A reanalysis of the center for epidemiological studies depression scale (CES-D) using non-parametric item response...

The "Center for Epidemiological Studies Depression Scale" (CES-D; Radloff, 1977) is a questionnaire used world-wide to measure depressive symptoms. Although the original four-factor-structure has been widely accepted and replicated, some studies point to other factor-structures like a one- and two-factor-structure. The goal of the current study was to evaluate the factor structure of the CES-D (one-, two- and four-factor-structure), which was found using classical test theory (CTT), with two non-parametric item-response-theory-models (Mokken-Scaling; Monotone-homogeneity-model; MHM and Double-monotonicity-model; DMM). To this end, a representative German sample was analyzed (N = 2,507). Regarding the one-factor-model, neither the MHM nor the DMM were supported; the two-factor-model and the four-factor-model, however, satisfied the assumptions of the MHM but not of the stricter DMM. Sum scores therefore constitute ordinal scales for the two-factor and four-factor-structure. This justifies, for example, the use of percentile ranks or medians and associated statistical techniques. Especially the use of the 16 negative formulated items seems appealing. Further studies should explore, whether the use of these 16 instead of the original 20 items is of advantage regarding the practical validity of the CES-D (e.g. higher sensitivity and specificity in screening applications and better temporal stability).