Measuring depression in a non-western war-affected displaced population: Measurement equivalence of the beck depression inventory

Abstract

Depression is commonly seen in survivors of conflict and disaster across the world. There is a dearth of research on the validity of commonly used measures of depression in these populations. Measurements of depression that are used in multiple contexts need to meet measurement equivalence, i.e., the instrument measures the same construct in the same manner across different groups. The Beck Depression Inventory (BDI) was administered to female trauma survivors in the United States (n = 268) and female survivors of war in Sri Lanka (n = 149). Three metrics of measurement equivalence-structural, metric, and scalar-were examined. Two- and three-factor structures of the BDI that have been identified in other populations did not provide a good fit for our data. However, a bifactor model revealed a similar general distress dimension across populations, but dissimilar secondary dimensions or subfactors. The Sri Lankan subfactor comprised of predominantly somatic symptoms and the United States subfactor comprised of cognitive and somatic symptoms. While intercepts of individual BDI items differed, their differences seem to be offsetting. Total BDI scores across these two populations are roughly comparable, although caution is recommended when interpreting them. Making comparisons on subscales is not recommended.