Assessment of Item Bias in BECE 2024 Mathematics Multiple Choice Using Differential Item Functioning in Public Schools, Zamfara State

Abstract

The study assess the item bias on students using Differential Item Functioning (DIF) in BECE 2024 mathematics examination in Zamfara State. The research adopted descriptive research of ex-post facto type. The sample size of the study was 400 students drawn through a multi-stage sampling approach from all the Fourteen (14) Local Government Area of the state. A proportionate stratified random sampling technique was used to select 280 male and 120 students from the selected schools in each of the Local Government Area in Zamfara State. The instrument used for data collection was 30 items multiple-choice Mathematics items administered by BECE 2024. Both face and content validity were ascertained to support the validity earlier established by the examination board. The data collected were analyzed using logistic regression. The finding revealed that 17items function differently between male and female examinees, while the second finding shows that 11 items functions differently between mixed and single-sex schools examinees. Based on the finding the study recommended among others, that test experts and developers should explore the use of DIF approach to detect items that are bias in mathematics examination items. It was also recommended that (BECE) should carry out differential item functioning analysis for all the test items as part of test development process.