Consistent attending versus consistent responding in visual search: Task versus component consistency in automatic processing development

Abstract

The current research examined whether the total processing, from stimulus to response, must be consistent for automatic processing to improve task performance. Consistent versus inconsistent attending and responding ti.e., response translation) were factorially combined. Results showed that the consistency of attending produced substantial practice effects. However, consistency of the motor response translation component did not affect the asymptotic performance level. The results indicate that automatic processing develops for consistent com' ponents of a task even when the entire task is not consistent.The consistency with which subjects are able to deal with stimuli has important implications for performance in memory, detection, and search experiments. For example, in a visual search paradigm, Schneider and Shiffrin (1977, Experiment 2) found that reaction times in conditions in which subjects could consistently attend and respond to stimuli were 580 msec; when attending was not consistent , reaction times were 1,290 msec. Previous research has shown that, in general, when subjects consistently attend and respond to stimuli, their performance becomes less dependent on processing load , requires less effort, is more accurate, and is faster (e.g

Attention in the acquisition and expression of automaticity.

“…These experiments suggest that the physical response that is executed to report the decision is not an important part of the memory trace that supports automatic performance. The results corroborate findings in letter search experiments, in which the physical response was changed from trial to trial with no effect on performance, provided that the mapping of stimuli to response categories remained consistent (Fisk & Schneider, 1984;Shiffrin & Schneider, 1977, Experiment 3). Together, these results suggest that output or response selection does not exert strong constraints on the development and expression of automaticity.…”

“…The present experiments suggested that the results of target selection are stored in memory, which completes the analysis of the four kinds of attentional selection described by Treisman (1969). Previous data have suggested that the results of input selection (Logan & Etherton, 1994) and analyzer selection (Logan, 1988(Logan, , 1990 are encoded in memory during automatization, but the results of output or response selection are not (at least in the sense of storing associations with physical responses; Logan, 1990; see also Fisk & Schneider, 1984;Shiffrin & Schneider, 1977). The present Experiments 4, 7, and 8 suggested that the results of output selection may be encoded in some cases, tied together with the results of other kinds of analyses.…”

Visual-memory search: An integrative perspective

“…Finally, the decisions are consistently mapped on the physical responses, as in CM conditions. CVM conditions are therefore very different from the conditions involved in the experiments of Fisk and Schneider (1984) and Logan (1978), described previously, where the mapping of the stimulus sets to the target-present vs. target-absent decisions was consistent, but the mapping of the decisions to the physical responses was not. Shiffrin and Schneider (1977, Experiment 3) tested a CVM condition with homogeneous stimuli (i.e., consonants) in a multiple-frame task.…”

“…They varied the proportion of times that stimuli served as targets from 100% (perfectly consistent mapping) to 33% and found graded effects on the accuracy of target localization in a multiple-frame visual search task. Fisk and Schneider (1984) delineated the consistent mapping effect more precisely by showing that variations in the physical responses associated with target-present and target-absent decisions did not prevent automatization in the multiple-frame task. Logan (1978) reported similar findings, using response time in a single-frame task as the measure of performance.…”

Item to decision mapping in rapid response learning

“…So far, the evidence from previous studies has been mixed. In many of the previous studies of instance learning, motor mapping has been assumed to play a minimal role (Dennis & Schmidt, 2003;Fisk & Schneider, 1984;Logan, 1990). Instead, it has been assumed that instance learning reflects the formation of an association between items and response categories or classification actions (Logan, 1990).…”