Psychologists have developed an awareness that they need to exercise caution in discussing supposedly universal traits or patterns of behavior. Even apparently simple cognitive acts like categorizing stimuli or remembering details of an experience involve important cultural components. For instance Ji, Zhang, and Nisbett (2004) reported that Chinese participants may group words according to how they relate whereas American participants group them according to taxonomy (e.g., in a monkey–banana–panda triad, Chinese participants group monkey–banana, but Americans pair monkey–panda).
Considerations of complex personality issues raise even more difficulties. For example, researchers have assumed that the need for self-esteem is a universal trait because the well-established body of literature has consistently shown such a need. However, that literature is based on North American culture; when psychologists have investigated self-esteem among the Japanese, the results diverge in critical ways from those involving North American participants (Heine, Lehman, Markus, & Kitayama, 1999). In fact, self-esteem as conceptualized in Western psychology may be fairly irrelevant in Eastern psychology.
Given that psychological theory is typically Western-based, the constructs that psychologists use and even the language may make it difficult to understand cultural differences and even to speak about them. Thus, researchers may be discussing what they think are etics that are really emics. In fact, there is still disagreement as to whether some psychological constructs are basically universal with different cultural manifestations or whether they are largely cultural (Smith et al., 2006).
One of the dimensions receiving considerable attention because of its cultural relevance involves the role of individualism and collectivism. According to Matsumoto and Yoo (2006), this is the most widely studied dimension in the field, and varied research projects have illustrated consistent behavioral differences associated with the individualistic-collectivist (IC) continuum. It will serve as a useful illustration of the nature of the problems associated with studying behaviors across cultures.
An initial caveat in discussing the effects of one’s location on the IC continuum is that one must recognize that attributing differences in behavior to IC orientation is difficult because the investigator must be able to separate effects of culture (e.g., IC orientation) from other sources of variability. On an ecological level, there is a myriad of factors that affect behavior in different countries including affluence, population density, religious practice, and climate. As psychologists define culture, these ecological variables stand apart from cultural variables (Matsumoto & Yoo, 2006, p. 237).
Research on individualism and collectivism has revealed that people from collectivist cultures remember events differently than people from individualistic cultures. Individualist people tend to remember situations from the viewpoint of themselves as part of the situation. On the other hand, collectivist people tend to focus on the social situation with the perspective of an outsider looking in. A typical interpretation is that individualists focus on self and collectivists focus on other. If this difference in memory is true, then simple measurement of memories may be problematic across cultures because of the different ways that people conceptualize their world (Cohen & Gunz, 2002).
However, even if one were to resolve this dilemma, another problem arises in cultural research. Multiple cultures may exist and notable individual differences appear within a given country; unfortunately, researchers often equate country with culture, making the assumption that what is true generally within a country reflects differences in culture compared to another country. Thus, Iwata and Higuchi (2000) studied state and trait anxiety, finding that their Japanese participants had less positive views of themselves and higher levels of both state and trait anxiety than Americans. They attributed the variation across countries to differences in Japan and the United States on the IC continuum. But, as Matsumoto and Yoo (2006) pointed out, Iwata and Higuchi interpreted their data under the assumption that Japan is a collectivist society in which people are compliant in order to maintain social harmony and underestimate their own positive traits.
Matsumoto and Yoo identified seven assumptions that Iwata and Higuchi made regarding Japanese people. None of these assumptions was empirically tested. Matsumoto and Yoo did point out that Iwata and Higuchi’s assumptions may be correct and that differences between Japanese and American students resulted from the relevant differences on the IC continuum. Their point was that psychologists should engage in research to verify such assumptions. The attribution of differences across groups to cultural factors when there is no empirical support for such an interpretation is known as the cultural attribution fallacy, a specific case of a larger problem that Campbell (1961) called the ecological fallacy.
Interestingly, Iwata and Higuchi (2000) couched their discussion of the Japanese-American differences using the American pattern as the norm. This type of inference poses behaviors of Americans as a standard against which others are compared, which Arnett (2008) has pointed out is common in discussing and interpreting research findings. This orientation to interpretation of research holds true for other types of comparisons as well, including gender differences (Hegarty & Buechel, 2006).
Another complicating factor here is that the IC dimension, as dominant as it is in cultural research, may not be the only construct with explanatory power. For example, as Matsumoto and Yoo (2006) pointed out, psychologists have listed 25 additional dimensions that distinguish cultures. As they noted, countries that differ on one dimension may also differ on others. For instance, on the IC continuum, the United States stands at first in individualism, whereas Japan is 27th out of 70 countries (which puts both Japan and the United States below the median in collectivist tendencies). At the same time, Japan is 8th in uncertainty avoidance, whereas the United States is 59th (Hofstede, 2001). Matsumoto and Yoo also noted that Japan ranks 4th in long-versus short-term orientation, with the United States at 26th out of 36 countries, which might explain differences in anxiety level differences among Japanese and Americans (Iwata & Higuchi, 2000) as well as or better than IC.
Finally, researchers must provide empirical assessment of characteristics of people in a particular culture in order to draw inferences about behaviors based on perceived differences across countries. Noting the great within-group variability in traits within a culture and, often, small across-group differences, McCrae and Terraccianno (2006) demonstrated that “there does not appear to be even a kernel of truth in the stereotypes of national character” (p. 160). There may be consensus about people in a given culture (e.g., the belief by Americans that the English have no sense of humor), but McCrae and Terraccianno note that consensus (i.e., reliability) does not equal validity. Hence, valid measurements of traits in a sample must accompany inferences of behaviors based on those traits.