The Effect of Joyful Learning Based on Deep Learning on Motivation and Interest in Learning Natural and Social Sciences Among Fifth-Grade Students

Abstract

This study examined the implementation of Joyful Learning based on the Deep Learning approach and its effect on fifth-grade students' motivation and interest in learning Natural and Social Sciences (IPAS) at SD Inpres Sanging-Sanging, Gowa Regency, Indonesia. The study used a quantitative design involving all 48 fifth-grade students as respondents through total sampling. Data were collected using questionnaires and classroom observation, while data analysis was conducted using validity testing, reliability testing, normality testing, descriptive analysis, and regression analysis with SPSS 26. The classroom observation showed that the implementation of Joyful Learning was generally categorized as good, although problem-solving activities still required pedagogical strengthening. The instruments were valid and reliable, with Cronbach's alpha values of 0.777 for Joyful Learning, 0.899 for learning motivation, and 0.866 for learning interest. The SPSS coefficient tables showed a positive but statistically non-significant effect of Joyful Learning on learning motivation (B = 0.395; beta = 0.267; t = 1.875; p = 0.067) and learning interest (B = 0.062; beta = 0.057; t = 0.389; p = 0.699). These findings indicate that Joyful Learning has pedagogical potential for creating an enjoyable and student-centered classroom atmosphere, but its statistical effect in this sample was limited. The study recommends strengthening teacher creativity, classroom interaction, and problem-solving facilitation in implementing Joyful Learning in primary education.