Non-parametric Method of Detecting Differential Item Functioning in Senior School Certificate Examination (SSCE) 2019 Economics Multiple Choice Items

Abstract

This study analyzed Non-Parametric Methods of Detecting Differential Item Functioning (DIF) in SSCE 2019 Economics Multiple Choice Items. Three (3) research questions were raised to guide the study. The survey research design was employed. The population of the study consists of 14,400 SS II Economics students from Edo South Senatorial District of Edo State, Nigeria. The number of sample size used in the study was 1,440 SS II Economics students. The stratified random sampling technique was adopted for this study. The instrument used to gather information was 60 Economics multiple choice items of NECO SSCE 2019 June/July. The instrument was validated by experts in Economics and Measurement and Evaluation and Quality Assurance Unit of National Examination Council (NECO) officials. Reliability was determined using Cronbach’s Alpha Statistics and coefficient of 0.72 was obtained to detect the items that functioned differentially by sex, three (3) DIF methods (Mantel Haenszel, Standardized p-diff and transform item difficulty (delta plot) were used. The findings showed that the three (3) methods displayed DIF. It was found that standardized p-difference and transform item difficulty are most suitable methods than Mantel Haenszel statistics in detecting 2019 Economics Multiple Choice Items prepared by SSCE 2019. It was also recommended that other non-parametric statistics be used to detect differential item functioning.