
Item Analysis of the Rasch Model Items in the Final Semester Exam Indonesian Language Lesson
Author(s) -
Nur ꞌAzizah,
Muchlas Suseno,
Bahrul Hayat
Publication year - 2021
Publication title -
world journal of english language
Language(s) - English
Resource type - Journals
eISSN - 1925-0711
pISSN - 1925-0703
DOI - 10.5430/wjel.v12n1p15
Subject(s) - rasch model , indonesian , test (biology) , reliability (semiconductor) , consistency (knowledge bases) , computer science , documentation , item response theory , mathematics education , subject (documents) , item analysis , quality (philosophy) , psychology , statistics , linguistics , artificial intelligence , psychometrics , mathematics , programming language , world wide web , paleontology , philosophy , power (physics) , physics , epistemology , quantum mechanics , biology
This study aims to analyze and describe the characteristics of the items for the Final Semester Examination for Indonesian Language courses at the PKN STAN official high school. The item analysis in this study was carried out with a modern approach (item response theory) using the Rasch model with Winstep software. This research is a quantitative research with descriptive method. The data was obtained through the documentation method, namely 25 multiple choice questions with 314 respondents. The results of the analysis show that the unidimensional requirements are not met. There are several items in the test that do not meet the local independence requirements, while the monotonicity requirements have been met. The ability of the subject on this test is greater than the level of difficulty of the questions indicated. The reliability of the test is generally satisfactory with the quality of the items and the consistency of the subject's answers are quite good. Persons or subjects who do not fit are 28 people, while items that are not fit are 5 items. There are 11 items indicated bias.