The Rasch Model for Analysing the Indonesian Language Examination Instrument

DOI

10.2991/assehr.k.220129.070How to use a DOI?

Keywords

item response theory; rasch model; difficulty

Abstract

This study aimed to analyze instrument for evaluation in the Indonesian Language Test, based on its difficulty level. The Item Response Theory [IRT] was applied in this study. We investigated the examinee’s responses to the instrument parameter. The respondents were 259 Vocational High School students of the 10th graders. We used Rasch Model to evaluate the difficulty level and create five categories from very difficult to very easy questions. Employing the R-Studio program for analysis, the result showed that out of 40 question items, one item was categorized as very difficult, one was categorized as difficult, twenty-nine items were classified as moderate, five items were categorized as easy, and four items were categorized as very easy. We grouped the difficult, moderate, and easy question items into the average difficulty category. As a result, 72.5 % of the items used in year-end assessment were considered moderately difficult, while 34 items, or 85% of the total items, match the Rasch model. The percentage of the items to be included in the question banks were the items that have moderate difficulty levels and match the Rasch Model, which was 60 %. It suggests that Rasch Model is applicable for instrument evaluation.

Copyright

© 2022 The Authors. Published by Atlantis Press SARL.

Open Access

This is an open access article under the CC BY-NC license.

Download article (PDF)

TY  - CONF
AU  - Muhammad Fajar Suminto
AU  - Sri Wening
PY  - 2022
DA  - 2022/01/31
TI  - The Rasch Model for Analysing the Indonesian Language Examination Instrument
BT  - Proceedings of the 5th International Conference on Current Issues in Education (ICCIE 2021)
PB  - Atlantis Press
SP  - 383
EP  - 388
SN  - 2352-5398
UR  - https://doi.org/10.2991/assehr.k.220129.070
DO  - 10.2991/assehr.k.220129.070
ID  - Suminto2022
ER  -