ANALYSIS OF ITEM DIFFICULTY AND DISCRIMINATION DRIFT IN THE 2023 AND 2024 MATHEMATICS WEST AFRICAN SENIOR SCHOOL CERTIFICATE EXAMINATIONS USING THE FOUR-PARAMETER LOGISTIC MODEL OF ITEM RESPONSE THEORY

Abstract

Without empirical evidence-based information on drift analysis of examination items, high-stake examination bodies may continue to use or compare items across years without knowing whether their measurement properties have changed. So, there is need for Analysis of Item Difficulty and Discrimination drift in the 2023 and 2024 Mathematics West African Senior School Certificate Examination using the four-parameter logistic model (4PLM) of item response theory. The study determined whether the two examination years exhibit comparable distributions of difficulty and discrimination after accounting for guessing and the possibility that highly proficient candidates may answer an item incorrectly. Three research questions guided the study. The study adopted longitudinal research design repeated measure. The target population is comprised of   22500 SS3 Students in public secondary schools across the six educational zone of Anambra State. The sample comprised 1125 students obtained through multi-stage sampling procedure using simple random technique and purposive sampling.2023 and 2024 May/June WASSCE Mathematics multiple-choice questions was adopted instrument used to generate data. The test was dichotomously scored 1 for correct answer and 0 for wrong answer using the scoring guide and the results were used to determine the extent of drift in the difficulty and discrimination, parameters across administrations, Data were analysed using Item Parameter Drift (IPD) analysis within the framework of IRT. Results showed that the majority of the 2023 and 2024 WASSCE Mathematics Multiple- Choice Questions exhibited moderate drift in item difficulty parameters across administrations, indicating that most items maintained relatively acceptable stability over time. But the 2024 examination recorded a higher proportion of high-drift items than the 2023 examination, indicating comparatively lower stability in item difficulty characteristics. The findings of the present study therefore suggest the need for careful psychometric review, recalibration, revision, or replacement of high-drift items before they are reused in future WASSCE Mathematics examinations. Such measures would help strengthen score comparability, preserve measurement accuracy, and improve the overall integrity of the examination system.