Statistical comparison of empirical weights (wV,emp) and standard deviations (σAV,emp) obtained from the psychometric (behavior) and neurometric (fMRI) functions pertaining to the audiovisual conditions of high and low visual reliability with the MLE predictions (σAV,pred, wV,pred) and unisensory standard deviations (σuniV, σuniA)
| Comparison | VR+ | VR– |
|---|---|---|
| wV,emp – wV,pred | ||
| Psychophysics | 0.063 | 0.129 |
| V1–V3 | 0.230 | 1 |
| IPS0–4 | 0.631 | 1 |
| Low-level auditory | 0.064 | 1 |
| σAV,emp – σAV,pred | ||
| Psychophysics | 0.188 | 0.063 |
| V1–V3 | 0.241 | 0.001 |
| IPS0–4 | 0.020 | 0.275 |
| Low-level auditory | 1 | 1 |
| σAV,emp – σuniV | ||
| Psychophysics | 0.188 | 0.438 |
| V1–V3 | 0.241 | 0.001 |
| IPS0–4 | 0.022 | 1 |
| Low-level auditory | 0.883 | 0.963 |
| σAV,emp – σuniA | ||
| Psychophysics | 0.063 | 0.313 |
| V1–V3 | 0.104 | 0.169 |
| IPS0–4 | 0.002 | 0.086 |
| Low-level auditory | 1 | 1 |
Numbers denote p values (p < 0.05 printed in bold). Psychophysics parameters were compared using two-tailed Wilcoxon signed rank tests on individual parameters (random-effects analysis, df = 4). Neurometric parameters from V1–V3, IPS0–4, and low-level auditory regions were compared using a two-tailed bootstrap test (5000 bootstraps) on parameters computed across the sample (fixed-effects analysis). All comparisons of neurometric parameters were Bonferroni corrected across the three regions of interest. A, auditory; V, visual; VR±, high/low visual reliability.