Abstract
The inversion effect is often considered a hallmark of "holistic" processing with face objects. Because non-face objects show smaller inversion effects than faces, they are thought to be recognized using feature-based processes, not "holistic". Previously, we showed that one of the keys to obtaining large inversion effects (or "holistic" processing) is the strategic combination of object features. Here, we extend this line of research by investigating the role of stimulus quality on these combinatorial effects. In Exp 1, four sets of novel objects and four sets of face objects were created by manipulating specific top and bottom features. Top diagnostic sets varied only in the top features, bottom diagnostic sets varied only in the bottom features. Conjunction sets varied across both top and bottom features, with OR conjunctions allowing the use of either the top or the bottom, and with AND conjunctions requiring the use of both the top and the bottom. Results showed large inversion effects for both face and non-face objects in the AND condition, but only for face objects in the OR condition. There were no inversion effects in diagnostic conditions. These results are consistent with our previous conclusion that strategic feature combination is related to the size of object inversion effects. In Exp 2, stimuli from the novel OR and AND conditions were presented at three different levels of stimulus quality (signal-to-noise ratio). Results showed that lower stimulus quality produced larger inversion effects, but only in the AND condition. The results suggest that the feature combination process that contributes to inversion effects is recruited more when stimuli are degraded. The results are discussed in relation to the principle of inverse effectiveness used in the study of sensory integration.
Meeting abstract presented at VSS 2012