Enhancing Mathematics Comprehension: A Decision Tree Analysis Using Orange Data Mining

Comprehension skills are essential in mathematics learning, as students' understanding is influenced by various internal and external factors. Recognizing these factors is crucial for educators to design effective teaching strategies. This study aims to classify and predict students' mathematical comprehension based on gender, attitude, learning styles, and self-confidence. A total of 53 eleventh-grade students from SMA Negeri 8 Ternate participated. Primary data were analyzed using data mining techniques—specifically, classification and prediction using the Decision Tree method via Orange Data Mining software. The analysis identified learning style as the most influential factor in students’ mathematical comprehension. The Decision Tree's root node represented comprehension data from 31 students, of which 19 students (61.3%) were classified as having understood the material. The internal node revealed two branches: students with an auditory learning style (8 students) showed a 100% understanding rate, whereas students with kinesthetic or visual styles (11 students) demonstrated a 47.8% understanding rate. The model's prediction accuracy based on the four attributes was 65%. Findings highlight the significance of tailoring instruction to students' learning styles. The relationship between visual, auditory, and kinesthetic learning preferences—when considered alongside gender, attitude, and self-confidence—can offer valuable insights into learning patterns. This study provides a practical reference for educators in developing effective and personalized teaching methods. By leveraging insights into learning styles and associated factors, instructional approaches can be optimized for improved mathematical comprehension.