A Re-Inquiry of Hofstede's Cultural Orr and Häuser

A RE-INQUIRY OF HOFSTEDE'S CULTURAL DIMENSIONS:

A CALL FOR 21st CENTURY
CROSS-CULTURAL RESEARCH
LINDA M. ORR, The University of Akron
WILLIAM J. HÄUSER, The University of Akron

Given the impact of Hofstede 's Cultural Dimensions over the past quarter of a century, many
scholars and practitioners have utilized these dimensions. However, numerous researchers have
questioned his methodology, while others misused the dimensions in terms of the original purpose.
Yet surprisingly, very few studies have performed an exact replication. This study summarizes
Hofstede's work and critiques his cross-cultural model. In order to test Hofstede's constructs on
different populations, three quantitative analyses were performed using domestic U.S., Asian, and
Australian samples. This study found serious problems with Hofstede's factor structure. Additionally,
the study suggests the need for re-examining the cultural dimensions within the global information
based context of the early 21^' century. This is not meant to criticize Hofstede, but instead to pinpoint
fallacies to enable researchers to build from his work in more appropriate directions.

INTRODUCTION applicability of its implications. Hofstede's

work has inspired a great improvement in the
In 1980, Geert Hofstede published Culture's discipline by specifying a theoretical model
Consequences. This influential study soon which serves to coordinate research efforts
became a major source of reference about value (Redding 1995). This theoretical model has
differences around the world. Culture's served as the foundation for many other
Consequences has been translated into research efforts. In sum, Hofstede's initial four
numerous languages since its original (later five) fundamental dimensions of culture
publication and was fully revised in 2001. still serve today as basic, fundamental criteria
Additionally, Culture's Consequences has been in most interdisciplinary, cross-culturally
cited more than any other book in social comparative research.
sciences (Yoo and Dunthu 2002). Hofstede has
been cited over 5,000 times, with more than However, Hofstode's work has been
3,000 of these citations being to Culture's misconstrued and misinterpreted in many
Consequences (IACCP 2007). Hofstede's work subsequent studies. More importantly,
has made it beyond higher-level research and surprisingly few exact replications, attempting
has worked its way into everyday teachings. to empirically examine Hofstede's factors, have
Many of his citations are in basic principles of been conducted. Furthermore, many subsequent
marketing, international business, advertising, studies have taken and utilized Hofstede's work
and consumer behavior textbooks. with surprisingly little questioning of his
results. Thus, due to the tremendous impact that
Hofstede's work has inspired a multitude of Hofstede has had on the scientific community,
international marketing research and has been an exact replication study is a necessary step.
the dominant research paradigm in cross- While replication studies may not hold the
cultural studies of national attitudes for some highest regard within academic research
time because of the simplicity of its theory and because they do not bring "anything new to the
table," replication studies for works this
The Marketing Management Journal
important must be done. If not, as researchers,
Volume 18, Issue 2, Pages 1-19 we will continue to utilize faulty theories and
Copyright O 2008, The Maiketing Management Assoeiation
All rights ofrcproducdonin any form reserved
models. Thus, a re])lication study of Hofstede's

Marketing Management Journal, Fall 2008

A Re-Inquiry of Hofstede's Cultural Orr and Häuser

work is necessary in order to assess the millennium. Thus, given the widespread
appropriate content, as well as the reliability acceptance of Hofstede's instrument, it was
and validity of Hofstede's cultural dimensions. used for this research.

