@article{oai:nagoya.repo.nii.ac.jp:00002095,
 author = {林, 文俊 and HAYASHI, Fumitoshi},
 journal = {名古屋大學教育學部紀要. 教育心理学科},
 month = {Dec},
 note = {Personality impressions are often formed on the bases of physiognomic cues. The present investigation aims at examining which physiognomic features are, in general, assumed to be related to which personality traits. Female undergraduates served as Ss. Each S was presented five stimulus persons (Sps) taken randomly from 90 characters picked up from various comic books, and was asked to rate them both on 20 five-point scales of physiognomic features and on 20 seven-point personality scales. N of Ss was 216 : each SP was assigned to twelve Ss. Prior to the major statistical analysis, mean ratings for each scale by each SP were calculated. Major findings obtained are as follows : 1. Of the 400 correlations between 20 physiognomic by 20 personality mean ratings (N=90,the same as the number of SPs), 163 were significant at the 1% level. The signs and values of these significant correlation coefficients are comparable with those obtained in our previous investigation (Ohashi et al., 1976,1977). 2. From a factor analysis of mean ratings of the 20 physiognomic scales, four factors were extracted. They were interpreted as (I) bushy eyebrows and masculine face, (II) round and flat face. (III) bright eyes, and (IV) knitted eyebrows and tightened mouth, respectively. In a similar vein, three factors were extracted from the mean ratings of the 20 personality scales. They were interpreted as (1) intimacy, (2) social desirability, and (3) dynamism, respectively. 3. Calculating each SP's factor scores for these seven factors, correlations between physiognomic and personality factors were obtained. The factor (1) correlated positivery with (III) and negativery with (IV), the factor (2) negatively with (II), and the factor (3) positivery with (I), (III), and (IV). 4. Multiple regression analysis was applied to the above data. The following equations were obtained : Y^^^ (1) = 0.63X(III)-0.62X(IV), R^2=0.72 Y^^^ (2) = -0.47X(II), R^2=0.21 Y^^^ (3) = 0.52X(IV)+0.50X(I)+0.46X(III), R^2=0.72 As to personality factor (1) and (3), it is possible, in some measure, to predict each factor score of SP by means of a linear combination of four physiognomic factors. But the factor score of SP on personality factor (2) is not predictable by the same way. 5. In addition to the multidimensional approach mentioned above, a typological approach based on a cluster analysis was attempted. The results obtained, however, was not clear., 国立情報学研究所で電子化したコンテンツを使用している。},
 pages = {41--55},
 title = {<原著>相貌と性格の仮定された関連性 (3) : 漫画の登場人物を刺激材料として},
 volume = {25},
 year = {1978}
}