Deep Learning Approach to Improve Fourth-Grade Students’ Interest in Mathematics

Low interest in mathematics can restrict students’ participation, persistence, and conceptual learning. This study aimed to describe the planning and implementation of a deep learning approach and to determine changes in fourth-grade students’ interest in mathematics after the approach was applied. Classroom action research was conducted in one fourth-grade class at SD Negeri Karanganom 3, Cirebon, Indonesia, using a pre-cycle phase and two action cycles based on planning, action, observation, and reflection. The cycle-level dataset comprised 24 students. Data were collected through structured observations of lesson-plan quality, teacher performance, affective activity, and learning interest, together with a five-item written test. Learning interest was observed through enjoyment, involvement, attention, and attraction. Data were analyzed using frequencies, means, and percentages against a 75% success criterion. Lesson-plan quality increased from 59.26% in the pre-cycle to 92.60% in Cycle I and 100% in Cycle II, while teacher performance rose from 61.90% to 80.95% and 100%. The proportion of students classified as good or very good in learning interest increased from 37.50% to 62.50% and 87.50%. Mathematics mastery also increased from 20.83% to 62.50% and 91.67%, with mean scores of 52.08, 68.96, and 87.50. The findings indicate that iterative planning and implementation of mindful, meaningful, and joyful learning activities were accompanied by substantial improvements in mathematics interest and classroom learning outcomes.