Pengembangan Perangkat RBL-STEM untuk Meningkatkan Keterampilan Berpikir Kombinatorial Mahasiswa Menyelesaikan Masalah Pewarnaan Pelangi Anti Ajaib dan Aplikasinya pada Distribusi BBM Bersubsidi dengan Teknik Graph Neural Network

Abstract

Students often face difficulties in solving complex mathematical problems in real-world contexts due to their low thinking abilities. To enhance students' thinking skills, effective learning approaches, such as the RBL-STEM model, which offers research-based and practically applicable learning in real-world contexts, are necessary. This study aims to investigate the activities of RBL-STEM learning, describe the process and outcomes of developing RBL-STEM tools, and analyze the data from these tools to improve students' combinatorial thinking skills. The study utilizes a mixed-methods approach with an explanatory sequential design that incorporates both quantitative and qualitative techniques. The development of RBL-STEM tools includes the design of student tasks, worksheets, and combinatorial thinking skill tests. The results of tool development indicate good validity. A trial was conducted with 23 students, and the implementation of RBL-STEM learning using these tools was considered practical and effective, with a 95% implementation rate. Furthermore, the students were highly engaged and provided very positive responses to the learning experience. Pretest and posttest analyses revealed an improvement in students' combinatorial thinking skills in solving the problem of magical rainbow coloring. The study also identified three levels of combinatorial thinking skills: high, medium, and low. Statistical analysis and phase portrait analysis confirmed the research findings, demonstrating an enhancement in students' combinatorial thinking skills. Thus, RBL-STEM learning has the potential to enhance students' combinatorial thinking skills in real-world contexts, such as in the application of subsidized fuel distribution using graph neural network techniques.