Abstract
Goal: Perceptual learning is the improvement in performance that results from practice with simple tasks. Such learning is highly specific to basic stimulus attributes, including spatial frequency, orientation and spatial location. Last year we reported that both voluntary and involuntary covert attention can speed and strengthen learning, and transfer it to different target-stimuli. Here we ask whether attention would also help perceptual learning transfer to new spatial locations.
Methods: Observers trained for five consecutive days with an orientation discrimination task (left vs. right) using two Gabor stimuli on the horizontal meridian. For four observers each trial began with a spatially neutral cue. For four other observers an uninformative spatial cue preceded the display; half the time it preceded the target (valid cue) and half the time it preceded the distractor (invalid cue). For a third group of four observers, a spatial cue always preceded the target (valid cue). Critically, the transfer task was identical to the training, except that for five consecutive days, observers reported the orientation of Gabors that were either at the original location or were shifted above and below their original locations.
Results and Conclusion: When observers began the transfer task at the new stimulus locations, nearly all learning was lost for the neutral and invalid cue conditions, as observers' performance levels dropped back to threshold level. However, when valid cues were used, the performance decrement was significantly ameliorated (for observers who trained with valid and invalid cues) and completely absent (for observers who trained only with valid cues). These results support the finding that attention strengthens perceptual learning for simple visual stimuli; crucially, they indicate that exogenous attention transfers learning to new stimulus locations.
Supported by NIH R01 EY016200.