. Cronbach's alpha coefficient was 0.92 and the split-half and Spearman-Brown reliability was 0.91.Conclusion: The best cut-off point for this questionnaire was 9 for two-stage studies and 15 for one-stage studies. Obtained construct validity and reliability coefficients indicate that this questionnaire is suitable for screening psychopathology in university students.