EVALUATION OF THE QUALITY OF FORMATIVE ONLINE MULTIPLE-CHOICE QUESTIONS TEST USING THE THREE-PARAMETER ITEM RESPONSE THEORY MODEL

Abstract

The question bank for the Biology - Genetics course has been used for years without a systematic quality assessment based on difficulty and discrimination parameters, relying instead on subjective instructor judgment.This study aims to evaluate the quality of 100 online multiple choice questions (MCQs) on the Cell Structure.The three-parameter logistic item response theory (3PL-IRT) model was applied. Data were collected from 1278 test results of students in the 48 and 36 courses during the 2022–2023 academic year and analyzed using the “ltm” package in R.The analysis revealed that 97% of the questions demonstrated acceptable discrimination across varying difficulty levels: 1% very difficult, 12% difficult, 35% moderate, 49% easy, and 3% very easy questions. Discrimination levels included 42% very good, 26% good, 29% average, 1% poor, and 2% very poor. Overall, 87% of the questions were considered high quality, though 10 questions with high guessing levels require revision to enhance reliability.The use of 3PL-IRT models is essential for evaluating and refining question banks, ensuring their reliability and effectiveness in assessing student abilities and supporting learning in the Biology - Genetics course.