Abstract
The unit of observation in all current cross-cultural research is a culture or society. There are two consequences. First, if the unit of analysis is also a culture or society, demonstrable face or content validity must establish that the units of observation in a sample of cultures or societies are a random sample from the population from which the sample is deemed to be drawn. Second, where units of analysis (e.g., individuals) are aggregated into units of observation (i.e., cultures or societies), appropriate data analysis must be used. Unless these conditions are met, two varieties of error may result: The sample used may not be a valid, representative subset of the population, or analysis of aggregated data as if they were nonaggregated may distort results. Examples are given using the Standard Cross-Cultural Sample (SCCS). Validity appears to be greater for the SCCS as a sample of societies than a sample of cultures. A method for drawing new samples is suggested.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have