Abstract
A major challenge for undergraduate students is reading scientific literature. This is especially true in biophysics, where many of the concepts may not have been covered in undergraduate courses. Students can become overwhelmed, which may lead to less overall engagement with the scientific literature. In response, we have developed a guided reading protocol that combines pre-reading strategies, structured notetaking, and ChatGPT to help students clarify unfamiliar concepts in an interactive way. To test the protocol, participants in this study were given an initial survey to determine their experience with reading scientific literature. After this they were given an abridged biophysics paper and the protocol. The ChatGPT transcripts were analyzed using open coding and the students were given a post-study survey. We found most students did not appear to regularly engage with the literature, possibly due to content barriers they encountered. Analyzing their transcripts, we observed that students asked for definitions, explanations, summaries, and simplifications. Overall, students reported that using ChatGPT was a positive experience and they expected to use ChatGPT in the future. From this work, we expect this new protocol may be a way to keep novice students from becoming discouraged when reading scientific papers and keep them engaged with the current literature.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
1. Removed the word Biophysics from the title and changed it to A ChatGPT Assisted Reading Protocol for Undergraduate Research Students. 2. Streamlined the abstract and reduced the wordiness. 3. In the Methods section, we addressed the fact that we did not ask the students for which version of ChatGPT they used. 4. We included anonymized versions of the survey questions and responses as spreadsheet files in Supplemental File 1 (initial survey) and Supplemental File 2 (post-study survey). 5. Included new figures summarizing results previously reported in the manuscript and results from the survey data, in order to help organize the data and make it clearer. 6. In paragraph 2 of the Introduction, we elaborated on the methods that Lennox et al found that students were using to read primary scientific literature and also enumerated some of the important differences in skills needed to read scientific literature as opposed to everyday reading. 7. In the Methods section, we also addressed that our protocol asks students to analyze headings and figures, but the abridged paper does not include figures. 8. In our discussion of the ChatGPT transcripts, we note this and mention that it may be worthwhile to include specific instructions to summarize in the updated versions of the reading protocol. We also address this again in our discussion of Figure 8 (Results and Discussion: Post-implementation Survey). 9. We added text to the conclusion addressing that this type of method may be applicable more generally in classwork assignments. 10. Lastly, we also added a statement to the Acknowledgment section stating that none of the authors had any competing interests.
