Implementation Fidelity and Learning Outcomes of Discovery Learning in Elementary School Science: A Systematic Literature Review of Indonesian Studies

Abstract

Science learning in Indonesian elementary schools is often dominated by lecture-based instruction, leaving many students below minimum mastery criteria. Discovery learning is widely proposed as a remedy, yet the evidence is scattered across research designs, outcomes, and curriculum labels (IPA and IPAS). This study aims to synthesize recent evidence on the implementation and effects of discovery learning in elementary school science learning. A systematic qualitative literature review was conducted on 11 Indonesian journal articles, selected from 26 records identified and published in 2020–2026, comprising four quasi-experiments, three classroom action studies, and four literature syntheses. Data were extracted with a matrix and analysed through thematic content analysis, descriptive comparison of pretest–posttest gains, and narrative synthesis. All sources reported positive effects on cognitive learning outcomes, learning interest, or critical thinking. In the three quasi-experiments reporting pretest–posttest means, the experimental groups’ gains exceeded the control groups’ by 5.87 to 26.11 points, and classroom action studies reached 85–88% mastery by the second cycle. Effects appeared larger where baseline instruction was strongly teacher-centred and weaker where syntax fidelity was low. Discovery learning is therefore effective when its six syntactic phases are implemented with guided worksheets and teacher scaffolding. Practically, teachers should run all six phases with concrete stimuli and sufficient time for presentation and generalization, while teacher training and IPAS implementation should prioritise syntax fidelity and guidance skills.