GHQ, WEMWBS and SF-12: what is the difference? Longitudinal analysis of the overlap and divergence of the three main health measurement scales in Understanding Society
|Day:||Thu 4 Jul|
Understanding Society introduced the Warwick-Edinburgh Mental Wellbeing Scale (short version), a measure that is stipulated to capture complete mental health and is advertised as the new "gold standard" for measuring well-being. With the availability of seven waves of data, it is now possible to conduct in-depth longitudinal analyses of the psychometric properties and associations of the WEMWBS. These results are compared and contrasted with competing and more established health scales, namely the General Health Questionnaire of negative mental health, and the Short-Form Health Survey (12 items) of health functioning.
Longitudinal confirmatory factor analyses are employed to test the factor loadings, variances, and stability of the GHQ, WEMWBS and SF-12. Distributional differences for gender, age groups, disability, objective health, psychological traits (BIG5) and various socio-economic factors are controlled for. Structural equation models and quantile regressions are employed to investigate whether these differences are especially pronounced at the top levels of well-being, as implied by positive health theories.
The overlap between the three scales is moderately high to very high, and there are no large time-trend differences between the scales. Even though the 3-factor structure can be retained in principle, there are major cross-loadings between the three scales. Best model fit is achieved by distinguishing between pooled negative mental health and well-being items, rather than between the scales. Nonetheless, within the WEMWBS the eudaimonic vs hedonic items can be statistically separated.
Psychometrically, the WEMWBS does not appear to provide much unique information that extends above and beyond the GHQ and SF-12.