Attention is captured by distractors that uniquely correspond to controlled objects: An analysis of movement trajectories

Atten Percept Psychophys (2015) 77:819–829 DOI 10.3758/s13414-014-0801-5 Attention is captured by distractors that uniquely correspond to controlled objects: An analysis of movement trajectories James D. Miles & Robert W. Proctor Published online: 3 December 2014 # The Psychonomic Society, Inc. 2014 Abstract When reaching for a particular target, movements tend to deviate toward distractors. Previously, we have shown that cursor movements deviate to a greater degree toward distractors when the distractor color corresponds to that of the cursor and the target color does not, even when this relationship is task-irrelevant (Miles & Proctor, 2011). In the present study, we investigated whether this correspondence effect is due to attention capture or to the activation of responses based on the task response rules associated with the colors of the distractor and target (viz. a flanker effect). Participants moved a central rectangular cursor to an upper left or upper right location, depending on the cursor color. The colors of the target (correct response side) and distractor (incorrect response side) were independent from one another and were either corresponding or noncorresponding with respect to the cursor color. In Experiment 1, reaction times were delayed when the distractor color corresponded to that of the cursor, but only when the target color did not correspond to the cursor color. No color correspondence effect was found for movement trajectories or movement times. However, in Experiment 2, when responses were time-pressured, initial movements toward the distractor were much more common when the distractor color exclusively corresponded to the cursor color. On the basis of these results, we argue that attention capture best explains the increased tendency to move a controlled object to a distractor that uniquely shares its features. J. D. Miles (*) Department of Psychology, California State University, 1250 Bellflower Boulevard, Long Beach, CA 90840-0901, USA e-mail: R. W. Proctor Purdue University, West Lafayette, IN, USA Keywords Attentional capture . Goal-directed movements . Perception and action When reaching to one of an array of potential targets, movements tend to deviate toward neighboring distractors in other potential target locations (Song & Nakayama, 2009; Spivey, Grosjean, & Knoblich, 2005; Tipper, Howard, & Houghton, 1998; Tipper, Lortie, & Baylis, 1992; Welsh & Elliott, 2004). Most of the reaching studies above have placed emphasis on the location, number, and type of distractor. The properties of the object that is controlled during the movement—often some variant of a computer cursor—are largely ignored. However, the correspondence, or match, between the features of the controlled object and those of the intended target and other distractors in the environment may also influence response selection, which would be reflected in response latencies and movement trajectories. These situations are commonplace; for example, most sorting tasks involve identifying some specific feature of an object and, on the basis of that feature, selecting the correct location at which to place the object. The potential goal location may contain features that correspond (match) or do not correspond (mismatch) to those of the object, even if the features are irrelevant to the task itself. Evidence already exists that certain types of distractors influence movement trajectories when they share physical or conceptual similarities with the targets (e.g., Finkbeiner & Friedman, 2011; Song & Nakayama, 2006; Welsh, 2011; Welsh & Elliott, 2004). For example, Welsh (2011) found that an invalid cue immediately prior to movement to a target location caused movements to deviate in its direction, but only when the cue included the same feature that was used to select the target. Miles and Proctor (2011) also showed that the relation between the cursor and distractor colors influenced responses. Participants moved a colored cursor (red or green) 820 to a rectangle on the left or right side of a display (the target) while ignoring a rectangle on the opposite side (the distractor); one of the rectangles was red and the other was green. Responses were slower and deviated toward the distractor side when the distractor matched the color of the cursor, even when the color correspondence was irrelevant to task performance. This color correspondence effect only occurred when the cursor color determined the response (“if the cursor is green, move to the right rectangle; if the cursor is red, move to the left rectangle,” or vice versa). The authors concluded that color correspondence effects transpire between a controlled cursor and distractor, but only when participants are forced to first attend to the cursor color. Two mechanisms could potentially underlie the correspondence effects between cursors and distractors observed in Miles and Proctor’s (2011) study. First, attention may be captured by distractors that uniquely share the color of the cursor, which in turn activates a response in their direction. Once attention is captured by the color-corresponding distractor, a response code for a movement toward it is activated, which in turn interferes with execution of the response toward the target (e.g., Becker, 2010; Song & Nakayama, 2009; Welsh, 2011; Welsh, Elliott, & Weeks, 1999). In this case, competing responses toward multiple targets increase in activation until one of them reaches some threshold, and the amount of activation of a particular response is at least partially determined by the amount of attention the target of that response receives. Although some debate exists, there is consistent evidence that attention capture is contingent on taskrelevant parameters (Folk, Leber, & Egeth, 2002; Folk & Remington, 1998; Folk, Remington, & Johnston, 1992). Such contingent attention capture may be specific to a particular feature, such as a single color, rather than to the general feature category (e.g., all colors). This view has been substantiated by prior work showing attention capture following the presentation of a cue in the color that indicated a task-relevant target, but no similar capture for other colors (Ansorge & Heumann, 2003, 2004; Lamy, Leber, & Egeth, 2004; Leblanc, Prime & Jolicœur, 2008). Extending the logic of contingent capture to controller correspondence effects, attention capture by a particular target or distractor may be contingent on the features of the controller. In this case, attention is initially focused on the task-relevant color of the controlled object (the cursor), which in turn makes any other stimulus matching that color more salient. The now-salient corresponding stimulus then captures attention and elicits response activation toward its location. We refer to this as the attention-mediated account. Alternatively, the color of any target or distractor on the display may directly activate the response associated with its color on the basis o (...truncated)