Meta-Analysis of Factor Analyses of the General Health Questionnaire – Short Forms GHQ-28 and GHQ-30

Abstract

Abstract. Two meta-analyses of exploratory factor analyses of the General Health Questionnaire short forms, GHQ-28 ( N = 26,848, k = 40) and GHQ-30 ( N = 43,151  k = 25), were conducted to determine the consistent factors found in each test and any common factors across them. Five databases (PsycINFO, PubMed, BASE, Semantic, and Google Scholar) were searched in 2021. Reproduced correlations derived from the original studies’ factor matrices and aggregated across studies were factor analyzed for the meta-analyses. For the GHQ-28, the standard four subscales of somatic, anxiety, social dysfunction, and depression were clearly identified and strongly supported by a four-factor structure. For the GHQ-30, a four-factor solution identified factors of anxiety, depression, social dysfunction, and social satisfaction, the first three factors shared a number of items with the same scales found in the GHQ-28. These shared factors appear similar across tests and should help bridge research using the GHQ-30 and the GHQ-28. Confirmatory factor analyses supported the four-factor models in both tests. The four standard subscales of GHQ-28 were strongly supported and can be recommended. The three similar factors in the GHQ-30, as well as the social satisfaction factor, appear reasonable to use.