Development of a Computerized-Adaptive Test to Measure English Vocabulary Size with IRT

Abstract

To speak English fluently like a native speaker, having a vocabulary of 20,000 words are required. In many universities in Indonesia, Vocabulary Size Test (VST) is still held with Paper-Based Test (PBT), so it is less effective and efficient. The aims of this study were 1) to develop Computerized-Adaptive Testing (CAT), which helps get students with a standardized English vocabulary size, and 2) to implement it in measuring vocabulary size. The approach in this study is Research and Development (RD) with product development procedures following the steps of Borg and Gall. The findings of this study are, 1) the CAT program has Item Response Theory (IRT) parameters that can change according to the test takers' answers, and the test will stop if the estimated ability of the test takers is known (Standard Error/SE close to 0.01), and 2) with the application of the CAT program it can be seen that the total vocabulary of English Language and Teaching Department students is still 20,000. Based on these findings, this study implies that for students' vocabulary mastery to be on par with native speakers' abilities, the English Language and Teaching Department needs to evaluate the learning process and determine a particular program to increase student vocabulary.