Skip to main content
. 2024 Feb 6;11:e54369. doi: 10.2196/54369

Table 2.

Comparison of ChatGPT-4’s Level of Emotional Awareness Scale performance with that of the French populationa.

Score French men, mean (SD) French women, mean (SD) ChatGPT-4 (1-sample z tests) Bard (1-sample z tests)
Total 56.21 (9.70) 58.94 (9.16)

  • ChatGPT-4 score=97

  • Men: z=4.20; P<.001

  • Women: z=4.15; P<.001

  • Bard score =97

  • Men: z=4.20; P<.001

  • Women: z=4.15; P<.001
MCb 49.24 (10.57) 53.94 (9.80)

  • ChatGPT-4 score=79

  • Men: z=2.81; P=.004

  • Women: z=2.55; P=.01

  • Bard score=79

  • Men: z=2.81; P=.004

  • Women: z=2.55; P=.01
OCc 46.03 (10.20) 48.73 (10.40)

  • ChatGPT-4 score=77

  • Men: z=3.03; P=.002

  • Women: z=2.71; P=.006

  • Bard score=75

  • Men: z=2.84; P=.004

  • Women: z=2.52; P=.01

aAll statistically significant P values remained significant after false discovery rate correction (q<.05).

bMC: main character.

cOC: other character.