Distinguishing Authentic Voices in the Age of ChatGPT: Comparing AI-Generated and Applicant-Written Personal Statements for Plastic Surgery Residency Application
Viren Patel, Anthony DeLeonibus, Michael Wells, Steven Bernard, Graham S. Schwarz
Abstract
BACKGROUND: ChatGPT, a generative artificial intelligence model, may be used by future applicants in the plastic surgery residency match. METHODS: Ten personal statements (5 generated by ChatGPT, 5 written by applicants) were rated by 10 reviewers, blinded to the source of the essay. RESULTS: A total of a 100 evaluations were collected. There was no significant difference in ratings for readability, originality, authenticity, and overall quality (all P > 0.05) when comparing computer-generated and applicant essays. CONCLUSION: Personal statements prepared by ChatGPT are indistinguishable from essays written by actual applicants. This finding suggests that the current plastic surgery application format be reevaluated to better aid in holistic evaluation of students.