Selection history is relative

Abstract

Visual attention can be tuned to specific features to aid in visual search. The way in which these search strategies are established and maintained is flexible, reflecting goal-directed attentional control, but can exert a persistent effect on selection that remains even when these strategies are no longer advantageous, reflecting an attentional bias driven by selection history. Apart from feature-specific search, recent studies have shown that attention can be tuned to target-nontarget relationships. Here we tested whether a relational search strategy continues to bias attention in a subsequent task, where the relationally better color and former target color both serve as distractors (Experiment 1) or as potential targets (Experiment 2). We demonstrate that a relational bias can persist in a subsequent task in which color serves as a task-irrelevant feature, both impairing and facilitating visual search performance. Our findings extend our understanding of the relational account of attentional control and the nature of selection history effects on attention.

Keywords: Attentional capture; Feature-based attention; Selection history.

Figure 1 -

RGB (red, green, blue) values of the colors used (values in parentheses) and their positions in CIE (Commission International de l’Éclairage) 1976 color space (u'- and ν'-coordinates).

Figure 2 –

Time course of trial events during the training and test phase of Experiment 1. Trials in both phases were preceded by a fixation cross; after fixation on the cross was registered, the task began. In the training phase (A), participants were instructed to fixate the uniquely-colored stimulus and received rewards depending on the set of colors. In the test phase (B), participants were instructed to make a speeded saccade towards the singleton shape while ignoring the colors. There was no feedback for correct responses, but the words “Too Slow!” would appear if the target shape was not fixated before the timeout limit.

Figure 3 –

Mean performance for the training phase. (A) RT and (b) accuracy between the different High Value (HV) conditions for the two sets of colors. Error bars represent the standard error of the mean.

Figure 4 –

Proportion of oculomotor capture from the test phase for (A) the group with cyan as the high-value target and (B) the group with orange as the high-value target. Error bars represent standard error of the means.

Figure 5 –

Sequence of events for Experiment 2. In the training phase (A), each trial began with a fixation cross for a random duration between 400 and 600ms, followed by four squares for 800ms, or until response. “Incorrect” was displayed in response to an incorrect response, and “Too Slow” was displayed if a response was not registered before the timeout. Both feedback displays were presented for 1000ms, after which a blank screen appeared for 500ms. In the test phase (B), each trial began with a fixation cross for a random duration between 440 and 580ms, followed by either a diamond among three circles or a circle among three diamonds. The feedback and blank displays were the same as in the training phase.

Figure 6 –

RT for the test phase of Experiment 2 for context, trained and relative target colors, for the (top) yellow-context condition that encouraged a redder-than bias and for the (bottom) red-context that encouraged a yellower-than bias. Error bars represent standard error of the means.