The analysis of bridging constructs with hierarchical clustering methods: An application to identity

- Hierarchical clustering (HC) is proposed for evaluating measures of bridging constructs, as identity.
- HC is compared to exploratory factor analysis (EFA).
- HC resulted more suitable for factorially-designed surveys and with small samples.
- Heat maps are indicated as graphic tools for hierarchical clustering with bridging construct measures.

When analyzing psychometric surveys, some design and sample size limitations challenge existing approaches. Hierarchical clustering, with its graphics (heat maps, dendrograms, means plots), provides a nonparametric method for analyzing factorially-designed survey data, and small samples data. In the present study, we demonstrated the advantages of using hierarchical clustering (HC) for the analysis of non-higher-order measures, comparing the results of HC against those of exploratory factor analysis. As a factorially-designed survey, we used the Identity Labels and Life Contexts Questionnaire (ILLCQ), a novel measure to assess identity as a bridging construct for the intersection of identity domains and life contexts. Results suggest that, when used to validate factorially-designed measures, HC and its graphics are more stable and consistent compared to EFA.