Academic journal article Attention, Perception and Psychophysics

The Multiple-Weighting-Systems Hypothesis: Theory and Empirical Support

Academic journal article Attention, Perception and Psychophysics

The Multiple-Weighting-Systems Hypothesis: Theory and Empirical Support

Article excerpt

Published online: 9 December 2011

© Psychonomic Society, Inc. 2011

Abstract Observers respond faster when the task-relevant perceptual dimension repeats across consecutive trials (e.g., color-color) relative to when it changes (orientation- color)-the phenomenon termed the dimension repetition effect (DRE). Similarly, when two (or more) different tasks are made to vary randomly across trials, observers are faster when the task repeats, relative to task changes- the phenomenon termed task-switch cost (TSC). Hitherto, the DRE and TSC effects have been discussed independently of each other. Critically, either effect was explained by assuming a single mechanism giving rise to DREs or TSCs. Here, we elaborate strong conceptual similarities between the DRE and TSC effects; we introduce the concept of criterion-specific intertrial sequence effects, with DREs and TSCs being different manifestations of criterionspecific effects. Second, we review available evidence suggesting that none of the single mechanism explanations can readily account for all the findings in the literature. Third, we elaborate on the multiple-weighting-systems (or MWS) hypothesis, a recently proposed account that postulates the existence of several, independent mechanisms sensitive to intertrial sequences. Finally, we test predictions derived from the MWS hypothesis in two novel experiments and discuss the results from both the single- and multiplemechanism perspectives.

Keywords Dimension repetition effect . Task switching . Multiple weighting systems . Dimensional weighting

Introduction

What we did recently has a substantial effect on our current performance. For instance, we are likely to set a table faster when we put all plates down first, as compared with setting down plates and cutlery in mixed order. Conceptually, one can think of three broadly defined processes that are potentially facilitated by such action sequencing: (1) visual selection (e.g., detecting plates among cups and bowls), (2) perceptual or semantic analysis (e.g., deciding whether the selected plate belongs to the appropriate set), and (3) action execution (e.g., transferring the plate to the table). When setting the table, all three processes (selection, analysis, action) are involved, thus making it difficult to isolate the specific cognitive processes affected by recent experience. Evaluating the role of different cognitive processes as a source of history effects on human performance is the main focus of the present study.

Typically, history effects are investigated by examining changes in a dependent variable as a function of the intertrial sequence-that is, repetitions versus changes of stimulus and/or response properties across consecutive trials. Although many such effects have been described in the literature, they can be broadly classified as being either (1) feature specific or (2) criterion specific in type. Feature-specific intertrial effects relate to repetition/changes of specific paradigm properties: changes in the exact defining feature of the stimulus (e.g., from green to red) or changes in the required response (e.g., from a left- to a right-buttonpress). By contrast, criterion-specific effects relate to more abstract changes in task-relevant stimulus properties (e.g., switching from color to orientation as critical stimulus property) or changes in stimulus-response (S-R) mappings (e.g., rom manual to vocal responding).

The present article focuses primarily on criterion-specific intertrial effects. To set the stage, first, different types of criterion effects are outlined and their conceptual similarity is elaborated. Next, unresolved discrepancies between dominant accounts of criterion-specific intertrial effects, all postulating a single mechanism underlying these effects, are elaborated, and several recently proposed, integrative accounts are outlined-with a focus on our own account, which postulates the existence of multiple mechanisms capable of producing criterion-specific intertrial effects. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.