Inappropriate Benefits and Identification of ChatGPT Misuse in Programming Tests: A Controlled Experiment
Hapnes Toba, Oscar Karnalim, Meliana Christianti Johan, Terutoshi, Tada, Yenni Merlin Djajalaksana, Tristan Vivaldy

TL;DR
This study investigates ChatGPT's potential for aiding programming tests, revealing it significantly speeds up completion time without compromising performance, and explores methods to identify its misuse and student perceptions.
Contribution
The paper provides empirical evidence on ChatGPT's impact on programming test performance and proposes ways to detect its inappropriate use, offering insights into its role as an educational tool.
Findings
ChatGPT enables students to complete programming tasks twice as fast.
Students' programming performance remains comparable with or without ChatGPT.
ChatGPT-generated code is efficient, using complex data structures.
Abstract
While ChatGPT may help students to learn to program, it can be misused to do plagiarism, a breach of academic integrity. Students can ask ChatGPT to complete a programming task, generating a solution from other people's work without proper acknowledgment of the source(s). To help address this new kind of plagiarism, we performed a controlled experiment measuring the inappropriate benefits of using ChatGPT in terms of completion time and programming performance. We also reported how to manually identify programs aided with ChatGPT (via student behavior while using ChatGPT) and student perspective of ChatGPT (via a survey). Seventeen students participated in the experiment. They were asked to complete two programming tests. They were divided into two groups per the test: one group should complete the test without help while the other group should complete it with ChatGPT. Our study shows…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsArtificial Intelligence in Healthcare and Education · Online Learning and Analytics · Topic Modeling
