社会与工程心理学研究室知识库

以往研究发现个体主义/集体主义不同面向之间呈现弱相关或者不相关，且这种几乎为零的相关性并不是由于个体内水平的不稳定性所致。研究者们提出行为剖图假说来解释集体主义的不同面向在个体间不存在相关性而个体内存在跨时间一致性这一现象，认为集体主义是一个联结松散的行为剖图。以往研究主要在美国、加拿大等个体主义水平较高的西方文化进行，在日本也仅进行了个体间相关性的研究，没有进行个体内相关性的研究。中国同日本一样，均为集体主义文化的典型代表国家，本研究假设行为谱图假说在中国也同样适用。

研究一以横切面角度为主，从广东省某高中抽取了404名高中生进行纸笔测试，采用图片分类任务、框架直线任务、自我膨胀任务、归因任务、“亲亲性”任务和集体主义量表，证实了个体间不存在或者存在非常低的相关性(平均相关系数为0.072)。研究二从纵向角度出发，使用自我包括他人任务、图片分类任务、归因任务和“亲亲性”任务等测量工具，分为两个阶段:阶段一是通过电子调查问卷收集了708人(主要为山东省内在读专科/本科大学生)的数据，获得了同研究一相同的结果，文化任务之间的平均相关系数为0.05 ;阶段二是在三个月后通过重测获得了其中131人的第二次数据，除了归因任务相关系数为0.252之外，其他任务的重测相关系数为0.327至0.515，为中等相关，支持了行为剖图假说在中国的适用性。

本研究获得了同以往以西方文化为代表的国家相一致的研究结果，有以下贡献:第一，通过重复以往的类似研究，在中国得到了相似的实证结果，扩展了行为谱图假说在东方国家的跨文化适用性，为该假说的跨文化适用范围提供了更多的可能性。第二，在解释影响中国集体主义文化的因素时出现了很多理论，如现代化理论、传染病理论、大米理论和气候一经济理论，这些理论的研究结果并不总是一致。如Talhelm等人使用自我膨胀任务、“亲亲性”任务、图片分类任务等进行研究，提出以种植水稻为主的地区的互依我水平高于以种植小麦为主的地区;而Van De Vliert等人使用集体主义量表等测量工具，提出气候恶劣且缺乏充足经济资源的地区互依我水平更高，在将两种理论所采用的量表得分进行比较时，却出现了互相矛盾的现象。本研究的研究一同时采用Talhelm等人使用的自我膨胀任务、“亲亲性”任务、图片分类任务和Van De Vliert等人使用的集体主义量表进行研究，发现这些任务之间的相关性基本为0，为上述出现的矛盾现象提供了可能的合理解释。

结论:本研究通过个体间水平的低相关或者零相关以及个体内水平的稳定性这一实证结果，支持了行为剖图假说在中国的适用性。

According to previous studies, the correlations between measures of individualism and collectivism at individual level seem to be negligible; however, within-person variability was found to be cross-temporally stable, thus calling into question the unitary construct of collectivism. Researchers conceptualized individual differences through the lens of behavioral profile, which suggests that collectivism might consist of loosely connected, cross-temporally stable subdomains. Previous studies in this field were mostly conducted in western cultures such as America and Canada, and only interindividual variability was examined in Japan, representing eastern cultures. Similar to Japan, China is an interdependent-culture in East Asia. Herein, we proposed that behavioral profile hypothesis could also find its support in Chinese culture. We replicated previous studies, obtaining the consistent results.

Study 1 was a cross-sectional study, which included 404 students from a senior high school who participated in a paper-pencil test, consisting of following 6 cultural tasks: triad task, Framed Line Task, self-inflation scale, attribution task, loyalty/nepotism task, and Van De Vliert's collectivism scale. The results showed negligible or very low correlations (mean r=0.072) across these tasks. Study 2, which was a longitudinal study with two time portions (time 1 and time 2), included 708 participants who filled out an online questionnaire in the first portion of the study. The questionnaire included 4 cultural tasks: inclusion of other in the Self Scale, triad task, attribution task and loyalty/nepotism task. The obtained results were in line with those obtained in Study 1，where the mean correlation was 0.05. Among these 708 participants, 131 were included in the second portion and filled out the same questionnaire on line after 3 months, after which the test-retest analysis was made based on the data collected from 131 participants at time 1 and time 2. The size of test-retest correlations was moderate (0.3277<0.515), except for one task with correlation coefficient of 0.252. As predicted, there were interindividual variability and intraindividual cross-temporally stability observed in Chinese samples.

The present study results were consistent with previous studies conducted in western cultures. The contributions of the current research are as followings: First, we replicated similar researches, which revealed consistent results in China, thus obtaining support for loosely-connected behavioral profile beyond America and Canada, and increasing its cross-cultural university. Second, there are different theories explaining factors influencing collectivism in China such as modernization theory, pathogen prevalence theory, rice theory and climate-economic theory. Each theory was based on different measures, which led to different results. E.g., Talhelm et al. adopted self-inflation scale, loyalty/nepotism task, triad task and proposed that individuals from rice counties were more interdependent than individuals from wheat counties; Van De Vliert et al. used collectivism scale and proposed that individuals having to cope with climatic demands without adequate resources were more interdependent. However, these two propositions seemed contradict with each other when comparing the scores measured by different tasks according to these two different theories. Herein, we measured individual attitudes by using self-inflation scale, loyalty/nepotism task, triad task and Van De Vliert’s collectivism scale in the same research (Study 1), finding that correlations across tasks were negligible, with one exception of a low correlation. This suggests that different culture tasks/different facets may explain contradiction of the propositions, so more facets of collectivism should be considered and adopted in the cross-cultural study of different regions in China in the future so as to obtain more robust results.

Conclusion: The present study demonstrated intra-individual cross-temporally stability among various facets of collectivism in spite of zero or low interindividual correlations, which gave a clear support for the application of loosely-connected behavioral profiles in Chinese culture.