This study aims to describe the implication of the Aptitude Treatment Interaction (ATI) model integrated with character values to increase the students’ skill in solving mathematics story problems. This study applied a quasi-experimental research type using a non-equivalent control group design involving two classes with 30 students each. Data was collected using a test instrument for solving mathematics story problem. Data were analyzed using n-gain descriptive statistical analysis to see the increase in students' skill in solving mathematics story world problems. The results showed that the average score of student's aptitude in solving mathematics story problems is 91.26 which is in the category of very high. There is an increase in the students’ ability with score of an n-gain of 0.77 which is in the category of high. In addition, the results of observations related to the implementation of learning model of the ATI with a percentage of 87.5% in the category of very good. Thus, the character-based ATI learning model can be used to increase the students’ skill in solving mathematics story problem. In addition, it accommodates the character of students who are concerned with learning mathematics so that learning goals can be achieved both from cognitive and attitudinal aspects.

This study investigated the integration of artificial intelligence (AI) tools into secondary school chemistry education in Zimbabwe, assessing their impact on student engagement and academic performance. Grounded in Vygotsky’s Sociocultural Theory and Cognitive Load Theory, the research employed a mixed-methods approach within a pragmatic framework. Quantitative data were collected through pre-test and post-test assessments and structured surveys, comparing an experimental group using AI tools with a control group employing traditional methods. Qualitative data from student and teacher interviews and classroom observations were analysed thematically. ANCOVA analysis revealed a statistically significant difference in post-test scores between the experimental and control groups, F (1, 117) = 188.86, p < .005, η² = 0.617, demonstrating a large effect size of AI integration on academic performance. Students in the experimental group exhibited a mean improvement of 20%, controlling for pre-test differences. Additionally, interaction effects between AI use and gender (F (1,115) = 0.17, p = .684) as well as prior chemistry knowledge (F (1,115) = 0.05, p = .829) were not statistically significant. Furthermore, 85% of the experimental group reported higher engagement levels, confirming AI’s role in fostering motivation and conceptual understanding. AI tools facilitated personalized learning paths, interactive simulations, and real-time feedback, optimizing cognitive efficiency and deep learning. Despite these advantages, significant challenges emerged, including limited internet access, insufficient technological resources, lack of teacher training, and curriculum integration difficulties. These barriers highlight the need for strategic investments in digital infrastructure, professional development for educators, and curriculum revisions to fully integrate AI into chemistry education. The findings underscore AI’s transformative potential in STEM education within developing nations. Addressing infrastructural and pedagogical challenges is critical to maximizing AI's impact, ensuring equitable access, and fostering long-term sustainability in educational innovation.