<![CDATA[Students now use generative artificial intelligence to complete coursework, yet the link between how they judge its usefulness and the grades they earn has rarely been tested in courses that demand methodological reasoning. This study analysed the relationship between perceived usefulness of artificial intelligence, framed by cognitive flexibility theory, and learning outcomes in a Research Methodology course. This correlational survey involved 52 students of a mechanical engineering education study program at a public university in Jakarta, selected purposively from a population of 109. Perception was measured with a 26-item Likert-scale questionnaire that met validity and reliability criteria (Cronbach's alpha 0.952), and learning outcomes were taken from final course grades. Eighty-three percent of respondents used artificial intelligence while taking the course, and ChatGPT was the most frequently used tool (63%). The Pearson Product Moment correlation was r = 0.403 with a significance of 0.003, indicating a moderate positive relationship. Perceived usefulness shared about 16% of the variance in final grades. Students who reported that artificial intelligence helped them adjust their thinking strategies tended to earn higher grades. Lecturers can use these findings to design assignments that direct the use of artificial intelligence towards source evaluation and the construction of research arguments]]>
Copyrights © 2026
