Abstract
Extended straight trajectories are easily detected in dense motion noise (Watamaniuk et al., 1995). The first part of the trajectory appears to direct attention to the subsequent segments, leading to this enhanced detectability. Our evidence for this cueing effect is that a contrast increment is far more detectable at the end of a trajectory sequence than at the beginning. We have shown that the improvement in thresholds at the end of a single trajectory is consistent with a reduction in uncertainty; the observer is monitoring far fewer motion detectors. What happens if there are competing trajectories? Using this contrast probe technique, we measured increment thresholds at the beginning and end of one 200 ms trajectory presented in dense motion noise when a second distracting trajectory was presented simultaneously. The distracting trajectory was either 100 ms or 200 ms in duration, and was initiated at the same time as the test trajectory. The improvement in increment thresholds at the end of the test trajectory depended on the strength of the competing distractor. For the case of the 200 ms distractor there was little difference in increment thresholds for the beginning and end of the test trajectory; for the 100 ms distractor there was a modest improvement for the end segment, compared to the phenomenal improvement in thresholds in the absence of a distractor. The distracting trajectory prevents the observer from focussing attention on the subsequent parts of the test trajectory, producing little reduction in uncertainty.