<![CDATA[This study aimed to examine whether the Think-Pair-Square (TPSq) cooperative learning technique could improve the learning process and mathematics achievement of Grade 8 students. Classroom action research was conducted with 32 students, comprising 12 boys and 20 girls, at SMP Islam YLPI Pekanbaru, Indonesia. The intervention was implemented in two cycles, each consisting of planning, action, observation, and reflection, with three instructional meetings and one achievement test per cycle. Students participated in eight heterogeneous groups through individual thinking, paired discussion, and four-member comparison and consolidation. Data were collected using open observation sheets and essay-based achievement tests and analyzed descriptively through activity patterns, individual development scores, group recognition, minimum mastery attainment, indicator-level mastery, mean scores, and score distributions. The results showed improved implementation fidelity and student participation across both cycles. Classical mastery increased from 62.50% at baseline to 68.75% after Cycle I and 81.25% after Cycle II, while mean achievement increased from 78.50 to 79.13 and 83.16, respectively. Five of eight groups achieved the highest recognition category in Cycle II, compared with none in Cycle I. Students demonstrated strong mastery of prism and pyramid concepts, surface area, and volume, although only 50.00% mastered proportional changes in volume following changes in edge length. These findings indicate that TPSq strengthens mathematics learning through individual accountability, peer explanation, and collaborative consolidation. However, the single-class design limits generalizability. Future studies should involve broader samples and examine targeted scaffolding for proportional reasoning.]]>
