Gender-related Differential Item Functioning in a Social Studies Test: A Three-parameter Logistic IRT Analysis

Abstract

This study investigated gender-related differential item functioning (DIF) in a multiple-choice Social Studies achievement test using a three-parameter logistic (3PL) Item Response Theory framework. The instrument comprised 50 items selected from an initial pool of 120 items developed from the Upper Basic (Basic 7–9) Social Studies curriculum. It was administered to 1,245 Upper Basic Nine students in public co-educational secondary schools in Rivers State, Nigeria, comprising 602 males and 643 females selected through multistage sampling. Content and face validity were established by four experts, while a pilot study involving 60 students produced a Kuder–Richardson Formula 20 reliability coefficient of 0.84. Item difficulty, discrimination, and guessing parameters were estimated across gender groups through marginal maximum likelihood, and DIF was evaluated using likelihood-ratio procedures. The findings indicated significant uniform DIF in the difficulty parameters of several items, with a larger number of items reported as relatively more difficult for male examinees. By contrast, the discrimination and guessing parameters did not show significant gender differences in the reported analyses. The findings demonstrate the value of item-level psychometric evaluation for identifying potentially differential item functioning and supporting fairer assessment. Routine review of difficulty, discrimination, guessing, and DIF evidence may strengthen the development and quality assurance of Social Studies achievement tests.