Analisis Kualitas Butir Soal Ulangan Harian Matematika Di Kelas IV MI Al-Islamiyah Menggunakan Rasch Model

Authors

  • Amalia Rahmayani Universitas Pendidikan Indonesia
  • Tiurlina Universitas Pendidikan Indonesia
  • Fitri Alfarisa Universitas Pendidikan Indonesia

DOI:

https://doi.org/10.37150/perseda.v5i3.1716

Keywords:

Daily Test, Mathematics, Rasch Model.

Abstract

 The daily test question is one of the evaluation questions given by the teacher to students to
measure understanding of the material that has been given but the daily test questions at MI
Al-Islamiyah have not been analyzed for items. This is also evidenced by the results of research from Wiguna in 2018 which states that most teachers find it difficult to analyze tests and the ability to compose learning outcomes tests is still low. The quality of the items can be known by analyzing the items, so that the teacher can know the quality of each item as well as the ability of the students so that the items can adjust to the average ability of the students. This study aims to see the quality of daily test items using the Rasch Model. The Rasch model can see the quality of the questions based on the level of difficulty. The approach used is a quantitative approach with a descriptive method. The object of this research is daily test questions in mathematics with the research subject being 152 students of class IV MI AlIslamiyah. The data collection technique was carried out using documentation methods such as teacher-made daily test sheets, answer keys and student response sheets. Based on the results of the analysis showed that all items matched the Rasch Model with an average item difficulty level which was in the range of -2.51 to 1.46. The value of the test information function is 27.134 on the student's ability of 0.0368 logit and the measurement error is 0.19197. The student's ability level is 64% moderate as evidenced by 77.78% of the items have been mastered by the students.

Downloads

Published

2022-12-06

Issue

Section

Articles
Abstract viewed = 214 times