ABSTRACT
In visual search experiments, subjects are asked to detect particular target stimuli presented among irrelevant nontargets. The difficult y of visual search is thus determined by whether a target is unique in some elementary feature or only in its conjunction of features. Feature integration theory has been applied to letter search by considering the conjunction of a shape's parts. There have been many visual search experiments using simple shapes such as letters and digits. It is supported by physiological evidence for early analysis of different stimulus attributes in different brain areas. To speak of stimulus similarity is fairly unambiguous when stimuli differ only along simple dimensions like size and color. From the preceding account it is clear that we shall have to consider two sorts of stimulus similarity in visual search. Unlimited-capacity parallel models predict no difference between simultaneous and successive conditions, contrasting with their inexact prediction of some unknown effect when the total amount of material changes.
