Multitrait-multimethod assessment of anxiety and depression scales.

Abstract

This multitrait-multimethod study explored the ability of three methods to separate the two constructs of anxiety and depression. Methods used were true-false (Taylor Manifest Anxiety Scale and Multiscore Depression Inventory), adjective check lists (Multiple Affect Adjective Check List and Depression Adjective Check List), and 5-point rating scales (Positive Affect and Negative Affect scales of the Positive and Negative Affect Schedule). Participants were 148 community college students of diverse ethnic background from a large southwestern city. Convergent validity was evident by moderate monotrait-multimethod correlations (.61, .62, and .78 for anxiety and .56, .52, and .49 for depression). The Positive Affect scale, representing the absence of depression, correlated negatively with all other scales, as expected. Correlation coefficients were reflected to remove the negative sign. Discriminant validity was comprised when many of the multitrait-multimethod correlations (ranging from .23 to .70) were higher than the monotrait-multimethod correlations and by the consistent strength of the strength of the multitrait-monomethod correlations (.75, .48, and .37). The contribution of method to the correlations was demonstrated by large differences between multitrait-monomethod and multitrait-multimethod correlations.