<![CDATA[This study examined the effect of Kahoot-assisted Discovery Learning on eighth-grade students’ science learning outcomes in human respiratory system material. It extends earlier local pre-experimental evidence by employing a nonequivalent control group design, incorporating a direct-instruction comparison class, and using ANCOVA to account for baseline achievement. The study was conducted at SMPN 16 Kupang in April 2026 with 56 students selected through purposive sampling. Class VIII E (n = 28) received Discovery Learning integrated with Kahoot as an immediate formative-feedback and concept-verification tool, whereas class VIII F (n = 28) received direct instruction. Learning outcomes were measured using a 20-item multiple-choice test administered before and after instruction; expert content validation of the test yielded 72/80 (90.00%), supporting its content suitability for use. Student activity and teacher learning management were observed to provide process evidence supporting the achievement data. Descriptive analyses included mean scores, classical mastery, and normalized gain, while inferential analyses included normality, homogeneity, and ANCOVA tests using SPSS 26 at a significance level of 0.05. The experimental class improved from a pretest mean of 51.25 to a posttest mean of 80.00, with an N-Gain of 0.60 and classical mastery of 85.71%. The control class improved from 49.29 to 66.25, with an N-Gain of 0.34 and mastery of 46.43%. ANCOVA yielded F = 11.615 and p = 0.001, indicating a significant difference in posttest achievement after accounting for baseline scores. Student activity was also higher in the experimental class (79.48%) than in the control class (74.60%), while teacher learning management was rated very good in both groups. These findings indicate that Kahoot contributes most meaningfully when used not merely as a game platform, but as immediate formative feedback and concept verification within a structured Discovery Learning cycle.]]>