Thus, the purpose of this study is to perform a Hofstede's work is based on "mental
very close replication of Hofstede's original programs." Due to the process of socialization,
study. To do this, first, a review of Hofstede's these mental programs are developed in the
methodology is presented. This review was family in early childhood and reinforced in
conducted because many researchers may not schools and organizations, and other areas
be fully aware of the almost haphazard manner throughout our lives, experiences, and
in which the cultural dimensions were initially upbringings. Thus, due to the shared common
developed. The results of three studies designed experiences of people living in the same
to analyze face validity are preformed. country, these mental programs contain a
Following these analyses, three re-inquiry component of national culture. They are most
studies are presented. Hofstede's instrument clearly expressed in the different values that
was obtained and reproduced with no predominate among people from different
alterations and administered with multiple countries (Hofstede 1980).
international samples. The examination of the
data in this study was conducted with more From 1967 to 1972, Hofstede administered
statistical rigor than any other replication study 117,000 questionnaires to employees of IBM in
known to the authors. As we have progressed over 60 different countries (Hofstede 1980).
well into the new millennium, cross-cultural His study resulted from the collaboration of
research is no doubt one of the most important researchers from five countries and his survey
subsections of marketing and international was pre-tested in ten countries (De Cieri and
business research. It is imperative that we have Dowling 1995). By 1980, he had developed his
a solid foundation and understanding from own cultural dimensions, Individualism-
which to build future research. Collectivism, Power Distance, Uncertainty
Avoidance, and Masculinity-Femininity.
BACKGROUND
Power distance is defined as the degree that
As mentioned, Hofstde's famous study is unequal distributions of power are expected and
widely recognized as a major break-through in accepted. Power distance "represents a nation's
cross-cultural social science studies. There are unique score on how to deal with social
almost no publications, either from the inequality. Inequality can occur in areas such as
disciplines of sociology, anthropology, history, prestige, wealth, and power; different societies
law, economics or business administration, that put different weights on status consistency
do not refer to Hofstede's work and his five among these areas" (Hofstede 1984, p.
fundamental dimensions of culture when 65). Uncertainty avoidance is the extent to
explaining correspondences and distinctions which people feel threatened by ambiguous
between cultures (IRIC online 2002). Geert situations and have created beliefs and
Hofstede is among the 20 most cited Europeans institutions that try to avoid these (Hofstede and
in the 2000 Social Science Citation Index Bond 1984, p. 419-420). Individualism-
(Institute for Research on Intercultural Collectivism "describes the relationship
Cooperation 2001), 57* in the world, with 416 between the individual and the collectivity
articles referring to him. In fact, Hofstede's which prevails in a given society," where
influence is becoming even more pronounced, "individualism is defined as a situation in
with the number of citations increasing, not which people are supposed to look after
decreasing, each subsequent year. Simply put, themselves" and "collectivism is defined as a
Hofstede's dimensions are still utilized widely situation in which people belong to in-groups or
even as we have progressed well into the new collectivities which are supposed to look after
Marketing Management Journal, Fall 2008
A Re-Inquiry of Hofstede's Cultural Orr and Häuser

them in exchange for loyalty" (Hofstede 1984, order to permit longitudinal and cross-national
p. 148 and Hofstede and Bond, 1984, p. 419- investigation. The first instrument consisted of
420). Finally, masculinity-femininity "describes 180 items, which were chosen through existing
the division of social roles between women and surveys, pilot studies, and literature review
men in a society." The predominant (e.g., Baehr 1954; Herzberg et al. 1957;
socialization pattern is for men to be more Hinrichs 1968; Vroom 1964; Wherry 1954).
assertive and for women to be more nurturing. After the initial sui^vey, individual site survey
"Masculinity is defined as situation in which administrators were still customizing the
the dominant values in society are success, surveys to their spixific needs. Thus, most of
money, and things" and "femininity is defined the surveys at this point varied considerably
as a situation in which the dominant values in from site to site. Therefore, a 1970 task force of
society are caring for others and the quality of researchers, including Hofstede, took over with
life" (Hofstede 1984, p. 176; Hofstede and a new approach. They wanted to derive an
Bond 1984, p. 419-420). instrument that usetl the previous questions but
had no more than 60 items. The criteria for
A group of researchers calling themselves the these questions were:
Chinese Culture Connection (1987) conducted The core questions should cover all the major
further analysis of the Hofstede dimensions in area or dimensions of job attitudes (content
Asian cultures and added a fifth dimension, validity);
Confucian Dynamism. In the 2001 edition of • The areas of job attitudes covered should be
Culture's Consequences, Hofstede has included meaningful in terms of theories of human
a chapter on this dimension and called it long- motivation and organization;
term versus short-term orientation. Confucian • The questions should be reasonably
Dynamism conceptually incorporates many reliable,
diverse elements of Confucian cultures. • Core questionnaire items should universally
Empirically, however, Confucian Dynamism apply to all employees of the corporation;
has consisted of two negatively correlated sets • Questions should be translatable;
of items, described by the Chinese Culture • The questions should be chosen from those
Connection as a positive and a negative pole. used before, to permit longitudinal studies;
More specifically, "there were four positively • Questions should be acceptable to the
loaded values in this grouping, all reflecting corporation's managers;
Confucian work ethics." These four items were, • All core questions should be useful
ordering relationships, thrift, persistence, and a information to managers; and
sense of shame - all represented by single • The number should not exceed 60 items
items. "Counterpointed against this hierarchical (Hofstede 1975).
dynamism were four negatively loaded values
advocating checks and distraction at the In 1971, Hofstede and colleagues reduced the
personal, interpersonal, and social levels" number of items from 180 to 120. His decision
(Chinese Cultural Connection 1987, p. 150). was to eliminate il ems that did not frequently
These items were, reciprocation, personal appear in the litenture. Hofstede next ran an
steadiness, protecting your face, and respect for exploratory factor analysis with a sample of
tradition. Thus, Confucian Dynamism begins to 700 employees on ull 120 items. After varimax
address traditional eastern values. rotation, he was left with 15 factors. The first
three factors explained 77 percent of the
HOFSTEDE'S METHODOLOGY variance. Therefore, he kept the items for those
three factors and eliminated the other items. He
IBM had occasionally surveyed employees to was left with three dimensions: management,
judge attitudes toward job satisfaction prior to satisfaction, and culture. Hofstede and
1960. In 1967, a team of researchers was colleagues (1971) had 146 items at this point,
gathered together to standardize the surveys in including demogra])hic variables. Next, he took

