Table 3.
Naïve and non-naïve LOOCV (on individual participants) Bayes classification accuracy per dimension (i.e., category system).
| Predicted dimension | Prior-only | Semantic | Structure | Appearance | Naïve | Non-naïve |
|---|---|---|---|---|---|---|
| Semantic | 37.36% | – | 42.27% | 44.65% | 48.44% | 48.08% |
| Structure | 31.53% | 15.33% | – | 30.58% | 26.05% | 50.69% |
| Appearance | 35.69% | 39.49% | 33.24% | – | 39.60% | 41.27% |