Analysis of the distractor of the multiple-choice test using classical test theory (CTT) and item response theory (IRT)

Abstract

This study aimed to analyzed the functional distractor based on the results of the multiple-choice test using CTT and IRT. Data obtained mathematics test at one of the junior high schools (grade 7) in Sidoarjo in the 2021-2022 academic year. The test consists of 20 items, which are a collection of questions that have been standardized in the school curriculum. One hundred examinees attending this mathematics test. The analysis used Iteman 4.3 software for analysis distractor using CTT, and R program for analysis pseudo-guessing’s parameter using 3PL IRT model. The result showed that all items have distractors well function it (Prop. > 0,05) in CTT. Meanwhile, with IRT, 8 of the 20 questions were not good because the pseudo-guessing index was > 0.25. The result of this study provided important information for future study to examine the ability estimate when a test's fixed feature is the item-specific characteristic utilized for pseudo-guessing.