Replicating and Extending Cross-National Value Studies : Rewards and Pitfalls – An Example from Middle East Studies

AIB Insights 5 my 1980 monoGrAph culture’s consequences demonstrated how differences between the cultures of nations could be meaningfully described as relative positions on a limited number of measurable “dimensions.” These dimensions were empirically based on anthropologically meaningful personal values dominant in the respective nations, measured through questionnaires. This “dimension paradigm” turned out to appeal strongly to practicing managers and management scholars. In the 1980s I used the same paradigm – with different dimensions – for describing differences between the cultures of organizations and corporations within countries (Hofstede, 1991). In the 1990s and 2000s, several other researchers undertook developing their own cross-national dimension frameworks, the best known being Shalom Schwartz (1994) from Israel and Robert J. House with the GLOBE team (2004) from the United States.

my 1980 monoGrAph culture's consequences demonstrated how differences between the cultures of nations could be meaningfully described as relative positions on a limited number of measurable "dimensions." These dimensions were empirically based on anthropologically meaningful personal values dominant in the respective nations, measured through questionnaires.This "dimension paradigm" turned out to appeal strongly to practicing managers and management scholars.In the 1980s I used the same paradigm -with different dimensions -for describing differences between the cultures of organizations and corporations within countries (Hofstede, 1991).In the 1990s and 2000s, several other researchers undertook developing their own cross-national dimension frameworks, the best known being Shalom Schwartz (1994) from Israel and Robert J. House with the GLOBE team ( 2004) from the United States.
Even before the publication of my 1980 book, colleagues and students had asked for copies of the research instrument used.Hesitatingly I sent them a list, with the warning that the interpretation of the answers might depend upon the nature of their respondent population.My hesitation about uncontrolled use of the instrument has never disappeared.Comparing national cultures is basically complex anthropological research, not something for amateurs.I regularly receive and still answer questions which make me doubt the competence for this task of the correspondent.
Sometimes erroneous conclusions based on naïve uses of one of the versions of my instrument even pass the filter of peer-reviewed journals.This present note discusses an example, one of many, selected due to it being the most recent to come to my attention, not only of false conclusions drawn but also of valid conclusions missed.The case is a Brief Report by Oliver Fischer and Ahmad Al-Issa in the September 2012 issue of the respectable International Journal of Intercultural Relations.A recent overview of my work in AIB Insights (Littrell, 2012) suggested this to be an effective medium to reach potential users of my survey instrument.

The Essence of Cross-National Survey Studies: The Need for Matched Samples
In his recent guide to the methodology of cross-cultural analysis, Minkov (2013) lists and describes 26 large-scale studies of cultural dimensions across modern nations that appeared between 1980 and 2011.What all have in common is collecting their data from matched national samples of individuals."Matched" means similar in all measurable respects except their nationality.Thus, my 1980 monograph used employees in seven occupational categories within 40 national subsidiaries of the same global company to show nationality-based differences in workrelated values, while keeping corporate culture constant.Schwartz (1994) surveyed basic values of school teachers and university students from 38 nations.Project GLOBE (House et al., 2004) surveyed middle managers in local organizations in food processing, financial services, and telecom services from 62 societies, in order to understand their views on leadership.The World Values Survey (ongoing) uses public opinion poll methods to compare samples from the adult population in more than 100 societies.Valid cross-cultural studies compare apples with apples across countries; basing a country comparison upon apples in one country versus oranges in another (or even in the same) country or countries produces fruit salad.

On Using the VSM
The term Values Survey Module (VSM) first appeared in Hofstede (1980)."Module" stands for a set of questions that can be inserted into crosscultural surveys as a way of replicating and extending the country comparisons in my book.On the basis of an analysis of the first results, an improved version was issued in 1982, the VSM82.The most extensive application of the VSM82 is found in a PhD study by Hoppe (1990), comparing values of elites from 19 countries, who had attended courses in the Salzburg Seminar of American Studies.Hoppe's experience together with some of the other replications allowed issuing a third version of the questionnaire, the VSM94.In the meantime, the research of Michael Harris Bond from Hong Kong with the Chinese Value Survey (The Chinese Culture Connection, 1987) had led to the identification of a fifth dimension: Long-term vs. Short-term Orientation (Hofstede & Bond, 1988).In the VSM94, this dimension was added to the other four.
The VSM94 questionnaire and a corresponding 10-page Manual for users until mid-2013 could be found on my home website www.geerthofstede.eu(not to be confused with websites of licensees and pirates that use my name).In 2008, a fourth version of the questionnaire was issued, the VSM08, which is shown, with a new Manual, on the same website.Besides the English originals, the website contained translations of both Replicating and Extending Cross-National Value Studies: Rewards and Pitfalls -An Example from Middle East Studies Geert Hofstede, Emeritus, University of Maastricht, Netherlands versions into a number of languages, made available by users.The website stresses that the responsibility for the accuracy of these translations lies with the users.
In March 2006, Fischer and Al-Issa contributed an Arabic translation of the VSM94, which was added to the ones on our website.With this translation they collected data from 329 male and female students attending the University of Sharjah, United Arab Emirates (UAE), from seven Arab societies: Egypt, Jordan, Lebanon, Palestine, Saudi Arabia, Syria, and UAE.They used the formulas in the VSM94 Manual to compute scores for the five dimensions per country and across their total respondent population.When Fischer emailed me in June 2006 that they had successfully used their translated questionnaire, I answered, "I hope you have matched groups in other countries to compare your Arab students with?See the Manual!Yours, GH. " The Manual stressed the need for basing comparisons on matched samples only, comparing apples with apples, not with oranges.Therefore studies using the VSM on new populations cannot be expected to replicate the country scores in my publications.They usually do replicate the differences between country scores.
The results of Fischer and Al-Issa's 2006 study were nonetheless published six years later in their brief report in the International Journal of Intercultural Relations, September 2012.The report does not include any matched groups of students from outside the Arab region.Instead, the authors compare the dimension scores for their Sharjah University population with the scores for the Arabic-speaking region of my global company (IBM), collected around 1970.These scores (first four dimensions only) were based on the answers of IBM employees in Egypt, Iraq, Kuwait, Lebanon, Libya, Saudi Arabia, and United Arab Emirates: a total of 79 respondents in 1967-1969, and 62 more in 1971-1973(Hofstede, 2001: 482): 482).Contrary to what Fischer and Al-Issa wrote in their article, these surveys used Arabic translations of the IBM questionnaire, not the English original (Hofstede, 2001: 44).Because of too small samples per country, the Arab country data were not used in the 1980 edition of my book Culture's Consequences.In later publications, the data were bundled into an Arabic-speaking region.In terms of occupational categories, they were matched with the other countries and regions in the corporation.Among 53 countries and regions in IBM, the Arab region scored high on Power Distance (rank 7 out of 53), average on Uncertainty Avoidance (rank 27) and Individualism (rank 26-27), and just above average on Masculinity (rank 23; Hofstede, 2001: 500).
Fischer and Al-Issa called their article "In for a surprise, " because their average scores for the Arab-speaking region did not equal mine.As they were comparing apples with oranges, their surprise would only have been justified if the scores had been similar.

