Construct validity was established by testing whether the items in a scale all loaded on
a common factor when within-scale factor analysis was run. Appendix B shows that all
eigenvalues exceeded the threshold of 1,0, which supports each scale’s
dimensionality.