<![CDATA[Mathematical word-problem solving remains a persistent challenge for deaf students, largely due to limited access to linguistic input and the abstract nature of mathematical reasoning. This study examines the effectiveness of the Aptitude Treatment Interaction (ATI) instructional model as an adaptive mechanism for improving mathematical problem-solving performance among seventh-grade deaf students. Employing a quantitative quasi-experimental one-group pretest–posttest design, the study involved ten students enrolled at a special junior high school in Indonesia. Students were classified into high, medium, and low aptitude groups based on an initial diagnostic assessment and received differentiated instructional treatments aligned with ATI principles. Data were collected through a validated mathematical word-problem test, classroom observations, and response questionnaires. Due to non-normal data distribution, the Wilcoxon Signed Rank Test was applied, accompanied by effect size estimation. The results demonstrate a statistically significant improvement in students’ problem-solving performance following ATI implementation, with mean scores increasing from 56.33 to 82.00 (p < 0.05) and a large effect size, indicating substantial practical impact. Beyond instructional effectiveness, the findings highlight ATI’s capacity to operationalize differentiated instruction by aligning learner aptitude with targeted pedagogical treatment. This study contributes empirical evidence to the global discourse on adaptive mathematics instruction and offers theoretical implications for inclusive learning models addressing heterogeneous cognitive and linguistic profiles in mathematics education.]]>
