Abstract:
The present study investigated whether the Beck Depression Inventory-II (BDI-II)
measures the same construct in exactly the same way across the groups of interest
and time in South Africa. The degree to which items or subtests of the BDI-II have
equal meaning across qualitatively distinct groups of examinees (e.g., culture and
gender) was explored. Measurement Invariance (MI) of the BDI-II across race
(blacks and whites), gender and time (two weeks lag) was examined in a sample of
university students, from two universities located in diverse geographical regions of
South Africa (N = 919). Confirmatory factor analysis (CFA) was used to test the fit of
the hypothesized three-factor model established through exploratory factor analysis
(EFA), and the results from these analyses indicated that the BDI-II was most
adequately represented by a three lower-order factor structure (appropriately named
Negative attitude, Performance difficulty and Somatic complaints). Results based on
multigroup confirmatory factor analysis (MCFA) (i.e., means and covariance
structures [MACS]) indicated that there was factorial invariance for this three lowerorder
factor structure across groups and time, suggesting that the BDI-II provides an
assessment of severity of depressive symptoms that is equivalent across race,
gender and time in university students. Results indicated that MI was established at
the level of configural, metric and scalar invariance for race, gender and across time.
However, there was some evidence of differential item functioning (DIF) and
differential additive response style (ARS) across race, with two noninvariant
intercepts (items 5 and 14) and three item intercepts (items 11, 14 and 18) across
gender being identified. Additionally, results of latent mean differences were
presented to explain group differences. The study concluded with recommendations
for future studies.