Analyzing the Fischer and Al-Issa Database
Fischer and Al-Issa's misdirected surprise does not mean that their research has been useless.The comparison among the seven Arab societies in their sample is more or less valid, as their student samples were matched and of sufficient size (between 24 and 89 respondents per country).The scores Fischer and Al-Issa published do show some significant differences among their seven countries on the four original Hofstede dimensions.Scores on the fifth dimension varied much less, and are left aside.The differences between their dimension scores for the seven countries are relevant in view of the different ways in which the present "Arab Spring" has affected these countries.

Power Distance
Power distance describes the extent to which the less powerful members of institutions and organizations within a country expect and accept that power is distributed unequally (Hofstede, Hofstede, & Minkov, 2010: 61).Students from Syria scored the largest and from Egypt the smallest on Power Distance.Large PD societies tend to be unable to resolve power struggles peacefully; the smaller the Power Distance, the greater the likelihood of a compromise.

Uncertainty Avoidance
Uncertainty avoidance is the extent to which the members of a culture feel threatened by ambiguous or unknown situations (Hofstede et al., 2010: 191).Students from Egypt scored the strongest and from Lebanon the weakest on Uncertainty Avoidance.Strong UA societies tend to be intolerant of people with different religions and convictions; in weak UA societies, different religions and convictions can exist side by side.

Individualism versus Collectivism
Individualism describes a society in which the ties between individuals are loose (everyone is expected to look after him or herself and his or her immediate family) versus Collectivism, which describes a society in which individuals from birth onwards are integrated into strong, cohesive in-groups, which throughout people's lifetime continue to protect them in exchange for unquestioning loyalty (Hofstede et al., 2010: 92).In-groups by their very nature oppose out-groups.Students from the United Arab Emirates and from Palestine scored relatively collectivist, their classmates from Egypt, Syria and Lebanon scored relatively individualist.Individualism tends to increase with modernity; in the seven emirates that make up the UAE the old tribal links make " As they were comparing apples with oranges, their surprise would only have been justified if the scores had been similar." the survival of collectivism understandable; in Palestine, the continuous external threat supports the importance of one's in-group as the only truly reliable source of protection.

Masculinity versus Femininity
Masculinity describes a society in which emotional gender roles are clearly distinct -men are supposed to be assertive, tough and focused on material success, women are supposed to be more modest, tender and concerned with the quality of life -versus Femininity, a society in which emotional gender roles overlap -both men and women are supposed to be modest, tender, and concerned with the quality of life (Hofstede et al., 2010: 140).Students from Egypt and Syria tended to score more masculine; those from Lebanon more feminine.Fischer and Al-Issa do not specify the gender composition of their national samples.The Masculinity/Femininity dimension is the only one of the four in which gender affects the scores: women on average score higher in Femininity than men, which could influence the scores, but as they stand they explain the antagonisms in Egypt and Syria, and the traditional compromises in Lebanon.
This brief analysis shows potential implications of Fischer and Al-Issa's findings for their seven Arab nations.Insiders more familiar with the cultures of these nations may find more or different clues.The Sharjah student database was a treasure which the authors in their 2012 article overlooked.They went in for the wrong surprise.

Conclusion
The message many PhD research program supervisors send to their candidates is that if you don't collect your own data, you're not really engaged in research.However, in the modern world collecting data includes more than hunting and gathering among respondents.Social science researchers certainly should always familiarize themselves with the realities of their respondents' situation.But for the quantitative part of cross-cultural studies, students will only rarely have the resources and access for collecting suitable samples from matched populations of respondents across nations.In this case, hunting and gathering in the literature and in the increasing number of professional data bases from the Web deserves to be accepted as valid research.