Discriminating the odd: Boundaries of visual-haptic integration

We investigated the degree to which visual-haptic fusion occurs. If the nervous system uses a statistically optimal integration rule, the combined visual-haptic percept is a weighted average of the visual and haptic estimates and its variance is lower than that of either estimate alone. We used an oddity task to investigate whether the combined estimate is used in discrimination or whether independent visual and haptic estimates are used. Three horizontal bars were presented sequentially. Two of them were identical and had equal visual and haptic heights (standard stimulus). The third had a visual and/or haptic height differing from the standard (odd stimulus). Subjects indicated which of the three intervals contained the odd stimulus. If subjects used visual or haptic information independently without combining them, then discrimination would occur whenever the visual or haptic height in the odd stimulus differed from the height in the standard by more than the vision-alone or haptic-alone threshold. In contrast, if subjects relied on the combined visual-haptic estimate, discrimination should be most difficult when the visual and haptic heights differed in opposite directions from the standard's height (such that weighted averages of odd and standard stimuli are equal) and easiest when they differed in the same direction from the standard (weighted averages different). We found that discrimination was indeed most difficult when the weighted averages were the same and easiest when they differed. Thus, the fused visual-haptic percept is used for discrimination. However, if the conflict between the visual and haptic stimuli is too large, this difference in discrimination performance is not observed. In other words, visual-haptic fusion breaks with large conflicts. In some conditions, metameric behavior is observed: discrimination would be better if subjects shut the eyes or removed the hand from the bar