Expectancy effects: cost-benefit analysis of monaurally and dichotically presented speech.

In three experiments, cost-benefit analysis was used to determine the role of attention in the processing of auditory information. In two experiments consonant-vowel syllables were presented monaurally, while in the third the mode of presentation was dichotic. For all three experiments the ability to detect a target stimulus under conditions in which precued location information was valid or invalid was contrasted with detection in a neutral condition where no location information was provided. The results indicated that attention can be effectively deployed under monaural conditions when either a simple detection or a discriminative response is required. Similar conclusions cannot be reached when considering the effectiveness of attention under conditions of dichotic listening. These results are discussed in relation to accounts which argue that dichotic listening performance is critically dependent upon auditory disembedding and where attention is viewed as a late process involving response selection.