<![CDATA[This study aims to understand the metacognition processes of students in solving math problems with the assistance of Chat-GPT using a qualitative method and a case study approach. The research subjects are two eighth-grade students from a junior high school in Surabaya, selected through purposive sampling, with one subject representing successful students and the other representing unsuccessful students in the initial problem-solving test (PST). Data were collected through the initial PST, Chat-GPT-assisted PST, and interviews, then analyzed using metacognitive indicators: planning, monitoring, and evaluation. The successful subject actively engaged with Chat-GPT's answers, using metacognition processes to rethink their understanding, connect with prior knowledge, consider strategies, check steps, adjust errors, and verify solutions. Their errors stemmed from misunderstandings and incorrect information from Chat-GPT, but they adjusted and corrected their work upon prompt. In contrast, the unsuccessful subject only showed metacognitive engagement at the planning stage, struggling to understand the problem and prerequisite algebra material. They relied solely on Chat-GPT's answers without understanding or evaluating them, leading to ineffective problem-solving. The difference in metacognition processes between the subjects was attributed to cognitive factors, such as information received from Chat-GPT's answers, understanding of prerequisite material, and ability to comprehend the mathematical problem]]>
Copyrights © 2024
