Assessment of Unidimensionality and Differential Item Functioning of WAEC 2021 Physics Multiple-choice Test Items

Abstract

The study investigated assessment of unidimensionality and differential item functioning of WAEC 2021 Physic Multiple-choice Test items. The sample size for the research is one thousand one hundred and seventy-nine (1,179) students, which was obtained from a population of 7,387, made up of 5,304 students from public schools and 2083 students from private schools in Bayelsa State. The sample size of the study was obtained using multi-stage sampling techniques in stages. WAEC 2021 Physics paper 1 (multiple-choice test items) was used as the research instrument. Its validity was believed to have been obtained because it is a paper owned by an examining body WAEC. Cronbach Alpha was used to obtain the reliability and it yield a reliability coefficient of 0.77 as a measure of consistency, which reveals the instrument was good for the study. Principal Components Analysis (PCA) of Rasch residual was conducted using WINSTEPS to determine the unidimenality of the instrument. Frequency count and percentage was used to answer the research question while the hypothesis was tested using Chi-square goodness of fit statistic at 0.05 alpha level of significant. Raju Area index an item response theory (IRT) method was used to assess the presence of differential item functioning (DIF) in the study. One major finding revealed, WAEC 2021 Physics Multiple-choice test item do not significantly function differently between male and female testees. Based on the findings it was recommended for item writers at all levels, be it the classroom or examination bodies should be adequately trained on how to identify DIF items and able to construct items free of biasness.