This study examines differences in the structure of the Center for Epidemiologic Studies Depression (CES-D) scale across three generations of Mexican Americans. The covariance structure of the items was viewed as a function of several parameter matrices. Factorial invariance was evaluated by testing a series of hypothesis involving equivalence constraints on one or more parameter matrices. The findings revealed that the proposed 12-item model fits the data within each of the three generations adequately. Although no generational differences in factor loadings were revealed, structural variations of the CES-D in measurement error variances were found. Measurement error variances associated with certain items in the older generation are significantly greater than those in the middle-aged and young generations. In addition, socioeconomic characteristics, acculturation, and health status fail to explain the observed differences in the structure of the CES-D among the three generations.
ASJC Scopus subject areas