![serial vs parallel search strategy definition serial vs parallel search strategy definition](https://royalsocietypublishing.org/cms/asset/36babc67-8a7e-4083-85b4-9a3352bd7c97/rsos191553f05.gif)
The serial account of the dual-target cost in RSVP tasks provides a potential explanation for the dual-target costs reported in longer duration free-view tasks (e.g., – ). According to this, the dual-target cost in single-fixation displays arises because observers are unable to switch the status of the target-templates in VWM quickly enough to search for both targets. The results revealed a reliable decrease in accuracy on dual- compared to single-target searches, with performance best described by a signal detection (SDT) model that limited attentional control to a single target-template at any one time. Evidence to support this model was reported in a study that used a rapid serial visual presentation (RSVP) task to compare observers’ accuracy when they searched for one or two targets. This functional distinction imposes a temporal cost on dual-target search because observers are required to switch the status of consecutive target-templates to search for both targets. Items outside the current focus of attention are maintained in a passive state, which neither contributes towards, nor interferes with, the search process. Active items, such as the target-template during search, must be available to guide the cognitive operations required to differentiate the target from distractors. According to the serial account, the functional status of items in memory is determined by the current focus of attention (e.g.,, ). Previous research suggests that competition for attentional control during search may be resolved in one of two ways: via a parallel process that is informed by simultaneously active target-templates, or a serial process, that restricts attentional control at any one time to a single target-template. How the brain resolves this competition during search, and its effect on target detection, have yet to be fully determined. When more than one target is sought, interactions between the information maintained in memory and visual input must be distributed across multiple objects, increasing competition for memory and attentional resources. In a single-target search, this ‘target-template’ is thought to provide a top-down bias for visual neurons that respond to the target’s features in the display. According to a number of models, memory-directed search requires observers to compare objects in the scene with a mental representation of the target. The magnitude of the dual-target cost depends upon the similarity of the two targets as well as their relation to the distractors: targets occupying non-contiguous regions within a feature-dimension elicit a larger dual-target cost as the distance between them and the number of intervening distractors increases –. This ‘dual-target cost’ is observed when the targets are differentiated from non-targets (distractors) by values along a single feature-dimension (e.g., their colour or orientation – ) and by variation across a range of feature-conjunctions (e.g., colour and orientation – ). Previous research has shown that this task is difficult: dual-target search is often slower and less accurate than separate searches for the same two targets. Checking baggage for a gun or the components of an explosive device, for example, requires security personnel to find multiple targets in complex visual displays. While this may have imposed a frustrating delay on your journey, the ability to search for multiple targets can have more important consequences in safety critical situations. The requirement to maintain two target-templates simultaneously, however, appears to impose a reduction in the specificity of the memory representation that guides search for each target.Īs you rushed to catch your bus this morning, you may have been delayed by a last minute search for your keys and your phone. These results are consistent with a parallel model of dual-target search in which attentional control is exerted by more than one target-template at a time.
![serial vs parallel search strategy definition serial vs parallel search strategy definition](https://d3i71xaburhd42.cloudfront.net/a67da5c004f86d073c7bb870540be2cfaa8a2aa7/7-Figure4-1.png)
Response accuracy also decreased on dual- compared to single-target searches: dual-target accuracy was higher than predicted by a model restricting search guidance to a single target-template and lower than predicted by a model simulating two independent single-target searches. The results revealed an increase in reaction times for dual- compared to single-target searches that was largely independent of the number of items in the display. The current study investigated this possibility by comparing behavioural responses during single- and dual-target searches for targets defined by their orientation. Recent research suggests this ‘dual-target cost’ may be attributable to a limit in the number of target-templates than can guide search at any one time. Simultaneous search for two targets has been shown to be slower and less accurate than independent searches for the same two targets.