Abstract
Goals: We investigated how practicing a task that requires attentional selection of one component in bivectorial transparent dot motion affects observers sensitivity for different motion directions.
Methods: We measured observers 75%-correct coherence threshold for motion detection in eight major directions pre- and post- learning. During the learning phase, observers underwent 6 hourly sessions of speed discrimination involving two transparently moving families of dots moving either in opposite or orthogonal directions. Observers' task is to attend to dots in the target direction and indicate the speed change (faster or slower) in the middle of the trial, while ignoring the other (distracter) direction. In the control condition, observers discriminate speed change with a single motion direction.
Results: In the case of opposite motion, although practice led to an increased sensitivity for both the attended and the neglected motion direction, the sensitivity enhancement was significantly larger for the attended motion direction. Practicing with an orthogonal transparent motion display, however, resulted in an enhanced sensitivity for the attended direction and the direction between the attended and neglected motions, but not for the neglected direction. In the control condition - single motion direction during practice - we found a Mexican-hat shaped modulation of the sensitivity, with the strongest sensitivity enhancement at the practiced motion direction.
Conclusions: Our results provide evidence that learning-induced changes in sensitivity for different motion directions - present during practice - are strongly affected by attentional selection and reveal the direction tuning of such attentional modulation.