Implementation and Evaluation of a ChatGPT-Assisted Special Topics Writing Assignment in Biochemistry

• Manik R. Reddy ^[1] ; Nils G. Walter ^[1] ; Yulia V. Sevryugina ^[1]
  1. [1] University of Michigan–Ann Arbor

     University of Michigan–Ann Arbor

     City of Ann Arbor, Estados Unidos

     
• Localización: Journal of chemical education, ISSN 0021-9584, Vol. 101, Nº 7, 2024, págs. 2740-2748
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • The effective and responsible educational application of ChatGPT and other generative artificial intelligence (GenAI) tools constitutes an active area of exploration. This study describes and assesses the implementation of a structured, GenAI-assisted scientific essay writing assignment in nucleic acid biochemistry. Briefly, students created, evaluated, and iteratively refined ChatGPT essays in response to feedback and independent literature research, identifying several strengths and shortcomings of large language model writing and citation practices. The scaffolded assignment structure aimed to prepare students for GenAI-assisted writing, and the majority of the class cohort ultimately indicated an improved understanding of GenAI functionality and prompt engineering, as well as interest in additional GenAI usage and applications. Moreover, students valued the instructional guidance on engagement with GenAI tools and the prompt engineering opportuvvvnities afforded by this exercise. However, discontentment with AI-produced citations was common, and 26% of supporting references were found to be nonexistent. The content evaluation and prompt generation strategies uncovered here may facilitate successful ChatGPT-guided writing assignments in other scientific contexts.