Physiognomy Of The Chinese People: Effects And Categories

Physiognomy is a specific phenomenon, but few studies have examined scientifically it in the term of psychological view in China. The purpose of the present study was designed to investigate the physiognomy effect, the valence of physiognomy, and the category of physiognomy.A total participant volunteered to participate in the study. The preliminary study was designed to obtain reliable experimental materials for Experiment 2. In Experiment 1 (N=100), a 2 (location of pictures:left vs. right)×2 (sources of pictures:the U.S. Senate 2002 vs. H.K. Legislative Council 1998) experimental design to test the physiognomy effect in the context of real election. The order of the pictures was presented randomly across the participants. Participants were asked to make a choice themselves in both pictures via a question (i.e. who is more competent?). Experiment 2 (N=565) was made up of three dimensions to investigate the valence of physiognomy and the category of physiognomy respectively. The first one (N=212) was a 2 (types of pictures:fugui-physiognomy vs. pinjian-physiognomy)×2 (gender:male vs. female)×2 (age:youth vs. middle-age) mixed design, and similar to the former, the second (N=164) was a 2 (types of pictures:junzi-physiognomy vs. xiaoren-physiognomy)×2 (gender: male vs. female) mixed design. The purpose of the both was for the valence. The last dimension (N=189) was a 4 (types of pictures:fugui-physiognomy, pinjian-physiognomy, junzi-physiognomy and xiaoren-physiognomy) design, which was to dissociate the category from the valence. These dimensions made the category clear step by step and in which participants were told to classify pictures among two/four options (i.e. types of pictures) based on their first impressions. In Experiment 3 (N=280), a 4 (types of pictures:fugui-physiognomy, pinjian-physiognomy, junzi-physiognomy and xiaoren-physiognomy)×3 (types of presentation: photo, photo-trait sentence pairs, trait sentences) mixed design was adopted to investigate the content of four-category physiognomy and further confirm the hypothesis. The types of pictures were the within-subjects designs, others were the between-subjects designs. All experimental materials were showed on computer randomly.Results in Experiment 1 showed that people could correctly predict election outcomes from facial appearance of the candidates. The proportion of correctly predicted races for the U.S. Senate was 57.61%, which indicated there was a cross-cultural effect on predicting social outcomes from faces. And the proportion for H.K. Legislative Council was 54.53%. They were both against the chance level of 50%. The findings suggested that the physiognomy effect was in perceiver's mind. The results from Experiment 2 were as follow:(1) People correctly rated categories for faces computer generated and in which the proportions were 66.61%,65.94% and 59.15% respectively in three dimensions. (2) The percentages of the first two dimensions (66.61%,65.94%) were higher than that of the third one (59.15%), which indicated that the valence might weaken the accuracy of rating category, but it never decrease them to chance level. The results from Experiment 3 showed that the content differences among four-category physiognomy were tested via four non-overlap radar maps, which also proved previous hypothesis.