An Exploration of Generative AI-Resistant Computational Thinking: Remote Assessment Strategies
- Duncan A. Coulter(author)
Export Metadata
| Title | An Exploration of Generative AI-Resistant Computational Thinking |
|---|---|
| Subtitle | Remote Assessment Strategies |
| DOI | https://doi.org/10.64449/9781997499978-12 |
| Landing page | https://ujonlinepress.uj.ac.za/index.php/ujp/catalog/book/408 |
| Publisher | UJ Press |
| Published on | 2026-09-15 |
| Long abstract | The rise of generative artificial intelligence systems capable of performing basic programming tasks has negatively impacted the ability of such competencies to be assessed remotely. While this may lead to a return to in-person assessment for such competencies, as is the case in other numerical sciences, there are still use-cases where remote assessment is preferable, such as in the case of screening applicants to large courses. While the situation can be dealt with as a generalisation of the pre-existing problem of establishing genuine authorship via online proctoring tools, such methods have severe implementation constraints in terms of user acceptance, technical issues, and privacy concerns. Consequently, the need exists to explore other possible assessment strategies that make use of the asymmetric imbalance in performance across differing assessment modalities inherent in the current generation of generative artificial intelligence (AI). |
| Page range | pp. 325-361 |
