The development of RBL - STEM learning materials to improve students' computational thinking skills in solving convolutional neural network problems

Abstract

In 21st century learning, computational thinking skills have become one of the essential competencies that need to be emphasised in the development of knowledge. To enhance computational thinking skills, research-based learning (RBL) with a science, technology, engineering and mathematics (STEM) approach, known as RBL-STEM, can be used. This study aims to explore RBL-STEM activities, describe the process and outcomes of developing RBL-STEM materials, and analyse data. In this research, the RBL-STEM framework is used to improve students' computational thinking skills in applying Convolutional Neural Network (CNN) to identify coffee plant diseases using a quadcopter drone and its flight path with resolving dominating set. The research method used is Research and Development (R&D). This research develops RBL-STEM materials and produces learning material products in the form of semester study plans, student assignment designs, student worksheets and learning outcome tests. The results of the development of the materials show validity with a validity criterion of 92%. Implementation using RBL-STEM materials was found to be practical with a practicality criterion of 96.25% and effective with an effectiveness criterion of 94.32%. In addition, students were highly engaged and provided very positive feedback on the learning experience. Pre-test and post-test analysis showed an improvement in students' computational thinking skills when solving CNN problems.