Abstract
Inhibition of return (IOR) is indexed by slower reaction times to targets presented at previously attended locations (location-based IOR) or on previously attended objects (object-based IOR). Both location-based and object-based IOR can be found in response to moving objects. Jefferies, Wright, and Di Lollo (submitted) used a moving object that became occluded during its motion path to demonstrate that observer expectation is a powerful determining factor in IOR. Specifically, IOR to a moving object that becomes occluded occurs both at location-based and object-based coordinates only if the observer expects the object to continue to exist behind the occluder. If the observer expects that the object ceases to exist at the end of the motion sequence, IOR does not occur at either coordinate. The present series of experiments investigated the limits of this critical expectation. Experiment 1 examined the effect of disconfirming the observers' expectation that the object continued to exist behind the occluder. This was done by moving the occluder away from its initial location to reveal that the object had not persisted underneath it. We found that IOR was much diminished under these conditions. In Experiment 2, we asked whether a memory representation of the occluder, as opposed to a perceptual representation is sufficient to mediate IOR. We found that a memory representation of the occluder is as effective as its continuing presence on the screen in mediating IOR. In Experiment 3, we examined whether independent expectations can be developed simultaneously for separate objects, or whether once an expectation is developed, it applies to all objects present on the screen. Collectively, the outcomes of the present experiments buttress the claim of Jefferies, Wright, and Di Lollo that two factors, object continuity and observer expectation, mediate IOR.