Analyzing the Perceptions and Interactions of Students with Generative AI in Engineering Courses: A Preliminary Study and Insights

Abstract

This study analyzes students’ perceptions and interactions, employing Generative AI (GenAI) in 16-week university engineering courses. An AI educational framework called PRIOR (Plan, Review, Integrate, Observe, and Reflect) was deployed along with several activities conducted in the courses using GenAI tools in a controlled way during the semesters. Then, two instruments to collect data from students regarding perceptions and interactions with GenAI were created: A Likert scale survey with 27 closed-ended questions and two open-ended ones, and a matrix to identify both the type of questions for the GenAI tools and how students comprehend the information provided by AI prompts. The outcomes indicate that students primarily use GenAI for concept explanation, synthesizing information, and generating ideas for problem-solving. Nonetheless, for help-seeking, students prefer teachers’ support, and they indicate issues with GenAI, such as over-reliance, hallucinations, lack of trustworthiness, and decreased critical thinking. Finally, the analysis of students’ questions with GenAI suggests that it is necessary to improve prompt engineering, AI literacy, and engagement in GenAI-based activities.

Ähnliche Arbeiten