OpenAI in Mathematics Education and the Critical Thinking Skills of Early Childhood Preservice Teachers as Predictors of Mathematics Performance

This study examined early childhood preservice teachers’ (PSTs) critical thinking skills (CTS) and mathematics performance (MP) and how they perceived and experienced mathematics learning after the integration of OpenAI tools such as ChatGPT, focusing on CTS as predictors of MP in twenty-first-century competencies aligned with SDG 4 (Quality Education). Using an explanatory sequential mixed-methods approach, the researchers collected data through quantitative measures of CTS and MP, as well as through focus group discussions. Before AI integration, participants reported low engagement, limited mathematical background, and reliance on rote learning, which corresponded to low CTS scores (Challenged to Practicing) and a mean MP of 74% (Failed). Following AI integration, participants (n = 66) demonstrated significant improvements, with CTS scores rising to the Advanced level and MP increasing to 90% (Very Good, p < 0.001). Qualitative findings highlighted enhanced conceptual understanding, reflective reasoning, and confidence in problem-solving. Correlational and regression analyses confirmed CTS as a strong predictor of MP. The results suggest that when generative AI tools are employed as instructional supports rather than as answer-providing shortcuts, they can enhance both measurable academic outcomes and twenty-first-century skills. These findings underscore the potential of AI to support effective mathematics learning while emphasizing the importance of responsible use and guided pedagogical implementation.