Do different attention capture paradigms measure different types of capture?

Abstract

When something captures our attention, why does it do so? This topic has been hotly debated, with some arguing that attention is captured only by salient stimuli (bottom-up view) and others arguing capture is always due to a match between a stimulus and our goals (top-down view). Many different paradigms have provided evidence for 1 view or the other. If either of these strong views are correct, then capture represents a unitary phenomenon, and there should be a high correlation between capture in these paradigms. But if there are different types of capture (top-down, bottom-up), then some attention capture effects should be correlated and some should not. In 2 studies, we collected data from several paradigms used in support of claims of top-down and bottomup capture in relatively large samples of participants. Contrary to either prediction, measures of capture were not strongly correlated. Results suggest that capture may in fact be strongly determined by idiosyncratic task demands and strategies. Relevant to this lack of relations among tasks, we observed that classic measures of attention capture demonstrated low reliability, especially among measures used to support bottomup capture. Implications for the low reliability of capture measures are discussed. We also observed that the proportion of participants demonstrating a pattern of responses consistent with capture varied widely among classic measures of capture. Overall, results demonstrate that, even for relatively simple laboratory measures of attention, there are still important gaps in knowledge regarding what these paradigms measure and how they are related.

Attention capture, processing speed, and inattentional blindness

“…While it is surprising that implicit capture was almost completely unrelated to noticing the unexpected event, it is possible that the lack of relationship is due to reliability issues of the tasks themselves (see Roque et al, 2016). In addition to reliability issues, participants may have different strategies and approaches to performing attention capture tasks, meaning that capture effects may be measuring something different for different participants.…”

“…The same methods and undergraduate sample from Study 1 of Roque et al (2016) were used for the current study. Here we present an abbreviated version of the methods.…”

“…Specifically, the Additional Singleton Paradigm (Theeuwes, 1992) and the Irrelevant Singleton Paradigm (Yantis & Jonides, 1984; Jonides & Yantis, 1988) were used for indices of bottom-up capture and the Contingent Cuing Paradigm (Folk et al, 1992) and the Contingent Blink Paradigm (Folk et al, 2002) were used as indices of contingent capture. Capture scores were then calculated in a manner consistent with the literature for each task (see Roque et al, 2016). Specifically, in order to determine the relationship between these two different types of capture, attention capture scores were calculated from RT and accuracy proportion scores from each condition of the various classic attention measures.…”

“…That is we divided observers’ reaction time by the accuracy in each condition to create an inefficiency proportion score for that condition. These inefficiency scores that take into account both accuracy and RT are generally more reliable than scores based purely on RT (Roque et al, 2016). For search tasks involving search for a target of a specific color, the color of the search target was held constant (i.e., green) to reduce the possibility of carryover effects.…”

“…We explored the degree to which individual differences in susceptibility to attention capture and processing speed are important predictors of IB. One hundred forty-six participants (from Study 1 of Roque, Wright, & Boot, 2016) completed four classic attention (implicit) capture tasks designed to assess stimulus-driven and contingent capture. Following the completion of these capture tasks, participants completed a sustained IB (multiple object tracking) task where an unexpected event appeared during the final critical trial.…”