1972 年 43 巻 1 号 p. 22-30
The purpose of the present study is to examine how the affective values of color and form influence the combined affective values of colored form objects when we perceive them.
In a dark room, the same Ss, 45 male students and 30 female students, were required to rate 12 colors, 5 line forms and 12×5 colored forms on 20 seven-point scales, according to the semantic differential method. The scales were used partly to represent the three dominant factors, Evaluation, Potency and Activity, discovered by Osgood (1957) in his factor analysis of verbal stimuli and partly to cover all the adjectives appropriate to evaluating the colors and forms, furnished by the report of Tanaka, Oyama and Osgood (1963).
In order to predict the affective values of colored forms from the affective values of colors and forms, which had been obtained separately, Multiple Regression Analysis for scales was applied in this study. This prediction was successful as indicated by high R: multiple correlation coefficient, out of 20, 16 R were higher than 0.90. The predicted values were also compared with those obtained according to Osgood's congruity principle and showed much greater fitness to the obtained data than Osgood's. The result of this Multiple Regression Analysis indicates that generally the color component has stronger influence on the affective values of colored forms than the form component.
As the next step, each of the three matrices, color, form and colored form stimulus group was factor-analyzed. Each analysis was based on the mean scale value of all Ss for each stimulus. The result of factor analysis denotes that color stimulus groups and colored form stimulus groups can have the similar semantic structure. The most salient factor in these analysis was Evaluation; Potency and Activity were not so dearly differentiated. Compared with color and color formed stimulus group, line form stimulus group is less differentiated in the semantic space.
As the third step, the relation between the results of these two analyses was examined. The scales in which color components have overwhelming influence on colored form components in the Multiple Regression Analysis were as follows:
beautiful-ugly (color 0.95 form 0.01)
warm-cool (color 0.92 form 0.16)
cheerful-gloomy (color 0.91 form 0.12)
like-dislike (color 0.95 form 0.00)
healthy-unhealthy (color 0.94 form 0.02)
gay-sober (color 0.92 form 0.13)
manly-womanly (color 0.91 form 0.30)
interesting-uninteresting (color 0.88 form 0.13)
strong-weak (color 0.88 form 0.29)
According to the results of factor analysis, all of the above scales were included in Evaluation. The scales in which form components have relatively strong influence on colored form are,
rounded-angular (color 0.20 form 0.96)
stable-unstable (color 0.52 form 0.77)
ordered-chaotic (color 0.49 form 0.71)
hard-soft (color 0.69 form 0.64)
smooth-rough (color 0.62 form 0.60)
dynamic-static (color 0.72 form 0.56)
These scales are included in Activity or Potency.