The structure of the MHC-SF in a large American sample: contributions of multidimensional scaling

Abstract

Background: Multidimensional scaling (MDS) is a data reduction tool that produces a visual representation of the association between variables to facilitate the identification of underlying concepts and structures. This technique has been underused in well-being research.

Aim: This study used MDS to unravel the underlying concepts and structures in the 14 items of the Mental Health Continuum-Short Form (MHC-SF).

Methods: A sample of 22 606 American students was used. Two-dimensional MDS plots were generated for the entire sample, and separately for each gender group.

Results: MDS yielded three non-overlapping clusters of the items that correspond to the three hypothesized well-being concepts measured by the MHC-SF. The results also showed that the items vary along the two underlying dimensions of personal vs. social, and hedonic vs. eudaimonic. The same general structure was found in both females and males.

Conclusion: These results supplement and extend the findings of prior factor analytic studies on the MHC-SF by showing that the items of the scale can be distinguished along two separate continua. The results also support the tripartite model of mental well-being.

Keywords:

• MHC-SF
• multidimensional scaling
• factor analysis
• tripartite model of well-being

Declaration of interest

No potential conflict of interest was reported by the author.

Notes

1 ESEM is an extension of classic EFA. When the ESEM model does not have a structural part, ESEM factors are identical to EFA factors. However, ESEM provides many advanced features and modeling possibilities that are not available in EFA, including, for example, fit indices, standard errors for all rotated parameters, and modification indices. For technical details, see Asparouhov & Muthen (2009). In the context of the present article, readers who are not interested in considering the technical issues may regard ESEM and EFA as identical.