We take this opportunity to comment on the published article entitled “A comparison of cover letters written by ChatGPT-4 or humans ". Using parallel arms with eighteen cover letters each, the study compared cover letters composed by GPT-4 with others written by humans. By completing a questionnaire, three blinded assessors evaluated the letters based on impression, readability, criterion satisfaction and level of detail. Furthermore, readability assessments were based on the Flesch Kincaid grade level and Lix score. None of the parameters examined in the results revealed any noteworthy or pertinent differences between the two types of cover letters. Nonetheless, 61% of the assessors were able to determine with accuracy if a human or GPT-4 had written the letter. Furthermore, subjective evaluations revealed that GPT-4 performed better in terms of readability, but objective tests showed that GPT-4 had a higher score.
A few limitations of the research include the small sample size - just 36 cover letters were assessed - and the use of only three assessors. Furthermore, the study did not explore other elements such as content or formatting that may potentially affect how effective cover letters are. Future directions: Increasing the sample size and adding more evaluators would improve the generalisability and reliability of the study. Investigating other elements, i.e. content analysis and formatting strategies, that affect the quality of cover letters would also be advantageous. To establish how GPT-4 measures up against other AI systems, more research may compare it with those systems. Finally, the author of the cover letter bears the final responsibility for any writing as is the general rule when using AI .
- Deveci CD, Baker JJ, Sikander B, Rosenberg J. A comparison of cover letters written by ChatGPT-4 or humans. Dan Med J. 2023 Nov 23;70(12):A06230412.
- Kleebayoon A, Wiwanitkit V. ChatGPT, critical thinking and ethical practice. Clin Chem Lab Med. 2023 May 30;61(11):e221.