Search for a matching or mismatching letter pair

The perceptual matching task was modified in order to increase the error rate and thus to reveal more clearly whether internal noise more often changes an objective match into a perceived mismatch than vice versa (noisy-operator theory). In Experiment 1, subjects searched for a “same” pair in a list of “different” pairs or for a “different” pair in a list of “same” pairs. As predicted, the target pair was overlooked or missed more often on “different” lists. False alarms, though, were not higher on “same” lists, owing apparently to rechecking, which also produced slower search through “different” lists. To disable the rechecking mechanism, rapid serial visual presentation (RSVP) was used in Experiments 2 and 3. As predicted, misses exceeded false alarms on “different” lists, whereas the reverse held on “same” lists, both at the physical level (Experiment 2) and at the name level (Experiment 3). Letter set size and repetition were varied in Experiments 1 and 2, and the results indicate that when letter pairs are presented in close temporal and spatial proximity, subjects are influenced by extraneous interpair comparisons in addition to relevant intrapair comparisons in deciding whether or not a target pair is present.