Research Article
Comparasional Analysis of Statistical Results Using the Rasch and Anates Model in Chemistry Learning
@INPROCEEDINGS{10.4108/eai.19-9-2023.2340586, author={Fatma Harian Dini and Siti Khadijah Dalimunthe and Ayi Darmana and Retno Dwi Suyanti and Zainuddin Muchtar and Ajat Sudrajat}, title={Comparasional Analysis of Statistical Results Using the Rasch and Anates Model in Chemistry Learning}, proceedings={Proceedings of the 8th Annual International Seminar on Transformative Education and Educational Leadership, AISTEEL 2023, 19 September 2023, Medan, North Sumatera Province, Indonesia}, publisher={EAI}, proceedings_a={AISTEEL}, year={2023}, month={12}, keywords={influence problem-solving approach folk tales}, doi={10.4108/eai.19-9-2023.2340586} }- Fatma Harian Dini
Siti Khadijah Dalimunthe
Ayi Darmana
Retno Dwi Suyanti
Zainuddin Muchtar
Ajat Sudrajat
Year: 2023
Comparasional Analysis of Statistical Results Using the Rasch and Anates Model in Chemistry Learning
AISTEEL
EAI
DOI: 10.4108/eai.19-9-2023.2340586
Abstract
The purpose of this study was to describe the comparative results of the analysis of student scores through the Rasch and Anates model approach in terms of the validity, reliability, difficulty level, and distinguishing power of the questions. This research is a qualitative research with descriptive method. The data were obtained through 46 students with a total of 40 items. Furthermore, the data obtained were analyzed through the Rasch and Anates model approach. The results showed that the quality of the instrument for measuring concept understanding through the Rasch and Anates model approach was in a good category. Reliability analysis through both applications is in the high category of 0.75 through Anates and 0.76 Person Reliability with the high category, 0.90 Reliability Items with a very high category in the Rasch model. The average level of difficulty index in Anates is with five levels of categories, while the results of the analysis of the level of difficulty based on the Rasch model show that there are four categories of item difficulty levels, namely 7 questions which are very easy, 12 easy questions, 11 difficult questions, and 10 very difficult questions. difficult. The discriminatory power of items using Anates is in the bad category of 3 questions, the sufficient category is 4 questions, and 9 items are in the good category.
