Ballistic and controlled processes in the Attentional Blink effect

The Attentional Blink (AB) effect occurs in a Rapid Serial Visual Presentation (RSVP) of digits, two of which are targets. The subject’s task is to report these targets. When the first target is reported correctly, the results show a drop of performance on the second target for lags of two to five. We explored the possibility that a ballistic process causes the AB effect. The alternative hypothesis states that the subject could stop the processing of the first target and identify the second. To test this hypothesis, we presented a RSVP to 32 subjects where the position of the first target was fixed. The subjects were informed of that position. In a second experiment, we repeated the same design but instead fixed both targets positions in the first condition and the distance (lag) between the two targets in the second condition. The results show no effect of this manipulation on the AB effect. However the number of inversion errors at lag one was reduced when the subjects knew the positions of the first and second targets compared to the cases where the subjects knew only the distance between the two targets. These results question the ballisticity assumptions of the AB