Canonical Correlation Analysis: Methodology andApplication to Student Mental Well-Being

We present a methodological overview of canonical correlation analysis (CCA), a multivariate technique used to investigate the relationship between two different sets of variables observed on the same sample. We provide an explanation for the underlying method of CCA including the computation of canonical correlations, canonical variates as well as performing tests of significance. We then apply CCA to a dataset measuring students’ life circumstances and their mental well-being. Our results highlight the extent to which students’ mental well-being is associated with their life circumstances. We reveal the factors that contribute most strongly to this relationship: social support, financial status and work life balance contribute positively whereas anxiety and isolation have a negative impact. The analysis demonstrates the potential of CCA, but also its limitation in capturing multidimensional aspects of mental well-being.