Learning Assessment Model for the Society 5.0: Integration of AI and Teacher-Based Affective Assessment through Sipaq' Mandar Values

Abstract

The transition into Society 5.0 necessitates educational assessment models that are not only technologically advanced but also culturally meaningful. This study developed and tested an integrated learning assessment model for the Society 5.0, synergizing Artificial Intelligence (AI) for cognitive-psychomotor evaluation with teacher-based affective assessment grounded in Sipaq’ Mandar local values. Employing an explanatory sequential mixed -methods design, the research involved 300 junior high school students and 60 teachers in Majene Regency, West Sulawesi. Quantitative data were collected through pre-post tests and questionnaires, while qualitative data were gathered via interviews, FGDs, and observation. Result demonstrated a statistically significant improvement in cognitive-psychomotor achievement, with an average score increase of 11.7 points (p < .001, Cohen’s d = 1.12). The Sipaq’ Mandar values (siri’, siarondoi, sipakalaq’bi, makkeada, dipokannyang) were successfully operationalized into a valid and reliable affective assessment rubric. The resulting SIAMASEI-AI model was deemed highly feasible (M=4.43/5) and contributed 71.2% to holistic assessment affectiveness. The study concludes that the integration of AI with culturally contextual affective assessment offers a viable, balanced, and humanized evaluation framework, addressing both the technological demands and the character-building imperative of modern education.