A Comparison Between Japan and the U.S. in Correspondence Bias : A Study of a Psychological Process Producing Misunderstanding of Different Cultures

本研究的目的是为了验证日本和美国的对应偏差的程度是否不同，以验证日本的对应偏差比美国弱的说法。为了评估对应偏差的程度，我们在日本和美国进行了问卷实验。问卷首先描述了发表政治声明的情况，然后描述了发表政治声明的情况，然后描述了发表政治声明的情况。然后要求被调查者推断出一个人的真实观点。在对1488名被调查者进行初步研究的基础上，我们编制了8种不同情况、不同执行力度和不同对应偏差程度的问卷。共有1157名美国大学生和712名日本大学生回答了调查问卷。“因此，在美国受访者中，只有那些有欧洲背景的人 ...更多信息 可以提供适当的数据。我们的美国受访者是旧金山弗朗西斯科州立大学的本科生，其中大多数人与有亚洲背景的人一样有非欧洲背景。在我们的1157名美国受访者中，有欧洲背景的只有291人。由于每种情况的答复人数只有18人左右，因此不可能通过统计检验得出明确的结论。本研究的初步结果如下：在两个故事中的一个故事（即，印度教和伊斯兰教之间的宗教冲突），美国受访者表现出一贯更强的对应偏见。在另一个故事中（即，然而，没有观察到这种趋势。在这两个故事中，调查问卷中确定的人应该做出的陈述的内容是相同的，只是专有名称被改变了（例如，从“穆斯林”到“塞尔维亚人”）。作为一个初步的结论，因此，这是推断，对应偏差的大小可能会更强烈地受到情境差异比文化差异。少

The purpose of this study was to examine whether the magnitude of correspondence bias was different between Japan and the U.S. to test an assertion that correspondence bias is weaker in Japan than in the U.S. We conducted a questionnaire experiment in both Japan and the U.S. to assess the magnitude of correspondence bias.The questionnaire first presented a description of a situation in which a political statement was made followed by the statement, and then asked respondents to infer a true opinion of a person who made the statement. Based on a preliminary study with other 1488 respondents, we prepared eight kinds of the questionnaire, which differed in the situation, its strength of enforcement, and the magnitude of correspondence bias. A total of 1157 American undergraduates and 712 Japanese undergraduates answered the questionnaires.The above assertion contrasted "European culture" and "Japanese culture." Among the American respondents, accordingly, only those with European backgrou … More nd could provide appropriate data. Our American respondents were undergraduates of San Francisco State University, in which the majority had non-European background as those who had Asian background. Among our 1157 American respondents, those with European background were only 291. It was impossible to reach a clear conclusion with statistical tests because the number of respondents was only around 18 per each condition. It is necessary to continue the study to obtain more respondents with European background.The tentative results were as follows : In one of the employed two stories (i.e., a religious conflict between Hindu and Islam), American respondents showed consistently stronger correspondence bias. In the other story (i.e., an ethnic conflict in Kosovo), however, no such tendency was observed. The contents of the statements that were supposed to be made by the person identified in the questionnaire were identical between these two stories except that proper names were changed (e.g., from "Muslims" to "Serbs"). As a tentative conclusion, therefore, it is inferred that the magnitude of correspondence bias may be affected more strongly by situational difference than cultural difference. Less