Optical-geometrical characteristics of fragmented figures and integral perception thresholds in repeated tests.

The causes of reductions in the integral perception threshold for fragmented figures in repeat trials using the Gollin test were investigated. The first study involved four trials with 5-day intervals using one set of figures. The amplitude-frequency characteristics of the test stimuli were interpreted using correlation analysis, as were the thresholds in each trial. The second study involved two trials with a three-day interval. On repeat testing, half of the figures were known (from the first test) to the subjects. In addition, half of the figures (both familiar and unfamiliar) were fragmented such that their amplitude-frequency characteristics had values similar to those in the first test. The results showed that the decrease in thresholds were mainly due to optimization of the functioning of the mechanisms operating with the statistical rather than the subject characteristics of the images.