Home‎ > ‎Project‎ > ‎

Study 1 - Infrahumanisation effect in Japan

This study was conducted in order to test whether group members believe embarrassment (a complex, human emotion) to be more typical of their own group than a higher- or lower-status outgroup. So far, most studies on the so-called infrahumanisation effect have examined whether ingroup members rate a number of different emotions indicating human uniqueness as more typical of the ingroup than outgroups. However, some researchers believe that embarrassment is the most human emotion of all.


An online study was conducted with the software Surveymonkey. Participants were recruited via online forums, as well as with a snowballing technique starting with family, friends, and acquaintances of a Japanese research assistant. The survey was translated from English to Japanese, and re-translated into English to check for accuracy. The final online survey was presented in Japanese.

Design: 3 within-subjects conditions (i.e., each participant rates emotions for all three groups): ingroup (Japanese), higher-status outgroup (Americans), lower-status outgroup (Chinese)

Participants: = 211 Japanese participants (63.5% women; Age: 18-75 years, M = 39, SD = 13; Occupational status: 59% employed orself-employed, 26% at home or unemployed, 16% students, 1% retired)

Manipulation check for status: A single item assessment of status adapted from Major, Gramzow, McCoy, Levin, Schmader, and Sidanius (2002) was used. The relative status of each national group was assessed by asking participants: "There are many people who believe that different national groups enjoy different amounts of social status in this society. You may not believe this for yourself, but if you had to rate each of the following groups as such people see them, how would you do so?" Participants could then rate the social status for Americans, Chinese, and Japanese individually on a 7-point scale, from 0 (= low social status) to 6 (high social status). 

Typicality ratings: Following the typicality rating of Viki and Abrams (2003), participants were asked the follwoing: "Of all the characteristics listed below please tick those that you think would be most typical of Americans, Chinese, and Japanese, respectively. You can choose as many as you like but please try not to pick too many." Then, a table followed that listed in the left hand column 3 positive primary (i.e., "simple") emotions (pleasure, surprise, happiness), 3 negative primary emotions (fear, aversion, anger), 3 positive secondary (i.e., "complex") emotions (compassion, hope, nostalgia), 3 negative secondary emotions (guilt, resignation, shame), and embarrassment (a negative secondary emotion). Emotions were presented in random order. Next to each emotion, participants could tick a box whether the emotion was typical of Americans, Chinese, or Japanese, respectively. Each box that was ticked was coded as 1, each un-ticked box was coded as 0. A summary score was computed for positive primary emotions, positive secondary emotions, negative primary emotions, negative secondary emotions, and embarrassment.


Manipulation check: As intended, Japanese participants rated the higher-status outgroup Americans (M = 5.25, SD = 1.28, simple contrast: F(1,152) = 44.44, p < .001) to be higher in social status than the Japanese (M = 4.46, SD = 1.27), and the lower-status outgroup Chinese (M = 3.25, SD = 1.37, simple contrast: F(1,152) = 66.76, p < .001) to be lower in social status than the Japanese (repeated-measures ANOVA: F(2,304) = 103.74, p < .001).

Typicality of Emotions
Embarrassment: As expected, Japanese participants rated embarrassment to be significantly more typical of the ingroup Japanese (M=0.51, SD=0.50) than of the higher-status outgroup Americans (M=0.02, SD=0.14, simple contrast: F(1,210) = 204.11, p < .001), or the lower-status outgroup Chinese (M=0.10, SD=0.31, simple contrast: F(1,210) = 90.93, p < .001; repeated measures ANOVA: F(2,420)=119.68, p<.001; see Figure 1 a).

Overall infrahumanisation effect: A repeated-measures ANOVA with factors valence (positive, negative), type (primary, secondary), and group (ingroup, higher-status outgroup, lower-status outgroup) yielded the expected 3-way interaction between valence, type, and group (F(2,420)=72.29, p<.001), a 2-way interaction between valence and type (F(1,210)=8.02, p<.01), valence and group (F(2,420)=65.26, p<.001), type and group (F(2,420)=86.33, p<.001), and a main effect for valence (F(1,210)=27.09, p<.001) and type (F(1,210)=20.76, p<.001), but no main effect for group (F(2,420)=1.12, p>.10). Post-hoc simple contrasts revealed that positive primary emotions were seen as significantly more typical of the higher-status outgroup Americans (M = 1.16, SD = 1.13) than of the ingroup Japanese (M = 0.48, SD = 0.79, simple contrast: F(1,210) = 78.26, p < .001) or the lower-status outgroup Chinese (M = 0.41, SD = 0.67, simple contrast: F(1,210) = 94.22, p < .001), with no significant difference between the latter two groups. Negative primary emotions were rated to be more typical of the lower-status outgroup Chinese (M = 1.16, SD = 1.13) than of the higher-status outgroup Americans (M = 0.45, SD = 0.77, simple contrast: F(1,210) = 83.13, p < .001), or the ingroup Japanese (M = 0.28, SD = 0.63, simple contrast to lower-status outgroup: F(1,210) = 108.95, p < .001; simple contrast to higher-status outgroup: F(1,210) = 8.07, p < .01). In contrast, positive secondary emotions were reported to be more typical of the ingroup Japanese (M = 0.72, SD = 0.81) and the higher-status outgroup Americans (M = 0.72, SD = 0.82) than of the lower-status outgroup Chinese (M = 0.52, SD = 0.73, simple contrast with ingroup: F(1,210) = 9.24, p < .01; simple contrast with higher-status outgroup: F(1,210) = 7.85, p < .01). In line with results for embarrassment, negative secondary emotions were seen as more typical of the ingroup Japanese (M = 1.01, SD = 1.19) than the higher-status outgroup Americans (M = 0.16, SD = 0.44, simple contrast: F(1,210) = 106.62, p < .001), or the lower-status outgroup Chinese (M = 0.23, SD = 0.63, simple contrast: F(1,210) = 74.45, p < .001).

Figure 1a. Typicality of embarrassment by group
Note. Error bars represent standard errors for each condition.

Fgure 1b. Typicality of primary/secondary emotions by group
Note. Error bars represent standard errors for each condition.


Embarrassment, an emotion unique to humans, was rated by Japanese participants as more typical of the ingroup Japanese than an outgroup, independent of whether that outgroup was higher in status (Americans) or lower in status (Chinese). This finding is in line with the infrahumanisation effect, which was also replicated with more commonly used positive and negative primary and secondary emotions. Embarrassment is a social emotion which helps to keep social behaviour in check. Such an emotion is likely to be highly valued in interdependent cultures, such as Japan. Thus, embarrassment may indeed be more typical for the Japanese than for more independent cultures, such as Americans. However, the replication of the infrahumanisation effect with other emotions in a Japanese context suggests that the finding regarding embarrassment may not only be due to cultural differences, but may indeed reflect an infrahumanisation effect. 

A stricter test of infrahumanisation in embarrassment, in a different cultural context (Scottish university affiliations), and with attributions of embarrassment in particular situations, rather than overall ratings of typicality of that emotion for different groups, may shed more light on this effect. We therefore conducted a second study on the infrahumanisation effect in embarrassment.