Analisis Soal Sumatif Mata Pelajaran Pendidikan Jasmani Olahraga dan Kesehatan Kelas VIII Di Sekolah Menengah Pertama Negeri 25 Kota Padang

Abstract

Penelitian ini dilatarbelakangi oleh belum diketahuinya kualitas soal sumatif mata pelajaran Pendidikan Jasmani, Olahraga, dan Kesehatan (PJOK) kelas VIII di SMP Negeri 25 Kota Padang berdasarkan karakteristik butir soal. Penelitian ini bertujuan menganalisis validitas, reliabilitas, tingkat kesukaran, daya pembeda, dan efektivitas pengecoh soal sumatif PJOK. Penelitian menggunakan pendekatan deskriptif kuantitatif dengan sampel 53 peserta didik yang dipilih melalui proportional stratified random sampling dari populasi 272 peserta didik kelas VIII. Data dikumpulkan melalui dokumentasi berupa perangkat soal, kisi-kisi, dan respons jawaban peserta didik. Analisis dilakukan terhadap 20 butir soal pilihan ganda menggunakan korelasi point biserial untuk validitas, Cronbach’s Alpha untuk reliabilitas, indeks tingkat kesukaran, indeks daya pembeda, dan efektivitas pengecoh. Hasil penelitian menunjukkan bahwa seluruh 20 butir soal (100%) dinyatakan valid dan reliabilitas tes berada pada kategori tinggi dengan Cronbach’s Alpha sebesar 0,784. Tingkat kesukaran didominasi kategori mudah sebanyak 17 butir (85%) dan sedang sebanyak 3 butir (15%), tanpa butir berkategori sukar. Daya pembeda terdiri atas 1 butir kategori bagus sekali, 3 butir lumayan bagus, 8 butir belum memuaskan, dan 8 butir jelek. Dari 60 pengecoh, 50 pengecoh (83,33%) efektif dan 10 pengecoh (16,67%) tidak efektif. Disimpulkan bahwa soal sumatif memiliki validitas dan reliabilitas yang baik, tetapi kualitas keseluruhan belum optimal karena tingkat kesukaran kurang proporsional dan sebagian besar butir memiliki daya pembeda rendah, sehingga diperlukan revisi dan analisis butir secara berkala.

This study was motivated by the lack of empirical information regarding the quality of summative assessment items in Physical Education, Sports, and Health (PJOK) for eighth-grade students at SMP Negeri 25 Kota Padang. The study aimed to analyze item validity, test reliability, difficulty level, discrimination power, and distractor effectiveness. A quantitative descriptive approach was employed with a sample of 53 students selected through proportional stratified random sampling from a population of 272 eighth-grade students. Data were collected through documentation, including the test items, test blueprint, and students’ response sheets. Twenty multiple-choice items were analyzed using point-biserial correlation for validity, Cronbach’s Alpha for reliability, the difficulty index, the discrimination index, and distractor effectiveness analysis. The results showed that all 20 items (100%) were valid, while the test reliability was high with a Cronbach’s Alpha coefficient of 0.784. Regarding difficulty, 17 items (85%) were categorized as easy and 3 items (15%) as moderate, with no difficult items. Discrimination analysis identified 1 excellent item, 3 reasonably good items, 8 unsatisfactory items, and 8 poor items. Of the 60 distractors analyzed, 50 (83.33%) were effective and 10 (16.67%) were ineffective. It can be concluded that the test demonstrated good validity and reliability; however, its overall quality was not yet optimal because the difficulty distribution was unbalanced and most items had low discrimination power. Therefore, periodic item analysis and revision are required to improve the quality of future summative assessments.