Analisis Butir Soal Penilaian Tengah Semester Genap Mata Pelajaran Pendidikan Agama Islam Kelas V di SDN 1 Bumiharjo

Abstract

This study aims to determine the quality of the items about the Even Midterm Assessment of Islamic Religious Education class V subjects at SDN 1 Bumiharjo. This study used evaluative analysis methods. Data collection techniques use documentation. The data analysis techniques used are calculating empirical validity, reliability, item difficulty, item differentiating power, and distractor functions in multiple choice question types through counter software, namely using ANATES software version 4.0 accompanied by data presentation and drawing conclusions. The results showed that in terms of validity, overall the question already has a good rational validity, this is based on the results of a high content and construct validity analysis. Meanwhile, empirical validity shows that as many as 2 items of multiple-choice questions or 12.50% are declared valid, fill-in-the-blank questions as many as 9 questions or 75% are declared valid, and 5-point description questions or 83.33% are declared valid from all questions, both multiple choice, fill in and description. From the reliability of the questions, the form of the fill-in-the-blank questions includes a high reliability coefficient of 0.78 and for multiple-choice questions and descriptions are still relatively poor, the reliability coefficients are still low, namely 0.53 and 0.63. In terms of the difficulty of the questions, the multiple-choice type questions are still relatively not good, because the questions are still relatively too easy, namely as many as 10 (55.56%) multiple-choice questions. Then for the fill form questions, it is quite good, because the questions stated are sufficient as many as 10 (83.33%) of the total fill questions. The question of the type of description is quite good, because 3 (50%) of the questions from the fill-in-the-blank questions are classified as having a sufficient level of difficulty. from the distinguishing power of the items, in the poor category multiple choice type questions there are 3 (18.75%) question items, satisfactory there are 9 (50%) question items, good there are 4 (22.22%) question items, and excellent there are 2 (9.03%) question items from the entire multiple choice type questions. Then for the fill type question, the poor category has only 1 (8.34%) questions, satisfactory there are 4 (33.33%) question items, and good there are 7 (58.33%) question items from all items about the fill type. Then for the question of the type of long description, the poor category has 2 (33.33%) question items, satisfactory there are 2 (33.33%) question items, and good there are 2 (33.33%) question items from the entire description type question. From the distractor function, the multiple choice type question has a fairly good distrktor function, because based on the analysis, it can be concluded that there are 11 questions or 61.11% of the distractor function is able to perform its function properly