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.
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.
One of the downloadable files contains the official matrix of dimension scores as far as we know and trust them. You can download the file freely for research purposes (.xls, .txt, .sav). Every now and then, new studies reveal new data. Some of these are plausible but not well enough documented. For instance, K.B. Mediwake, in an unpublished study, compared samples of Chinese and Sri Lankans. Equating his scores for the Chinese with those in the official matrix, and accounting for the difference with his Sri Lankans, he came to the following scores for Sri Lanka: PDI 72, IDV 28, MAS 35, UAI 53, LTO 49, IvR 64. While plausible, such scores derived from a small-sampled two-country comparison should be treated with caution, and we do not put them in the matrix.
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.
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 thank our contributors and will be happy to publish additional translations.
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.
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.