Recently, our lab has begun to explore the nature of perception without attention. This research follows on recent studies by Mack and Rock as well as earlier studies by Neisser and colleagues. Our primary interests are in exploring the role of attention and inattention in the perception of complex dynamic events. These projects use both video and computer-based paradigms to explore the mechanisms underlying inattentional blindness.

      • Click here for a brief overview of inattentional blindness.

Sustained Inattentional Blindness for dynamic events
       In a series of earlier studies (see overview), Neisser and colleagues showed that when observers were actively engaged in an unrelated task, they sometimes failed to see a woman carrying an umbrella walk across the screen (an unexpected event, or UE). In those studies, observers pressed a key whenever one team of basketball players made a pass, while they simultaneously ignored passes made by another team. The displays showed both teams superimposed on a single display such that both displays were partially transparent, giving the events a ghostly appearance. In our lab, we have replicated the earlier studies and have extended them by varying the difficulty of the distraction task, the similarity of the UE to the ignored team, and the nature of the UE. We have also conducted studies with a choreographed version of the task in which both teams of players and the UE were filmed with a single camera shot. In these displays, the UE is more clearly visible and salient. As in the earlier studies, we find that observers often do not report having seen the UE. Even in the choreographed version, a substantial proportion of the observers show inattentional blindness for the UE. Ongoing studies are continuing to examine the factors underlying this failure to see the UE.

Visual Similarity, Directed Ignoring, and Primitive Features
       Another related series of studies uses a computer analogue of the video task described above. In this case, observers simply track animated shapes moving in random directions on a computer display. They are required to perform various counting tasks, ignoring some of the shapes while attending to others. On the fourth trial of this task, an unexpected event occurs: another shape moves across the screen. Thus, the procedure used in this task is essentially equivalent to that used by Mack & Rock (see overview) except that the UE lasts for 5 seconds rather than 200ms. Experiments in this project investigate the role of attention in perceiving so-called "primitive" features. What sorts of features are likely to be perceived without attention? This approach to exploring pop-out may be superior to visual search in that subjects devote no attention to the possible appearance of a target. The paradigm allows an assessment of what is perceived without attention. We can also directly examine the role played by the ignored items, essentially exploring what we call "directed ignoring." The precise control over the displays gained by computer implementation allows a more systematic exploration of the same issues raised by the video task. In combination, these approaches allow more wide-ranging generalizations about attention and inattention than either could provide alone.