KUALITAS SOAL UJIAN DEPARTEMEN PENDIDIKAN BAHASA DAERAH FPBS UPI TAHUN AKADEMIK 2016/2017

Abstract

Penelitian  ini dilatarbelakangi dari belum disusunnya tolak ukur penyusunan alat evaluasi mata kuliah berdasarkan uji kelayakan soal-soal ujian di Departemen Pendidikan Bahasa Daerah FPBS UPI. Tujuan penelitian ini adalah untuk mendeskripsikan kualitas soal ujian mata kuliah Departemen Pendidikan Bahasa Daerah FPBS UPI tahun akademik 2016/2017 di yang dikaji berdasarkan uji validitas, reliabilitas, tingkat kesukaran, dan daya beda sebagai dasar penentuan kelayakan soal-soal ujian. Penelitian ini didesain menggunakan pendekatan kuantitatif dengan metode deskriptif-analisis. Sumber data diabil secara purposive pada soal-soal mata kuliah kategori pembelajaran, kebahasaan, kesastraan, dan kebudayaan yang menggunakan jenis soal pilihan ganda dengan jumlah 160 item soal. Teknik yang digunakan dalam mengumpulkan data yaitu teknik studi dokumentasi. Teknik pengolahan data dilakukan dengan cara kompilasi, tabulasi, skoring, dan interpretasi. Hasil dari penelitian ini adalah, 1) tingkat validitas soal yang diuji sebanyak 59% valid serta 41% sisanya tidak valid; 2) tingkat reliabilitas soal yang diuji berada dalam kategori sedang; 3) tingkat kesukaran soal 58% sedang, 20% mudah, 14% sukar, 7% sangat mudah, dan 2% sisanya sangat sukar; 4) daya pembeda soal 43% cukup baik, 39% buruk, 10% baik, 6% sangat buruk, dan 2 % sisanya sangat baik; serta 5) tingkat kelayakan soal ujian yang dianalisis yakni 67% layak, 12% harus direvisi, dan 21% harus diganti.

Abstract

This research was motivated by the fact that the benchmarking for subjects evaluation device has not been compiled based on test item feasibility in the Sundanese Education Department of FPBS UPI. The purpose of this study was to describe the quality of test items in the Department of Sundanese Language Education FPBS UPI in 2016/2017 academic year, which were examined based on tests of validity, reliability, level of difficulty, and power of the different on the basis for determining the feasibility of exam question items. This study employed a quantitative approach with descriptive-analysis methods. Data sources were chosen purposively on test items of learning, language, literature, and culture subjects’ categories and used multiple-choice question type consisted of 160 exam question items. The technique used in collecting the data was document study technique. Data processing techniques were done by compiling, tabulating, scoring, and interpreting. The results of this study were:1) questions validity is 59% and the remaining 41% is not valid; 2) the questions reliability is mediocre; 3) question level of difficulty are distributed as it follows: 58% is moderate, 20% is easy, 14% is difficult, 7% is very easy, and the remaining 2% is very difficult; 4) the differentiating question 43% is good enough, 39% is poor, 10% is good, 6% is very poor, and the remaining 2% is very good; and 5) the question items feasibility split in 67% is feasible, 12% must be revised, and 21% must be replaced.