圖形殘効の實驗的研究 (I)

時間的要因について

抄録

Figural after-effects depend upon two temoral factors, namely the duration of inspection period (i) and the time interval between inspection and test periods (t). This study was planned to measure figural after-effects (A) as a function of two variables, i and t, whereas, up to the time, many experimenters (1, 3, 6, 12) had studied the after-effects as a function of one variable, either i or t.
Experimental conditions and procedure : Black outline circles on white paper were used as the inspection- and test- objects. (See Fig. I) The right test-circle was the standard stimulus and the left circle was the variable stimulus. Six test-sheets with various sizes of the left circle were prepared. But, after each inspection period only one test-sheet was used. The inspection- and test-sheets were presented at the distance of 3 meters from the subject, who observed them binocularly. The inspection- and test-sheets with their surroundings of white paper, were illuminated homogeneously at the brightness of 2 millilamberts. The durations of the inspection periods were 1, 2, 5, 15, 60 and 240 seconds. The simplified “method of comleete series” was used. As the measure of figural after-effects, the amount of apparent shrinkage of the test-circle in diameter was adopted.
Results and discussion : The upper curve in Fig. 2 shows the time course of development of the figural after-effects that are tested immediately after the inspection periods. It indicates that a 1 second inspection is long enough to produce a considerable amount of the figural after-effects and the prolongation of the inspection period can hardly bring about the increase in them. The lower curve shows the time course of development of the figural after-effects that are tested 5 seconds after the inspection period. It froms an ordinary growth curve. Since these two curves are obtained from the same series of experiments, the difference between the two should be attributed to the difference in t, that is the sole difference between their conditions. It is very interesting that Gibson & Rander and Hammer, who used the method of adjustment, which would require at least 5 seconds for one setting, reported the results which resemble our lower curve rather than our upper curve.
The curves in Fig. 3 show the time courses of disappearance of figural after-effects after the inspection periods of various durations. All of these curves start from almost the same level and decreasing with negative acceleration reach zero. Their qualitative characteristics are similar to one another and to those of Hammer's curve. Inspection of Fig. 3 reveals that, in general, the long r the inspection period is, the slower is the rate of the decrease in the after-effects.
The schematic representation of our results requires a three-dimensional graph like Fig. 4. And, according to the mathematical formulation of Mueller, who used Hammer's data, the disappearance course of figural after-effects was represented by the exponential function, A=A_oe-kt. Our results would suggest that in this formula the parameter k should be taken as a function of i and A_o a constant which is independent of i.