FUZZY LOGIC APPROACH FOR PREDICTING STUDENT ACHIEVEMENT IN SCRATCH

Abstract

21st-century skills such as critical thinking, problem-solving, and analytical thinking gained importance to survive in today’s world. There is growing research mostly focus on the prediction of students in higher education using machine learning and statistical models. However, predicting primary and middle school student’s performance also becomes important especially in learning computer programming. In this study, it was primarily proposed to a fuzzy logic system to predict student performance during the experiment then compare fuzzy logic prediction results to the experts’ results. Secondly, to test the theory that students’ interest in learning algorithms and coding can be increased using the creation of games in a visual programming tool for beginners. The fuzzy logic inference system has been employed to predict middle school student’s performance in the programming experiment which has been carried out using the Scratch environment with the participation of three different middle school students in Turkey. The success rate of three different middle school group success rates is estimated regarding task completion times, and the regression results with respect to the groups are %80, %97, %84.