Taking the help or going alone: ChatGPT and class assignments

Brian Hill ()
Additional contact information
Brian Hill: HEC Paris - Recherche - Hors Laboratoire - HEC Paris - Ecole des Hautes Etudes Commerciales, CNRS - Centre National de la Recherche Scientifique, GREGH - Groupement de Recherche et d'Etudes en Gestion à HEC - HEC Paris - Ecole des Hautes Etudes Commerciales - CNRS - Centre National de la Recherche Scientifique

Working Papers from HAL

Abstract: There is increasing speculation about the future role of ChatGPT and other artificial intelligence (AI) chatbots aiding humans in a variety of tasks. But do people do better when aided by these tools, as compared to when they complete tasks on their own? Can they properly evaluate and where necessary correct the responses provided by ChatGPT to enhance their performance? To investigate this question, this study gives university-level students class assignments involving both answering questions and correcting answers provided by ChatGPT. It finds a significant reduction in student performance when correcting a provided response as compared to when they produce an answer from scratch. One possible explanation for this discrepancy could be the confirmation bias. Beyond emphasising the need for continued research into human interaction with AI chatbots, this study exemplifies one potential way of bringing them into classroom: to raise awareness of the pitfalls of their improper use.

Keywords: ChatGPT; Human-AI chatbot interaction; Confirmation bias; Class assignments; AI in education; Future of work; ChatGPT Human-AI chatbot interaction Confirmation bias Class assignments AI in education Future of work (search for similar items in EconPapers)
Date: 2023-03-31
Note: View the original document on HAL open archive server: https://hal.science/hal-04381138
References: Add references at CitEc
Citations:

Downloads: (external link)
https://hal.science/hal-04381138/document (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:wpaper:hal-04381138

DOI: 10.2139/ssrn.4465833

Access Statistics for this paper

More papers in Working Papers from HAL
Bibliographic data for series maintained by CCSD ().