Item Analysis of the Developed Reading Preparation Materials for the English Proficiency Test

Abstract

This study is an integral part of research and development (R&D) on digitalized reading materials for the English proficiency test. It focused on item analysis of the fifty reading comprehension questions embedded in a newly developed Digitalized English Proficiency Test (EPT) Reading Preparation Materials platform. Although previous studies have examined item quality in teacher-made tests, little is known about the statistical quality of reading items developed specifically for a digital EPT preparation platform at the university level; this study addresses that gap and offers novelty by integrating item facility, item discrimination, and distractor efficiency into an item-level decision matrix for platform improvement. Because a fellow researcher independently constructed the items and they had never been empirically evaluated, this study aimed to analyze item facility, item discrimination, and distractor efficiency. It used a quantitative descriptive design and analyzed response data from 38 English Education students who completed all 50 items, following Arikunto’s (2018) classical test theory procedures. The results showed a mean item facility index of .64, with 50% of items falling in the medium/acceptable range. The mean item discrimination index was .34, and 109 of 150 distractors (72.7%) were functional. Based on the combined criteria, 6 items (12%) were recommended for retention, 41 items (82%) required revision, and 3 items (6%) were recommended for removal. The main limitation is that reliability was not calculated and the pilot sample was relatively homogeneous in English proficiency, which likely restricted discrimination estimates; therefore, findings should be interpreted as initial, sample-dependent evidence. These findings provide an empirical basis for iteratively improving the platform’s item bank before wider implementation.