Eureka: One shot viewing enables perceptual learning

Purpose. Early perceptual learning generally proceeds sequentially, in cascade fashion, beginning with improvement on easier type trials and only later, gradually, spreading to more difficult type trials. We now ask whether training adult human subjects (Ss) with sub-threshold stimuli only, would induce perceptual learning. Methods. The behavioral task was to detect whether there was an oddly oriented element (30 degrees difference) anywhere within a 7×7 array of homogeneous distractors. The limiting factor was the Stimulus Onset Asynchrony (SOA) between test stimulus and subsequent mask. We used a very brief 50ms SOA which was initially sub-threshold for almost all Ss. Results. There was virtually no improvement with these sub-threshold stimuli. Given reinforcement feedback, some Ss improved. For these Ss, improvement began only following several hundred trials, but then proceeded rapidly. We looked for a way of enabling rapid learning for all Ss, and at the onset of the practice session. We found that a single presentation of the stimulus, lasting several seconds, was sufficient to induce immediate onset of performance improvement, even without feedback. This one-shot presentation had to match the subsequent test stimulus: a schematic sketch of the stimulus would not do. We call this the Eureka phenomenon. Conclusions. For most Ss, practice, even with feedback, is not sufficient to induce perceptual learning. However, supplying clear knowledge of the expected image, is sufficient to initiate learning. Surprisingly, this information need not be given on a trial-by-trial basis and its application may be separated in time from the subsequent practice stage.