Quantitative Research Methods

Measurement

In cross-cultural studies, culture is usually one of the primary constructs, often defined by national origin or ethnicity and usually treated as a dichotomous (or trichotomous) independent variable that is expected to influence the rest of the constructs measured. This approach raises several considerations about measurement, especially about the primary influencing variables in cross-cultural quantitative research.

The first consideration is the use of dichotomous (polarized) versus continuous variables. National culture and ethnicity are categorical variables: Someone is considered Mexican or Japanese, Hispanic or East Asian. The variable of culture, therefore, is not continuous (although there are some scales out there that measure things such as Hispanicity or the level at which a person is located on a range of scores). Subjects are assigned to a category based on nationality or [Page 703]knowledge of one’s ethnic heritage. This approach can be used to describe cultures by comparing two or more groups on a given set of attitudes or behaviors. Or it can be used to test theory, for instance, to select two cultures that are expected to differ widely on one or more constructs that are theorized to influence a given set of attitudes or behaviors: If the theorized differences are supported, then the theory is supported. Lacking a strong theoretical rationale, many cross-cultural quantitative studies result in a comparison of two or more groups on a set of constructs, as in the first approach above.

A causal theory explains the relationships between constructs. Culture or ethnicity alone cannot explain causation. A culture must be associated strongly with some other construct for the theory to be tested. Some examples could be values (e.g., power distance or honor), socioeconomic differences (e.g., per capita income), or relational structures or attributes (e.g., network transitivity). These constructs should be measured or quantified to clearly associate the group with the level of the relevant associated constructs.

From the view of theory construction, continuous variables provide more useful information for prediction and explanation than categorical variables can provide. A conceptual rationale that explains why a particular culture or group should be high or low on a continuous construct—accompanied with measurement (as with values) or verification (as with economic data)—can advance theory.

Some continuous variables, such as individualism– collectivism, are treated as dichotomous variables and therefore typically provide weaker causal explanations than other continuous variables. If differences are found between cultures, they are often assumed to be due to individualism–collectivism, which treats this construct as an all-encompassing theory rather than a theory of the middle range. A strong rationale should ask what the characteristics of a culture are that suggest differences in individualism–collectivism and develop measures, especially continuous ones, for those characteristics instead.

...