ABSTRACT
P. M. A. Rabbitt suggested that visual search might be treated as a particular example of a more general range of tasks in which the subjects (S) make the same response to all members of one group of stimuli and make different responses to members of other groups. P. M. Fitts et al. have shown that Ss learning to discriminate between complex patterns improve their performance by identifying and using a range of critical cues and by ignoring other details as misleading or redundant. The process of discriminating relevant from irrelevant letters would clearly be most economical if the Ss used the minimal number of cues. Two experiments used a training-transfer paradigm to investigate learning of irrelevant symbols in visual search tasks. Groups of Ss were practiced at locating members of particular sets of relevant letters embedded among restricted vocabularies of irrelevant letters.
