Analisis Butir Soal Pilihan Ganda dengan Pendekatan Classical Test Theory pada Ujian Akhir Semester Mata Pelajaran Matematika

The purpose of this research is to analyze question items with a classical test theory approach. The type of research is descriptive quantitative. The object of this research is multiple choice items used in the end of semester examination of mathematics subjects in junior high school. The research subjects were students who took the test. The number of respondents was 50 people. The research instrument was 30 multiple choice questions. The data were analyzed using the Classical Test Theory approach with the help of RStudio analysis software. The results of the analysis of the level of difficulty of 30 items, it is known that as many as 28 items (93.33) are difficult categories. There are 2 items (6.67%) in the medium category. There is no single question that is classified as easy. Differentiability as many as 7 items (23.3%) in the category of acceptable with improvement. A total of 12 items (40%) in the corrected category. Meanwhile, 11 items (36.7%) were in the discarded category. All options A to E on each item have been selected, so the distractor function works well. The reliability of each item is in the range between 0.70 to 0.71 in the high reliability category. Overall, the reliability of the test instrument is 0.713, which is also in the high category.