This semester, for the second year running, I tried out a small experiment with one of my classes to see if the students cultural profile matched Hofstede’s results for New Zealand:
I asked the students to fill in the VSM94 questionnaire, then put their results (anonymously) into a shared Google doc that calculated the results according to the algorithms outlined in the VSM94 manual. The sample size was only 30, so take the results with a pinch of salt, but the outcome was very interesting, and replicated the results from last year. Despite the fact that the majority of my students were not born in New Zealand, the results of our survey correlated quite closely with Hofstede’s figures, apart from masculinity:
|Cultural Measure||Hofstede Result||Student Result|
Certainly wider New Zealand society has masculine traits with a strong emphasis on competing in sport, aggressive outdoor activities and terrible driving, but perhaps students have a different subculture? However it is the similarities that interest me more than the differences, suggesting that we absorb the culture around us quite quickly, given the short time that some of my students have been in New Zealand.
2 thoughts on “Validating Hofstede? Some reflections on cultural differences”
Last year, I wrote about a study done by Kruger and Roodt validating the VSM94 in a South African context. My recommendations were:
As acknowledged in the current study done by Kruger and Roodt, the study itself possesses limitations. For one, the study was conducted in a very specific cultural and business context, which is female management in South Africa. In fact, the original objective of the study was to research the influence of cultural values on female leadership behavior. Only after discrepancies were observed in the results was the study reoriented toward a review of the VSM-94 as an instrument to accurately measure cultural values according to five particular dimensions. Furthermore, although Hofstede’s work on the VSM-94 was dependent on data limited to a single multinational corporation, the same constraint applies to the current study. The authors of the current study also used an abbreviated version of the VSM-94 to produce their results. To adequately determine the reliability and validity of the VSM-94 as a measurement for research in cultural values, then a more thorough and comprehensive undertaking needs to be in place. A range of cultural and business contexts, not only limited to female managers in South Africa, has to be used to strengthen the argument against the use of the VSM-94.
While Kruger and Roodt raise pertinent questions about the effectiveness of the VSM-94 in the cultural and business climate of today, their findings should be framed in such a way that takes into account the limitations of their research, especially since the VSM-94 has been widely used for so long in a number of subsequent and important studies. Great care should be taken to check the problems of the VSM-94 against a large sample size that considers many and varying cultural and business contexts and uses the complete form of the VSM-94. The only clear conclusion that may be derived from Kruger and Roodt’s research is that the VSM-94 is inadequate in measuring the influence of South African cultural values on female leadership in a particular corporation. In order to satisfactorily answer the question of whether the VSM-94 is reliable and valid enough, further research in the field has to be done.
Thanks, Matthew, for your comments. My own post was of course not intended as a research contribution, rather it was just a way of trying to find a more engaging way of teaching the topic that would involve the class in a practical activity. I had no particular expectations of the exercise but was pleasantly surprised by the alignment of our results with Hofstede’s data, not just in this year’s class but last year’s as well. I think the topic is valuable in a world where my computing students will likely work in organisations that are multinational / outsourced to / outsourced from / overseas.