Hofstede Revisited: Is Making the Ecological Fallacy when Using Hofstede’s Instrument on Individual Behavior Really Unavoidable?

摘要

This paper suggests that avoiding the ecological fallacy is - under certain circumstances - possible. One problem in (cross-cultural) research is that there are often two levels of theorizing (individual and country) that needs to be taken into account when data are being analyzed and conclusions are drawn. Typically, as is the case with Hofstede's (1980) well-known research, cultural values are measured on country level. Consequently, researchers who make causal inferences from such group data to individual behaviors are making the ecological fallacy, i.e they (most often) wrongly assume that relationships observed for groups necessarily hold for individuals. If, however, a method could be found that could help determining how individuals behave from the study of aggregated data, the usefulness of Hofstede's research (and research on aggregated data in general) from a manager's point of view would definitely increase. In this paper the problem of the ecological fallacy, as well as different methods of avoiding it is discussed, and a particular solution of the problem is suggested.