The influence of irrelevant information on speeded classification tasks

Abstract

Multidimensional stimuli, which could vary on one, two, or all three dimensions within a particular series, were presented to Ss who were required to classify each stimulus on the basis of its value on a specified dimension. The prior relevance of the irrelevant dimensions and the difficulty of the task were varied. Latency and error data indicated that Ss were unable to gate the irrelevant information effectively. It was further concluded that this lack of perfect gating cannot be simply attributed to competing responses learned during the experiment.

Competing responses and the processing of irrelevant information

“…In a review of filtering in speeded classification tasks, Egeth (1967) labeled the explanation the competing-response (CR) hypothesis, and noted its applicability to several other studies, especially the Stroop (1935) color-naming task. However, Egeth and Bevan (1972) and Well (1971) have recently argued that the CR hypothesis is not unambiguously applicable to the results obtained by Hodge and Montague, They claim that Hodge and Montague confounded irrelevant information with task complexity. Furthermore, there also seems to be evidence that irrelevant information can degrade performance in the absence of competing responses (Morgan & Alluisi, 1967;Well, 1971), a result in sharp contrast with those obtained by Archer (1954), Fitts and Biederman (1965), Imai and Garner (1965), and Morin et al (1961).…”

“…Notwithstanding the criticism of Egeth and Bevan (1972) and Well (1971) that Hodge (1959) and Montague (1965) confounded irrelevant information with task complexity, it is difficult to imagine a situation in which the amount of relevant or irrelevant information can be varied independently of task complexity. In fact.…”

“…In the Well (1971) investigation (Experiment I), significant performance decrements were produced by the number of irrelevant dimensions. the level of discrimination difficulty.…”

“…Some of these studies (e .g., Egeth, 1966;Hodge, 1959;Hodge & Reid, 1971;Montague, 1965;Morgan & Alluisi, 1967;Well, 1971) have found that irrelevant information degrades performance, while others (e.g., Archer, 1954;Fitts & Biederman, 1965;Imai & Garner, 1965;Morin, Forrin, & Archer, 1961) have not. To account for the detrimental effects of irrelevant information, Hodge (1959) suggested that the occurrence of competing responses was a likely explanation.…”

“…In view of the questions concerning the CR hypothesis and given the results of Morgan and Alluisi (1967) and Well (1971), it seems appropriate to reanalyze the hypothesis and its role in the processing of *The author wishes to thank Clyde E, Noble and Howard Egeth for their critical reading of the manuscript. irrelevant information.…”