Abstract
When an object is partially occluded, the process of amodal completion can perceptually fill in the occluded object to help guide attention (Rensink & Enns, 1998). However, amodal completion for an objects’ guiding feature, such as orientation, can fail, causing search to be inefficient. (Wolfe et al., 2011). However, these tasks used orientation as the target-defining feature, and our question was whether we would see evidence for amodal completion when the task was to search through complex objects. To test this, we displayed objects that had different levels of internal complexity (Black bar, Hershey bar, and Remote Controller), yet had the same horizontal and vertical dimensions. The targets and distractors could be fully visible or partially occluded by a diagonal white bar or a gap. Distractors were either horizontal or vertical versions of those three stimuli. In the first experiment participants searched for a vertical. When there was no occlusion, the target popped-out, no matter the object type. However, when the target was occluded, search was more difficult when it was a black bar or Hershey bar compared to when it was the remote control. This object-specific effect might be due to surface features of the remote control interacting with the occluder to make detection of its vertical orientation easier. However, vertical is not an inherent feature of any of the objects, but instead a transient property. In the second experiments, participants searched for an object based on its identity. Occlusion had a larger effect on the complex objects compared to the simple bar, with search efficiency roughly 2.5× slower. In addition, the remote control failed to pop-out when there was no occlusion. Effects of occlusion on visual search are complex, object specific, and may depend on which surface features are occluded.