Shogbesan, Y. O, Asowo A. P and Ogunsaki, I. B.
This study investigated the extent to which moderating variables such as sex and school
type affect the occurrence of differential item functioning for compromised and secured
economics items. The study adopted the quasi-experimental research design with a
population which comprised all the secondary school students in Ogun State and a
sample of 1500 economics students selected through multi-stage sampling. The
instruments used for the study were “Economics Achievement Test (EAT)” -adapted from
NECO 2015 SSCE economics objective test paper III and an OMR-type answer sheet.
Data collected was subjected to DIF analysis which was calibrated in mirt of R language.
The results showed that sex and school type of examinees had a moderating effect on the
occurrence of differential item functioning (DIF) for 34 (56.7%) and 42 (70%)
compromised and secured item conditions. In addition, the effect of compromised and
secured items contributed more to the occurrence of DIF based on sex and school type
of the examinees. It was recommended that differential item functioning analysis should
be carried out on all items suspected to be compromised as a statistical evaluation by
test experts and examination bodies.