Abstract
Visual attention can be reflexively drawn to salient stimuli (saliency capture), sudden onsets (onset capture), or stimuli that are similar to a sought-after target (contingent capture). However, previous studies showed that a salient distractor can affect attention very differently depending on whether its effects are measured in the visual search paradigm or the spatial cueing paradigm. In the present study, we compared the effects of different distractors in both tasks (visual search, spatial cueing), to ascertain whether different forms of attentional capture (saliency capture, onset capture and contingent capture) are indeed related across these paradigms. In the visual search task, we monitored participant's eye movements in response to a salient red or green distractor, an onset distractor, or a target-similar distractor in search for a shape target. In the spatial cueing paradigm, we assessed validity effects to the same stimuli when the distractor (cue) was briefly presented prior to the target display. The results showed strong capture effects for the onset and target-similar distractor across both paradigms, but only weak effects of salient colour distractors. In visual search, capture scores derived from eye movement parameters were much more reliable than other measures (e.g., RT). Still, there were significant correlations in capture for all distractors across paradigms. Overall, target-similar distractors captured attention most reliably and correlated most strongly across paradigms. Performance in visual search also correlated with intelligence (g), though this was mainly limited to the ability to quickly select the target. Overall, the results indicate that the two paradigms tap into the same constructs of attentional capture, which are however difficult to measure with a high degree of reliability. Moreover, capture by salient items or sudden onsets are probably not related to a weak top-down control mechanism, as contingent capture and onset/saliency capture were not inversely related to each other.
Meeting abstract presented at VSS 2018