Abstract
BACKGROUND: Despite extensive evaluations of the Center for Epidemiological Studies Depression Scale (CESD), its shortest version, the Eight-Item version (CESD-8), is less investigated, with absolute lack of information on its psychometric properties in the Arab world. METHODS: To fill the gap, data collected via an anonymous online survey from Saudi samples of students (N = 979, 422) and employees (N = 314), were analyzed through exploratory factor analysis, confirmatory factor analysis, and multigroup analysis to examine the structure and measurement invariance of the CESD-8. Convergent validity and internal consistency tests involved correlating the CESD-8 with its subscales, item analysis measures, and intra class correlations. Criterion validity tests involved correlating the CESD-8 and its subscales with a single-item measure of happiness. RESULTS: Exploratory factor analysis produced two factors (negative affect and positive affect) with eigen values >1, which explained 86.4 % of the variance. In confirmatory factor analysis, the crude exploratory factor analysis model had good fit while the fit of the unidimensional CESD-8 and another two-factor structure (depressed affect and somatic complaints) was improved by correlating the residuals of the items of positive affect (CESD4 and CESD6). A three-factor model (depressed affect, somatic complaints, and positive affect) expressed the best fit in the absence of error correlations. This model was invariant across groups of students and employees, gender, and age. The scale and its three dimensions demonstrated adequate internal consistency (alpha coefficient range = 0.65-0.89), convergent validity (item total correlation range = 0.43-0.80 and range of correlations with the CESD-8 = -0.68-0.92), and criterion validity (range of correlations with happiness scores = -0.40-0.60). CONCLUSIONS: The CESD-8 is a valid short scale for quick identification of people with depressive psychopathologies. Using the CESD-8 to detect heterogenous depressive symptoms, rather than assessing depression as a whole condition, may influence our understanding of the dynamics and treatments of depression in specific groups/cultures, with emphasis on absence of positive affect in the definition of depression among Arabs. Replications of the three-factor structure in different cultures are needed.