Skip to main content
. 2023 Dec 28;9:e48904. doi: 10.2196/48904

Table 5.

Sentiment comparison between human-written and ChatGPT-generated text in the medical abstract and radiology report data sets

Sentiment Medical abstract data set Radiology report data set

Human (n=2200), n (%) ChatGPT (n=8800), n (%) Human (n=2200), n (%) ChatGPT (n=8800), n (%)
Negative 432 (19.6) 1205 (13.7) 204 (9.3) 493 (5.6)
Neutral 1588 (72.2) 5822 (66.2) 1942 (88.3) 7738 (87.9)
Positive 180 (8.2) 1773 (20.2) 54 (2.5) 569 (6.5)