Abstract
Previously, we showed that viewing only one trial per stimulus condition leads to significant perceptual learning in a face-identification task. Here, we ask whether such rapid learning is a general phenomenon, or whether it depends on particular aspects of the stimulus and task. Specifically, are similar results obtained with non-face visual stimuli, and does learning transfer across different kinds of stimuli? Observers performed an identification task on two consecutive days. Identification thresholds were measured by varying stimulus contrast using the method of constant stimuli. Experiment 1 manipulated the amount of practice observers received on Day 1in a 10AFC texture-identification task. The textures were band-limited noise patterns displayed at one of seven contrasts, in one of three external noise levels for a total of 21 conditions. On Day 1, observers received 0 (control), 1, 5, 10 and 40 trials/condition. Observers viewing fewer than 40 trials/condition performed an unrelated task for the rest of the session. On Day 2 all observers received 40 trials/condition. Observers who practiced 10 and 40 trials/condition performed equally well on Day 2, and significantly better than observers who practiced 0, 1 or 5 trials/condition. Unlike our results with faces, small amounts of practice did not elicit learning: performance in the 1 and 5 trials/condition groups was no different than the control group. In Experiment 2 observers trained for the entire session on either the face or texture-identification task on Day 1 and then transferred to the other task on Day 2. We found no transfer of learning in either case: Thresholds on Day 2 were the same as those from observers performing the tasks for the first time. Our results suggest that rapid learning is not a general phenomenon, but differs as a function of stimulus class. It may be that our past experience “jumpstarts” perceptual learning for faces, more so than for other less frequently encountered stimuli.
Support: NSERC PGS-A award (ZH), NSERC Discovery grants (42133 & 105494), and the Canada Research Chair Program.