Skip to main content
. 2023 Sep 29;15(9):e46222. doi: 10.7759/cureus.46222

Figure 4. Modified Bland-Altman plot for assessing agreement between the scores provided by three raters to the explanations of ChatGPT-3.5.

Figure 4

For ChatGPT-3.5, the limit of agreement was calculated as 0.55 and is as shown in the figure with a black dashed line.