Impulsivity and the window of attention

Potisek, Deborah L., Harold Greene, and Amanda Witkowski

It has been suggested that individuals with attention problems find it difficult to restrict attention to a limited spatial area.  We tested this conjecture within the context of visual search.


Participants were classified as having  low or high levels of impulsivity.  Their task was to search for a target embedded in arrow tip distractors. The arrow tips were organized such that they always pointed towards the target’s location.  Participants were either informed (i.e. Knowledge-driven participants) or  not informed (i.e. Image-driven participants) about the coded organization of the arrow tips.  In one condition, participants searched for the target without any hindrance to their peripheral vision (i.e. Large window condition).  In another condition (i.e. Small window condition), the same participants’ peripheral visual field was blurred by a gaze-contingent procedure controlled by an eye tracker.  Thus a 2(Impulsivity) X 2(Window condition) X 2(Informed about organization) mixed factorial design was utilized.   The measures of interest  were (i) the number of saccades made in search of the target (i.e. an index of search difficulty),  and (ii) the  percent of  saccades executed in directions not indicated by the arrow tips (i.e. an index of  participants’ use of the pointing information).  We used the term “errant saccades” to describe saccades that were not executed in the direction indicated by the arrow tips.


The results of the experiment indicated that a higher number of saccades were made by Image-driven participants than by Knowledge-driven participants (p<.05).   There was also a significant interaction such that the blurring of peripheral information increased numbers of saccades more for the Image-driven than for the Knowledge-driven participants (p = .01).   Thus, not surprisingly, the search task was more difficult for Image-driven participants.   With respect to the utilization of the pointing information in the arrow tip distracters, Image-driven participants had a higher percentage of errant saccades than Knowledge-driven participants (p= .001).   There was a trend towards a significant Impulsivity X Window condition  X  Informed about organization  interaction (p=.07). Blurring of peripheral information had no significant effect on Image-driven participants;  however, highly impulsive Knowledge-driven participants made increasingly higher percentages of errant saccades when peripheral information was blurred.  A reasonable interpretation is that these participants suffered a cost when the acquisition of useful information in their peripheral field was hindered. In contrast, Knowledge-driven  participants who were less impulsive made increasingly lower percentages of errant saccades when peripheral information was blurred.  For these participants, the information in the peripheral field was a hindrance, and blurring it facilitated performance.  Together, the results support the conjecture that  (at least some aspects of) attention deficit may be characterized by a wide visual attention window.  From our findings, one may reasonably propose that the extra information processed by a wide attention window  in attention deficit is not inherently a hindrance to visual processing.  Visual performance in attention deficit may be governed by an  interaction between task demands and the kind of information in peripheral vision.