Skip to main content
. 2025 Sep 24;9:e75421. doi: 10.2196/75421

Table 3. Auto-evaluator performance on matched (N=187) and mismatched vignettes (N=187).

Matched vignette, n (%) Mismatched vignette, n (%)
Dimension High Medium Low High Medium Low
Evidence 184 (98.4) 3 (1.6) 0 36 (19.3) 32 (17.1) 119 (63.6)
Realism 187 (100) 0 0 84 (44.9) 94 (50.3) 9 (4.8)
Completeness 156 (83.4) 31 (16.6) 0 19 (10.2) 46 (24.6) 122 (65.2)