Proactive-retroacrive effects in the successive comparison of visual extents III

Abstract

The purpose of the present investigation was to examine the effect of an extra line presented at various spatial loci upon the successive comparison of the lengths of two lines. Besides, the variation of this effect due to the changes of the time intervals between three lines was studied. Three lines were the black ones on a white background in horizontal position. The presentation of these lines were designed as follows. Three lines were presented in succession for the same brief duration, but the time intervals between lines were varied within a short period. A standard line (N) of 30 mm in length preceded the comparison line (V) and they were presented at an identical location. But the extra line (e) was presented at 0, 0.5, 1, 2, 3, 5 or 7 cm distant from the compared lines and e preceded N and V, followed them, or was interpolated between them.
The effect of the extra line was measured by comparing PSE obtained by the method of complete series with that of the control experiment where the extra line was not presented. As the-experimental results obtained by the presentations of three lines at an identical location were similar to ones reported in the previous papers, they were not mentioned here.
The chief Results were as follows :
I. Proactive effect of e in the sequence of e-N-V
(1) When e was 60 mm in length and the time intervals of both e-N and N-V were short, V was underestimated. The amount of this effect diminished as the spatial distance between e and N (V) was enlarged. But when the time intervals were more lengthened, V was overestimated.
(2) When e was 15 mm, V was overestimated at near position to e, while it was under-estimated at distant position from e.
II. Retroactive effect of e in the sequence of N-V-e
(1) When the time intervals of both N-V and V-e were short, V was overestimated under both conditions of 60 mm and 15 mm lengths of e.
(2) When the time intervals were relatively long, V was underestimated regardless of the length of e.
III. Proactive-retroactive effects of e in the sequence of N-e-V
(1) When e was 60 mm and the time intervals of either N-e or e-V were longer than that of the other, V was overestimated at near position to e.
(2) When e was 15mm and the time intervals were short, V was overestimated at near position to e, but it was underestimated at distant position from e.
The above results were ascribed to an assimilation-contrast phenomenon. The present investigation indicates that the assimilation-contrast phenomenon varies not only with the temporal intervals between lines but with spatial intervals between them.