A mixed method approach to Ways of Thinking

Higher order factors and hierarchical structures are foundational to the study of individual differences. However higher order factors inevitably explain less variance (often around 60%) than the lower order constructs and have a lower predictive validity (Reynolds and Clark, 2001). By transposing the data matrix, hierarchical structures formed by types of people (rather than composites of variables) can be examined. Using transposed principle components analysis (tPCA) with the Ways of Thinking instrument, higher order types of people were examined. Integrating Roodenburg’s (2014) participants linked the results to her qualitative typology thereby allowing for greater explanatory power of the person-centred types. The tPCA method explained 95% of the variance in the data. Using tPCA derived person-centred types, a discriminant function analysis was conducted, which found that 94.3% of cases could be correctly classified according to their Ways of Thinking facet scores. Further research is recommended in refining the person-centred methodology.