Pengembangan Model Pembangkit Jalur Belajar Remedial Terpersonalisasi Berbasis Pembelajaran Perancah

Wahyuningsih, Yulia (2024) Pengembangan Model Pembangkit Jalur Belajar Remedial Terpersonalisasi Berbasis Pembelajaran Perancah. Doctoral thesis, Institut Sepuluh Nopember.

Text 05111960010002-Dissertation(2).pdf - Accepted Version Restricted to Repository staff only until 1 July 2026. Download (6MB) | Request a copy

Abstract

Idealnya, seorang siswa yang telah mempelajari seluruh konsep dalam satu topik suatu mata pelajaran di kelas akan mampu menjawab pertanyaan dengan menerapkan beberapa konsep yang telah dipelajarinya. Sayangnya, tidak semua siswa dapat memenuhi kondisi ideal tersebut, sehingga tes remedial biasanya dilakukan pada siswa tertentu yang gagal mengkorelasikan konsep untuk menyelesaikan soal tes, terutama pada mata pelajaran yang tingkat hubungan antar konsepnya relatif tinggi. Penelitian ini mengusulkan model baru untuk menghasilkan jalur pembelajaran remedial yang dipersonalisasi berdasarkan metode pembelajaran scaffolding yang disebut Personalized Remedial Learning Path (PRLP). Model yang diusulkan melibatkan tiga proses utama dalam pengembangannya. Yang pertama adalah proses identifikasi konsep setiap soal tes menggunakan pengklasifikasi multi-label berdasarkan BiDirectional Encoder Representations from Transformers (BERT) untuk Bahasa yang disebut dengan Soal indoBERT (QuiB). Proses kedua adalah otomatisasi pembangkitan bobot CER untuk setiap konsep yang disebut proses Mastery. Kemudian yang terakhir adalah proses Scaffolding untuk menghasilkan jalur pembelajaran remedial yang dipersonalisasi berdasarkan metode pembelajaran scaffolding. Hubungan antar konsep sebagai Concept Effect Relationship (CER) direpresentasikan dalam grafik berarah berbobot, dimana nilai bobot setiap busur pada grafik menyatakan frekuensi kemunculan konsep secara bersamaan antar soal tes yang digunakan dalam tes formatif. Evaluasi model usulan dilakukan pada materi dinamika gerak lurus Fisika terhadap 102 siswa dari beberapa SMA yang gagal memperoleh nilai ketuntasan belajar minimal pada hasil tes formatif. Hasil evaluasi menunjukkan bahwa jalur pembelajaran remedial personal yang ditetapkan dapat meningkatkan nilai tes remedial siswa dan persentase siswa yang mencapai ketuntasan belajar minimal. Rata-rata nilai tes sebesar 65,27 (standar deviasi 11,59), dan 47,06% siswa yang diberikan PRLP (dari total 51 siswa) berhasil mencapai ketuntasan belajar minimal. Sebaliknya pada kelompok siswa yang tidak menggunakan PRLP (dari total 51 siswa) hanya memperoleh rata-rata nilai tes sebesar 46,08 (standar deviasi 13,84), dan hanya 3,90% siswa yang berhasil mencapai ketuntasan belajar minimal. skor
====================================================================================================================================
Ideally, a student who has studied all the concepts in one topic of a subject in class will be able to answer questions by applying several concepts he has learned. Unfortunately, not all students can meet these ideal conditions, so remedial tests are usually conducted for particular students who fail to correlate concepts to solve the test items, especially in a subject whose level of relationship between concepts is relatively high. This research proposes a new model to generate a personalized remedial learning pathway based on a scaffolding learning method called the Personalized Remedial Learning Path (PRLP). The proposed model involves three main processes in its development. The first is the concept identification process for each test item using a multi-label classifier based on Bidirectional Encoder Representations from Transformers (BERT) for Bahasa called the Question indoBERT (QuiB). The second process is the automation of generating CER weightings for each concept called the Mastery process. Then the last one is the Scaffolding process to generate personalized remedial learning paths based on the scaffolding learning method. The relationship between concepts as a Concept Effect Relationship (CER) is represented in a weighted directed graph, where the weight value for each arc in the graph states the frequency of concept co-occurrence among the test items used in the formative test. The evaluation of the proposed model is carried out for the topic of dynamics of rectilinear motion in Physics on 102 students from several senior high schools who failed to obtain the minimum learning mastery score on formative test results. The evaluation results show that the established personalized remedial learning path can leverage both students' remedial test scores and the percentage of students who achieved the minimum learning mastery. The average test score was 65.27 (standard deviation of 11.59), and 47.06% of the students provided with the PRLP (from a total of 51 students) successfully achieved the minimum learning mastery. In contrast, the group of students who did not use the PRLP (from a total of 51 students) only obtained an average test score of 46.08 (standard deviation of 13.84), and only 3.90% of students who managed to achieve the minimum learning mastery score

Item Type:	Thesis (Doctoral)
Uncontrolled Keywords:	pembelajaran perancah, jalur belajar terpersonalisasi, peta konsep, tes formatif, tes remedial; scaffolded learning, personalized learning paths, concept maps, tests formative, remedial tests
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.62 Decision support systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55001-(S3) PhD Thesis (Comp Science)
Depositing User:	Yulia Wahyuningsih
Date Deposited:	23 Feb 2024 08:49
Last Modified:	23 Feb 2024 08:49
URI:	http://repository.its.ac.id/id/eprint/107630

Actions (login required)

View Item