The Analysis Of Multiple-Choice Test Quality For Reading III Class In English Education Department, Universitas Pendidikan Ganesha Bali, Indonesia

Abstract

<jats:p>This study aimed at investigating the quality of 40-item multiple-choice taken from Reading III class in English Education Department. It was carried out in order to get clearer picture and provide feedbacks on how the Critical Reading test for the new curriculum should be developed and what factors should be taken into account when developing the test. The data were taken from 24 fourth semester students in English Education Department, Unidksha, Bali who took the reading test. The data were analyzed by conducting Classical Test Theory analysis with the assistance of jMetrik software. In general, the result shows that the test was consistent but it was easy and only 40% of total items are eligible to be used. Along the findings and discussion, some feedbacks are provided for future development for the reading test.</jats:p>

Evaluation of an online teacher-made test through blackboard in an English as a foreign language writing context

“…42% of the distractors functioned while 58% did not non-function. The findings reported in this study on the easiness of the test and inefficiency of distractors and discrimination indices in many items are in line with the findings by Paramartha's (2017) that revealed that the EFL reading test taken by a sample of Indonesian school students was easy, and less than half of the test items had eligibility.…”

Examining teacher-made English test in a language school

“…Likewise, a study conducted by Kholilah (2017) revealed that even more than half of all the test items had poor discrimination ability which could hardly inform who had a better and who had less ability. Paramartha (2017), on the same token, found many items on the examined had low discrimination power. On the contrary, this present study resonates very well with the study by Lebagi et al (2017) who uncovered that most items of the teachermade English test under investigation proved to be strong in discriminating the test takers on their ability.…”

“…The finding of this degree of construct validity of a teacher-made English test neither confirms nor contrasts the findings of other studies. Other studies in the field in Indonesian context, to the author's knowledge, did not seek to examine the construct validity of the test they investigated (Hakam & Irhamsyah, 2020;Indrayani et al, 2020;Jannah et al, 2021;Kholilah, 2016;Lebagi et al, 2017;Paramartha, 2017;Santy et al, 2020;Septi et al, 2020). It is with Conquest application that allows the generation of the degree of construct validity of the test.…”

“…The weak aspects, however, indicate that again test development expertise of the teacher is not yet sufficient. Similarly, a study conducted in an English Education program in Bali by Paramartha (2017) discovered that an English reading teachermade test could hardly discriminate among students with different levels of mastery. Almost half of all the items mostly require removal and some revision.…”