Research and VSM

Geert Hofstede first published his scholarly book Culture's Consequences in 1980. In hindsight, it represented a paradigm shift in the comparative study of cultures. Since then, the amount of use made of his work, as well as the amount of research carried out in the area of culture, has continually been on the increase. Today, several dimensional frameworks of culture exist as well as various other conceptions of what culture is. Judging from the frequency of application, as well as from the continued validity of the Hofstede dimensions against other kinds of data, the Hofstede model is still the most valid, practical one to date. But new developments are happening.

Dimensions do not exist - but they can serve
Geert stresses that dimensions of cultures do not exist in a tangible sense. They are constructs. A construct is "not directly accessible to observation but inferable from verbal statements and other behaviours and useful in predicting still other observable and measurable verbal and nonverbal behaviour" (Teresa Levitin, 1973). Culture itself is a construct, so are values. It makes no sense asking how many dimensions of culture there are.  This is like asking how many types of cloud exist - it is a matter of definition, and practical significance should be the criterion.

Replications
Many researchers have followed Hofstede's paradigm. Some have developed new research instruments for comparing country samples; the best known are Shalom Schwartz and Robert J. House with the GLOBE team. Note that GLOBE use some of the same dimension names as Hofstede, but for different constructs! Others have replicated  Geert's research, using updated versions of the IBM questionnaire on which his original study was based. These are  known as Values Survey Modules (VSM) ;  there has been a VSM82 and a VSM94; the most recent version is the VSM08.

Using the VSM: warnings
The VSM calls for professional research expertise. The VSM has been useful in estimating country scores for countries not part of the original research, by comparing carefully matched respondent samples between the new countries and one or more countries covered before. The usefulness of replications increases with the number of countries  included. One-country replications are meaningless because they have no match to compare with: the VSM is is like a thermometer that has to be re-calibrated at each use. Those considering two-country replications on countries studied before should rather search the literature. Anybody wanting to use the VSM should study the 2001 edition of Geert's scholarly book Culture's Consequences first.

Translations of the VSM
Researchers have translated VSM versions into various languages, and copied us on their translations. All available translations of the VSM08 and the VSM94 are accessible via this website. We neither doubt nor guarantee their quality; users should always check their tools. We will be happy to publish additional translations.

Levels of analysis
The Hofstede dimensions of culture are group-level constructs. Dimensions of national culture are about societies; dimensions of organizational culture about organizational units. Neither is about individual differences between members of society or organizations. Comparing survey responses between individuals does not yield similar patterns to the cross-population comparisons on which the Hofstede dimensions are based. This kind of issue is known among scholars as the distinction between levels of analysis. In the attached article on "multilevel research", Geert refers to distinguishing among flowers, bouquets and gardens.

Dimensions cannot tell what will happen
The Hofstede dimensions do not directly predict any phenomena or dynamics. Applying them to make sense of what happens in the world always has to take into account other factors as well as culture - notably national wealth, history, personalities, and coincidences. There is no quick fix to understand social life after taking a dose of Hofstede. But the dimensions, when well understood, do allow to predict a little better what is likely to happen. And they become more useful as you go from the specific case to the trend, average, or expectation. For instance, knowing about culture hardly helps to predict what car you will buy - but the trend among your compatriots to buy certain types of car in the coming years can be predicted fairly well.