Much research has established that causal attributions differ cross culturally. A classic finding is that Americans are prone to make fundamental attribution errors, whereas Chinese, Indians and Koreans are more likely to attend to situations (Morris peng, shweder)
The proposed mechanism: implicit theories that is present in American vs. Chinese society. Because Americans are more independent, so they are more likely to see individuals as agents. While Chinese are more interdependent (markus & kitayama), so they are more likely to see holistic influences on people’s behavior.
Given this, one related question is that how gender affects our understanding of the causes of male and female’s action. There is significantly less research on this subject. Indeed, the answer may exist, but it has not been analyzed. For instance, Morris and Peng’s classic paper doesn’t report the number of males and females surveyed in US and China or their likelihood to follow particular attribution patterns.
•Men are more likely to establish internal attribution.
•Women are more likely to establish external attribution.
In study 1, I have 200 participants from a summer school in Special A. Male and female were divided into two groups randomly. Then, I spread out two versions of questionnaires, which were totally the same except for whether male or female got a low score in school. The first group of participants got the version that female got a low score in school. And the second group of participants got the version that male got a low score in school. Thus I can distinguish difference of attribution, including male explained male’s behavior, male explained female’s behavior, female explained male’s behavior, and female explained female’s behavior. These two questionnaires were both in Chinese version so that students in China could understand the authors’ meaning accurately.
Authors used a scenario to determine whether participants were internal locus of control or external locus of control through their score. For example, if a participant chose 6 points in bad luck and 1 point in unintelligence, the authors could tell that this participant has more external locus of control.
Summer School Sample
Subjects: We sampled 200 high school students in a summer school.
Materials: Two types of questionnaires that were totally the same except for whether male or female got a low score in school.
(A) Why LiJuan got such a low score?
Li Juan has achieved one of the lowest scores in the school on this exam. There are lots of reasons for why people do poorly in schools. Why do you think she has done so poorly?
(B) Why WangPeng got such a low score?
Wang Peng has achieved one of the lowest scores in the school on this exam. There are lots of rea-sons for why people do poorly in schools. Why do you think he has done so poorly?
I looked at full set attribution items and performed factor analysis to reduce the items. I found 3 constructs emerged: internal malleable, internal stable, and external. I then created three new variables, internal malleable(intelligence and ability); internal stable (work hard and use effort in the exam); and external (four external factors including luck, unpredictably, exam is difficult, and exam is challenging).
I performed an ANOVA with 3 factors, respondent gender x target gender x language. We considered three dependent variables above: internal stable attribution, external stable attribution, and internal malleable.
1). Internal attribution
I found no main effect of respondent gender, target gender, and language. I found 2 way interactions between respondent gender x language: males and females are more internal in English than Chinese (p<.05). No other two way interactions were significant. I then found a marginal three-way interaction of respondent gender x target gender x language. There was an marginal internal trend. (p<.10)Females are more internal about males in English and more internal about females in Chinese. Males are more internal about male in English and more internal about female in Chinese.
2). External attribution
I found following main effects external attribution:
a. Male attributors are more external(marginal p=.07)
b. Attribution in English is more external (p=0.7). This is contrary to Morris & Peng.
Two way interaction:
respondent gender x language interaction: Males and females are more external in English. (p=.02) This is contrary to Morris & Peng.
No three way.
3). Internal malleable.
No effects significant.
We can see that Attribution in English is more external (p=0.7). Why language, which should prime cultural accessibility, produced opposite effect? Perhaps Chinese have stronger judgments and more confidence in Chinese language about norms about school work. Also, maybe Chinese have more pressure in their school life. So, Morris&Peng’s general idea should be more specific to domain considered. Thinking for Chinese at least about school performance, Chinese increased internal attribution, judgment, and harshness.
Second, 3 ways interaction show that males are more internal attribution about females in Chinese. Maybe males have more prejudice and stereotype to women in Chinese, however, prejudice and stereotype maybe mitigate when participants answer questions in English. (Chart 1 & 2) Females are more internal attribution about females in Chinese, maybe because there are lots of comparison and competition between females. External attribution also has similar effect.