<![CDATA[This quasi-experimental study examined whether Direct Instruction enriched with pedagogical deep learning improves Grade V students’ pencak silat crescent-kick learning outcomes more effectively than ordinary physical education instruction. Sixty-four students from SDN Mankul 1 participated in a pretest–posttest control-group design. The control class received ordinary learning, whereas the experimental class completed three meetings combining explicit demonstration, decomposed guided practice, peer feedback, safe target games, and reflection. Performance was assessed using a six-component rubric converted to a 0–100 scale. Because the data were non-normal, Wilcoxon signed-rank and Mann–Whitney U tests were used, supported by N-Gain and mastery analysis. The experimental group improved from 51.95 to 82.03, with a gain of 30.08 and an N-Gain of 0.63. The control group improved from 52.60 to 60.29, with a gain of 7.68 and an N-Gain of 0.16. Between-group differences in posttest, gain, and N-Gain were significant (p < .001), and all experimental students achieved mastery. The findings show that reflective Direct Instruction can transform imitation-based martial-arts learning into safe, meaningful, and self-regulated motor skill acquisition.]]>
Copyrights © 2026
