Experience and latency to achieve stereopsis.

This study assessed perceptual learning effects with complex random-dot stereograms. Observers were shown the same complex anaglyph five times daily, for four consecutive days, and latencies to achieve stereopsis were recorded. Two-factor repeated-measures analysis of variance yielded significant effects of days, trials, and days-by-trials interaction. Latency to achieve stereopsis decreases over trials each day, but this decrease is not completely transferred across days. It is concluded that obervers must, in some sense, "re-learn" how to perceive complex stereograms if subsequent presentations occur over more than one day.