Abstract
Methods
Cross-sectional study among 4,146 individuals (mean age 16.3 years). The students completed a health questionnaire that included
the GHQ-12. A confirmatory factor analysis (CFA) was conducted with four factor structure models. Three of the models were
theory-driven: unidimensional, two-dimensional (positive and negative questions), model proposed by Graetz (anxiety and depression,
social dysfunction, loss of confidence); and the fourth model was based on our exploratory factor analysis (EFA).
the GHQ-12. A confirmatory factor analysis (CFA) was conducted with four factor structure models. Three of the models were
theory-driven: unidimensional, two-dimensional (positive and negative questions), model proposed by Graetz (anxiety and depression,
social dysfunction, loss of confidence); and the fourth model was based on our exploratory factor analysis (EFA).
Results
The Cronbach’s alpha for internal consistency was 0.82. A three-dimensional structure was identified in the EFA. The first
factor included items 1, 2, 5, 6, 7, and 9 of the GHQ-12; the second, items 3, 10, and 11; and the third, items 4, 8, and
12. The three factors together explained 53.7% of the variance. The model with the best fit in the CFA was the three-dimensional
model proposed by Graetz, followed by the three-dimensional model derived from the EFA. These two models had acceptable goodness-of-fit
indices.
factor included items 1, 2, 5, 6, 7, and 9 of the GHQ-12; the second, items 3, 10, and 11; and the third, items 4, 8, and
12. The three factors together explained 53.7% of the variance. The model with the best fit in the CFA was the three-dimensional
model proposed by Graetz, followed by the three-dimensional model derived from the EFA. These two models had acceptable goodness-of-fit
indices.
Conclusions
In an adolescent population from Southern Europe, the GHQ-12 showed high internal consistency. The factor structure that best
fitted the data was the Graetz three-dimensional model. However, the high correlations observed between factors suggest that
the GHQ-12 should be used as a unidimensional scale, as currently done.
fitted the data was the Graetz three-dimensional model. However, the high correlations observed between factors suggest that
the GHQ-12 should be used as a unidimensional scale, as currently done.
- Content Type Journal Article
- Pages 1-8
- DOI 10.1007/s11136-011-0038-x
- Authors
- Alicia Padrón, Public Health Centre, Area 10. Consejería de Sanidad de Madrid. C/Francisco Gasco Santillán 2, portal B 2º, Getafe, Spain
- Iñaki Galán, National Centre for Epidemiology, Instituto de Salud Carlos III, Madrid. C/Monforte de Lemos 5, Madrid, Spain
- María Durbán, Department of Statistics and Econometrics, Universidad Carlos III de Madrid, C/Madrid 126, Getafe, Spain
- Ana Gandarillas, Department of Epidemiology, Consejería de Sanidad de Madrid, C/Julián Camarillo 4, Madrid, Spain
- Fernando Rodríguez-Artalejo, Department of Preventive Medicine and Public Health. School of Medicine, Universidad Autónoma de Madrid/IdiPAZ, C/Arzobispo Morcillo 2, Madrid, Spain
- Journal Quality of Life Research
- Online ISSN 1573-2649
- Print ISSN 0962-9343