Abstract:
The research aimed at examining whether the 2012-2016 May/June West African Senior Secondary Certificate Examination (WASSCE) in core subjects exhibited gender and location differential item functioning (DIF) in Ghana using the cross-sectional design. Six research hypotheses and one research question were formulated for the study. A sample of 36,035 candidates consisting of 8,994(English Language), 8,935(Mathematics), 9,089(Integrated Science) and 9,017 (Social Studies) candidates was selected from a population of 273,289 candidates each who sat for the examination from 2012-2016. The instrument for the study was the 50 multiple- choice test items each for Science, Mathematics and Social Studies and 78 (2015), 80 (2016) and 100 (2012, 2013 and 2014) English Language. MH, LR and IRT DIF detection methods were used to identify items that exhibited DIF. The findings showed that there was a significant gender differential item functioning. There was also a significant location differential item functioning as all three methods detected items that function differentially among the five regions under study. There was a high degree of agreement between the Logistic regression, Mantel Haenszel and 3PL Item Response Theory in identifying items with DIF. It was concluded that some items in exams used by WAEC exhibited significant DIF and it was recommended that DIF studies should be conducted by test developers on their test so that the items exhibiting Differential Item Functioning (DIF) could be revised or eliminated to enhance fairness.
