Feature-based attentional set as a cause of traffic accidents

Abstract Voluntary and relatively involuntary subsystems of attention often compete. On one hand, people can intentionally "tune" attention for features that then receive visual priority; on the other hand, more reflexive attentional shifts can "short-circuit" top-down control in the face of urgent, behaviourally relevant stimuli. Thus, it is questionable whether voluntary attentional tuning (i.e., attentional set) can affect one's ability to respond to unexpected, urgent information in the real world. We show that the consequences of such tuning extend to a realistic, safety-relevant scenario. Participants drove in a first-person driving simulation where they searched at every intersection for either a yellow or blue arrow indicating which way to turn. At a critical intersection, a yellow or blue motorcycle—either matching or not matching drivers' attentional set—suddenly veered into drivers' paths and stopped in their way. Collision rates with the motorcycle were substantially greater when the motorcycle did not match drivers' attentional sets. Portions of this paper were presented at the 2005 meeting of the Vision Sciences Society. Portions of this paper were presented at the 2005 meeting of the Vision Sciences Society. Acknowledgments We thank M. Chun, D. Simons, B. Scholl, and K. Curby for helpful comments and discussion. This research was conducted at Yale University, where S. Most was supported by NIH grant F32 MH066572. Notes Portions of this paper were presented at the 2005 meeting of the Vision Sciences Society. 1Software enquiries can be directed to the second author at robert.astur@yale.edu 2Even had no participants been discarded, the difference in collision rates between the match and mismatch conditions would have remained significant, p<.05.