Academic journal article Canadian Journal of Experimental Psychology

Qualitative Differences in Response Bias from Spatial Cueing

Academic journal article Canadian Journal of Experimental Psychology

Qualitative Differences in Response Bias from Spatial Cueing

Article excerpt

Abstract Two response bias parameters were examined in a spatial (periphera l) cueing paradigm: (i) "cued location bias": the relative probability with which a signal is incorrectly assigned to the cued location; and (ii) "likelihood ratio (beta)": the amount of evidence required to decide that a signal appeared at a particular, cued or uncued, location. The two parameters were found to be differentially sensitive to the stimulus onset asynchrony (SOA) between peripheral cue and target. Cued location bias was strongly affected by SOA, dec reasing within 300 ms after peripheral cue onset. In contrast, the difference in beta between cued and uncued locations (cued < uncued) was independent of SOA. This suggests that cued locat ion bias reflects strong but transitory pathway pre - activation following a peripheral c ue, consistent with accounts of response bias effects in terms of relatively "early" processes of st imulus coding (e.g., Hawkins, Shafto, & Richardson, 1988; Shulman & Posner, 1988). In contras t, beta effects predominantly reflect "late" decision making processes that differential ly weight the sensory evidence from cued and uncued locations according to their a - priori si gnal probabilities (e.g., MUller & Findlay, 1987; Shaw, 1982).

How does attention influence visual perception? One line of research on this que stion has examined the effects of spatial cues on the accuracy of detection of, or speed o f responding to, a subsequent target. This work has demonstrated that advance spatial cues enhan ce detection accuracy for, and speed reaction times (RTs) to, targets at cued relative to unc ued locations.

To explain the effects of spatial cueing on detection accuracy, it seems us eful to assume that there are at least two functional stages between stimulus and response: In the f irst, "coding" stage (which may consist of a number of substages), the stimulus is converted in to an internal representation. In the second, "decision" stage, the internal representation is used to determine a response (e.g., a decision that the target is present or absent). According t o signal detection theory (SDT; Green & Swets, 1966), the internal stimulus representation is chara cterized by a one - dimensional strength variable and the observer makes decisions about the p resence of a target by comparing the strength of a representation against a criterion. Thus, the decisive question is whether spatial cueing influences the quality of stimulus representations generated by the coding process and/or whether it causes criteri on shifts in the decision process.

Most studies investigating cueing effects have measured simple RTs to supra threshold luminance increments (e.g., Posner, Nissen, & Ogden, 1978; Posner, Snyder, & Dav idson, 1980). However, Duncan (1980), Sperling (1984; Sperling & Dosher, 1986), and Sh aw (1984) have argued that it does not follow from the mere presence of the cueing effect that stimulus coding at cued locations is enhanced by the allocation of attention. RT advanta ges for cued (more likely), relative to uncued (less likely), locations can equally well be e xplained by a simple response bias: that is, a tendency by the observer to reduce the amount o f evidence required to decide whether a change has occurred at the more likely locations (a nd, possibly, to raise the criterion for the less likely locations).

MUller and Findlay (1987) directly investigated the question of sensitivity versus criterion changes under spatial cueing conditions by combining a "cost - benefit" analysis (Posner & Snyder, 1975) with a SDT procedure. They found that with tasks requiring proces sing of stimulus shape, there were benefits in sensitivity (P(A) parameter) for cued (li kely) locations and costs for uncued (less likely) locations, relative to a neutral cueing condition in which all locations were equally likely. This indicates that cueing enhances the processi ng of stimulus shape. …

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.