Priming and the guidance by visual and categorical templates in visual search

Abstract

Visual search is thought to be guided by top-down templates that are held in visual working memory. Previous studies have shown that a search-guiding template can be rapidly and strongly implemented from a visual cue, whereas templates are less effective when based on categorical cues. Direct visual priming from cue to target may underlie this difference. In two experiments we first asked observers to remember two possible target colors. A postcue then indicated which of the two would be the relevant color. The task was to locate a briefly presented and masked target of the cued color among irrelevant distractor items. Experiment 1 showed that overall search accuracy improved more rapidly on the basis of a direct visual postcue that carried the target color, compared to a neutral postcue that pointed to the memorized color. However, selectivity toward the target feature, i.e., the extent to which observers searched selectively among items of the cued vs. uncued color, was found to be relatively unaffected by the presence of the visual signal. In Experiment 2 we compared search that was based on either visual or categorical information, but now controlled for direct visual priming. This resulted in no differences in overall performance nor selectivity. Altogether the results suggest that perceptual processing of visual search targets is facilitated by priming from visual cues, whereas attentional selectivity is enhanced by a working memory template that can formed from both visual and categorical input. Furthermore, if the priming is controlled for, categorical- and visual-based templates similarly enhance search guidance.

Keywords: SOA; cueing; priming; template; visual search; working memory.

Figure 1

The figure depicts an outline of the task of (A) Wilschut et al. (2013) as well as (B) the postcue manipulation, and (C) the precue manipulation of Experiments 1, 2 of the present study, respectively. (A) adapted from Wilschut et al. (2013). The time it takes to turn a memory into a template. Copyright by the Association for Research in Vision and Ophthalmology (ARVO).

Figure 2

Accuracy scores for Experiment 1, presented as a function of postcue type and SOA. Left part of the figure shows accuracy scores separately for the relevant and irrelevant set sizes, right part shows the total averages over the set sizes. Error bars show the standard error of the mean here and in other figures.

Figure 3

Experiment 1: Mean error slopes for the relevant and irrelevant set, and their difference, plotted as a function of postcue type and SOA.

Figure 4

Model simulations for overall accuracy and selectivity. The top row shows a constant priming at all SOAs. The bottom row shows priming that decreases with SOA.

Figure 5

Accuracy scores for Experiment 2, presented as a function of precue type and SOA. Left part of the figure shows accuracy scores separately for the relevant and irrelevant set sizes, right part shows the total averages over the set sizes.

Figure 6

Experiment 2: Mean error slopes for the relevant and irrelevant set, and their differences, plotted as a function of precue type and SOA.