An evaluation of the independence of the Health of the Nation Outcome Scales

Objective: In this study the independence of the scales/items in the Health of the Nation Outcome Scales (HoNOS) was empirically investigated. Method: Parallel analysis using random column permutation and bootstrapping were used to compare the factor structure, intercorrelations and Cronbach's alpha from the original HoNOS study and also recently collected HoNOS ratings. Random permutation ensures that the data has the same distributions as the data it is based on, but that the variables are now independent. Results: It is shown that both of the real HoNOS data sets are significantly different to the independent items data in many ways. An examination of fit statistics from confirmatory factor analysis is also used to show that the independence model is a very poor fit to the data. Conclusions: It is clear that the 12 HoNOS scales are unlikely to be independent. There is a need for more research to clarify the appropriate structure of HoNOS, and also to consider whether some of the items need either replacing or augmenting.