Part-whole information is useful in visual search for size x size but not orientation x orientation conjunctions.

In visual search tasks, subjects look for a target among a variable number of distractor items. If the target is defined by a conjunction of two different features (e.g., color x orientation), efficient search is possible when parallel processing of information about orientation is used to guide the deployment of attention to the target. Another type of conjunction search has targets defined by two instances of one type of feature (e.g., a conjunction of two colors). In this case, search is inefficient when the target is an item defined by parts of two different colors but much more efficient if the target can be described as a whole item of one color with a park of another color (Wolfe, Friedman-Hill, & Bilsky, 1994). In this paper, we show that the same distinction holds for size. "Part-whole" size x size conjunction searches are efficient; "part-part" searches are not (Experiments 1-3). In contrast, all orientation x orientation searches are inefficient (Experiments 4-6). This difference between preattentive processing of color and size, on the one hand, and orientation, on the other, may reflect structural relationships between features in real-world objects.