Investigating the Impact of Artificial Intelligence Tools on Students’ Academic Achievement

Sammanfattning

In a short time, the world has been introduced to AI tools that once were thought fringe tools have shown remarkable capacity for solving generic problems and generating solutions. OpenAI has released Codex which has already been shown to outperform most first-year students on typical programming assignments in introductory courses in programming. This thesis explores the student attitudes toward leveraging AI tools in their academic studies and the usefulness and frequency they use them. The thesis also explores the academic achievements of first-year students at KTH for the introductory programming course for the course offerings of 2021 and 2023 using static analysis tools. The study shows that students from 2023 are more likely to find AI tools more useful on average than in 2021, and they are using AI tools more frequently. It is shown that the students from 2023 achieved a higher code quality in the introductory course in programming compared to the students from 2021 according to the static analysis with a statistical significance (p-value < 0.05).