Marketing Management Journal, Fall 2008

A Re-Inquiry of Hofstede's Cultural Orr and Häuser

his new survey with 146 items and quantitative evidence to refer to, Hofstede
administered it to 5 separate populations, which conceptually decided which components should
follow. be a part of each of the three factors.
• Technical experts, France, 1968, n = 436
• Technical experts, U.K., 1968, n = 436 From the previous analyses, Hofstede devised
• Head office clerks, secretaries and other his new instrument. The final questionnaire
nonprofessional employees, U.K., 1969, n encompassed 60 core questions: 58 from factor
= 535 analysis and two new items. Subsequently,
• Unskilled direct manufacturing operators, Hofstede administered the same work
Japan, 1970, n = 231 satisfaction survey in other countries and
• Unskilled direct manufacturing operators, derived his "cultural" dimensions. The
U.K., 1970, n = 296 following paragraphs will describe how each of
the Hofstede cultural dimensions, as we know
To analyze the data, Hofstede preformed them today, are derived through the use of
separate factor analyses with each population. theoretical reasoning and factor analysis.
Management questions explained 21-27 percent
(depending on the sample) of the variance and Hofstede began administering the instrument in
the weakest factor, culture, explained 11-19 individual countries at this stage. Each factor
percent of the variance (see Tables 1 and 2). analysis was preformed separately for each
However, the three components that Hofstede country and then standardized, normalized
derived, management, satisfaction, and culture, means were calculated to derive a factor score.
were not one-dimensional. In fact, the "culture" Hofstede himself admits that factor structure
components contained thirteen constructs, some does not hold across populations (Hofstede
of which contained zero or only one item. In 1984, p. 43). In fact, as mentioned, he never
other words, even though there was no data or intended his instrument to be used at the

TABLE 1
Hofstede et al. (1971) Factor Analysis
Number of Number of factors % of total vari- % of variance
variables with eigen- ance explained explained by
values>l first factor
Satisfaction questions:
1. T. E. France 54 15 62 17
2. T. E. U.K. 54 15 62 16
3. Clerks U.K. 52 15 63 20
4. Operators Japan 51 15 68 24
5. Operators U.K. 54 15 65 20
Management questions:
1. T. E. France 50 14 62 22
2. T. E. U.K. 50 14 61 21
3. Clerks U.K. 50 13 63 24
4. Operators Japan 48 15 72 27
5. Operators U.K. 50 14 65 24
Culture questions:
1. T. E. France 42 12 58 13
2. T. E. U.K. 42 14 61 11
3. Clerks U.K. 40 11 59 17
4. Operators Japan 40 13 63 16
5. Curators U.K. 42 11 59 19

Marketing Management Journal, Fall 2008

A Re-Inquiry of Hofstede's Cultural Orr and Häuser

TABLE 2
Hofstede's Factors
Satisfaction Number of fVfanagement Number of Culture Number of
Factor Items Factor Items Factor Items
SI 3 Ml 5 Cl 2

