ANALISIS KUALITAS PREDIKSI SOAL UJIAN NASIONAL HASIL DIKLAT TEKNIS SUBSTANTIF UN GURU MATEMATIKA MTs DITINJAU DARI TINGKAT KESUKARAN, DAYA PEMBEDA DAN KUALITAS PENGECOH THE QUQLITY ANALYSIS OF NATIONAL EXAMINATION PREDICTION TEST AS RESULT OF MADRASAH TSANAWIYAH MATHEMATICS TEACHERS SUBSTANTIVE TECHNICAL NATIONAL EXAMINATION TRAINING PROGRAM AT PRIVINCES BALI, NTB AND NTT
| Main Author: | Sudarsono, Sugeng |
|---|---|
| Format: | Article info application/pdf Journal |
| Bahasa: | eng |
| Terbitan: |
Balai Diklat Keagamaan Denpasar
, 2022
|
| Subjects: | |
| Online Access: |
https://widyadewata.kemenag.go.id/index.php/widyadewata/article/view/39 https://widyadewata.kemenag.go.id/index.php/widyadewata/article/view/39/23 https://widyadewata.kemenag.go.id/index.php/widyadewata/article/view/39/24 |
Daftar Isi:
- ABSTRAK: Penelitian ini bertujuan untuk mengetahui tingkat kesukaran, daya pembeda dan fungsi pengecoh soal Prediksi Soal UN hasil Diklat Teknis Substantif UN Matematika MTs tahun 2019. Metode penelitian yang digunakan adalah analisis deskriptif kualitatif dengan teknik pengumpulan data berupa dokumentasi. Sumber data berupa soal ujian, kunci jawaban dan seluruh lembar jawaban siswa pada uji soal prediksi UN 2019 di MTsN 3 Jembrana pada bulan Maret 2020 Hasil penelitian menunjukkan bahwa untuk tingkat kesukaran didapatkan 4 butir (11 % ) tafsiran sangat mudah, 2 butir (3 % ) dalam tafsiran mudah, 19 butir ( 54 %) tafsiram sedang, 7 butir (20 %) tafsiran sukar, dan 3 butir (9 %) tafsiran sangat sukar. Hasil analisis daya pembeda soal didapatkan 3 butir ( 9 %) memimiki daya beda sangat jelek, 4 butir ( 11 %) memiliki daya beda jelek, 3 butir (9 %) memiliki daya beda cukup, 12 butir ( 34%) memiliki daya beda baik dan 13 butir (37 %) katagori Baik sekali. Hasil analisis fungsi pengecoh menunjukkan bahwa dari 105 distraktor, ada 41 pengecoh (39 %) katagori harus direvisi atau diperbaiki, 64 pengecoh (61 %) dalam katagori efektif. Secara keseluruhan soal Prediksi Soal UN hasil Diklat Teknis Substantif UN Matematika MTs tahun 2019 dikategorikan sebagai soal yang baik, sebab memiliki persentase tingkat kesukaran terbesar pada kategori soal sedang yaitu sebanyak 19 butir (54%), persentase daya pembeda terbesar pada kategori baik dan sangat baik yaitu sebanyak 25 butir soal (71 %) serta lebih banyak pengecoh yang telah berfungsi dengan baik yaitu sebanyak 64 pengecoh (61 %). Kata-kata kunci : tingkat kesukaran, daya pembeda, fungsi pengecoh , prediksi soal UN ABSTRACT: This study aims to determine the level of difficulty, distinguishing features and deception functions about the UN Test Prediction Result of the MTs Substantive Technical Training for Mathematics in 2019. The research method used is descriptive qualitative analysis with data collection techniques in the form of documentation. The data source is in the form of exam questions, answer keys and all student answer sheets in the 2019 UN prediction test in MTsN 3 Jembrana at March 2021. The results showed that for the difficulty level there were 4 items (11%) very easy interpretation, 2 items (3%) in easy interpretation, 19 items (54%) moderate interpretation, 7 items (20%) difficult interpretation, and 3 items ( 9%) interpretation is very difficult. The results of the analysis of the distinguishing power obtained 3 items (9%) had very bad power differences, 4 items (11%) had bad power differences, 3 items (9%) had sufficiently different power, 12 items (34%) had good power difference and 13 items (37%) categorized Very well. The results of the analysis of the deception function showed that of the 105 distractors, there were 41 deception (39%) categories that had to be revised or improved, 64 deception (61%) in the effective category. Overall, the Prediction Questions for the National Examination of the National Mathematics and Natural Sciences Sub-National Mathematics MTs in 2019 were categorized as good problems, because they had the highest percentage of difficulty in the medium problem category of 19 items (54%), the biggest percentage of differentiation in the good and very good categories namely as many as 25 items (71%) as well as more deceivers that have worked well, as many as 64 deception (61%). Key words: difficulty level, distinguishing features, deception function, prediction of UN questions