Exploring the Equivalence of Factor Structure in a Measure of Depression

Although equivalence of measures is a fundamental requisite in comparative research, re-searchers often have ignored or paid little attention to this important issue. This study employs exploratory factor analysis to evaluate the equivalence of factor structure of the short version (11 of 20 original items) of the well-known Center for Epidemiologic Studies-Depression (CES-D) Scale between Black and White women. The study used data from a nationwide probability sample that included 438 young (ages 24 to 59 years) Black females, 652 young White females, 340 older (ages 60 or older) Black females, and 764 older White females.The results revealed that the short CES-D Scale had different factor structures according to age and race. In comparative research, it is imperative that researchers employ compatible instruments to collect and analyze data. More specifically, the concept of interest must be truly equivalent between comparison groups. Meaningful results can be achieved only when equivalent measures of concepts or variables are used; otherwise, observed group differences can be an artifact of differences in measurement. For example, group differences in the total score of a multidimensional scale (i.e., a scale that has two or more factors) become ambiguous if the pattern of responses to different items varies between groups. As a result, the mean score of one group may be a function of one set of items, and the mean score of another group may be a functio