S2 1 M2 2 C2 2

S3 3 M3 2 C3 1

S3A 3 M4 2 C4 2

S4 3 M5 0 C4A 1

S4A 1 M6 0 C5 0

S5 1 M7 1 C6 2

S6 1 M8 1 C6A 1

S6A 1 M9 1 C7 3

S7 2 C8 1

sg 1 C9 0

S9 1 CIO 0
SIO 1 cn 0

C12 0

C1.3 1

individual level. Hofstede (1984, pp. 43 and 55) "culture" factor as he saw themes emerge. He
admits, derived these dimensions theoretically instead
"A between-cultures analysis had not of empirically. In other words, he examined the
been done at that time; first, because the "culture" factor and made educated guesses at
main purpose of the survey operation which items should make up each of his four
was organization development - that is, cultural dimensions.
use within parts of the organization -
within made the within-analysis The first dimension was power dislance.
obvious, and second, I must confess that Hofstede noticed that the question, "How
the difference between within- and frequently are employees afraid to express
between-culture analysis had not disagreement with their managers?" was
occurred to us at that time. If it had, we receiving similar answers within cultures (but
might have come to a very different not between). He then decided to choose this
selection of, in particular, the 'culture' one core question iïs his entire power distance
survey items.. .From the earliest! surveys dimension (Hofstede 1984). Two additional
onward, it had been clear that questions questions were added based on ecological
dealing with hierarchical relationships correlations and this formed the Power Distance
received systematically different dimension.
answers in different countries."
Next, according to Hofstede (1984, p. 55), "the
From this point forward, while it is somewhat uncertainty avoidance index was developed in
unclear in past writings, Hofstede derived his an analogous way. I had an earlier theoretical
four separate cultural dimensions from the interest in the phenomenon of stress which was

Marketing Management Journal, Fall 2008

A Re-Inquiry of Hofstede's Cultural Orr and Häuser

measured by the question 'How often do you analyses, numerous other limitations exist with
feel nervous or tense at work?'" Scores on this Hofstede's research. While the overall findings
question differed greater by country than by of Hofstede's research are extremely relevant to
occupation. Thus, Hofstede was able to deduce today's cross-cultural studies, and the rigor is
that a cultural dimension existed around possibly unmatched even today, major
people's differing reactions and exceptions to constraints exist with Hofstede's research. First,
uncertainty and anxiety. and through no fault of Hofstede, there is a
question of time relevancy. Researchers have
A potentially rich source of data was also questioned whether the dimensions developed
available in the "work goal importance" from data collected between 1966 and 1973
questions (Hofstede 1984). After normalizing were artifacts of the period of analysis (e.g.,
the data on these 14 questions, Hofstede Baumgartel and Hill 1982; Warner 1981).
realized that a structure emerged similar to Hofstede investigated the correlations between
Maslow's hierarchy of needs (Hofstede 1984). his data and other variables like geographic,
Through a review of the literature and a long economic, demographic, and political national
process of analysis and deduction, Hofstede indicators. Over forty years have passed since
later decided that these 14 questions measured the beginning of the study. Just a simple map of
two constructs - Masculinity and Individualism. the world looks very different today than it did
Subsequently, Hofstede subjected the above in 1966. While these correlations were
mentioned questions for all four factors, plus beneficial, they are not only out dated, but the
others that had shown a reasonable amount of cultures themselves have changed.
stability over time to a factor analysis with
orthogonal rotation. The final 32 items Second, Hofstede's research may suffer from
explained 49 percent of the variance and sampling problems. Several researchers have
became his initial set of questions to measure argued that the constraints derived from
his cultural dimensions. Hofstede's research population of IBM
employees (e.g.. Graves 1986; Merker 1982;
Even though the masculinity and individualism Triandis 1982). The use of employees from one
constructs were derived from the same company allowed Hofstede to reduce the other
variables and even though the above factor sources of variance and concentrate on culture.
loadings would empirically be three constructs, However, several criticisms have come from
Hofstede took a different approach. For reasons this fact. First, IBM employed mostly males at
which are not explained anywhere in the the time of the survey. In the words of Milton
literature, according to Hofstede (1984, p. 62), Bennett (1996), "the differences between men
"Factor one represents an Individualism-low and women is the greatest culture conflict of
Power Distance factor...I shall continue to treat all." More differences exist between men and
them as two dimensions because they are women than from country to country, especially
conceptually distinct...Factor 2 is a masculinity when analyzing things like
factor... and factor three corresponds to masculinity/femininity, power distance, and
uncertainty avoidance" (Hofstede 1984, p. 62). individualism/collectivism.
These items can be found in Table 3. As can be
seen, some items are included in overlapping In lieu of the sampling issue, all the subjects in
dimensions. The obvious confusion of this the survey were from the same corporate
methodology, along with its limitations, is culture. Additionally, although Hofstede
discussed in the following sections. surveyed many countries, all subjects were
employees of an American company.
LIMITATIONS Additionally, most employees were from white-
collar positions. Hofstede (1980) himself
Other than the obvious methodological discusses the problems of ethnocentrism that
problems which stem from the above described exist in previous scales. As Hamden-Tumer and
Marketing Management Journal, Fatl 2008
A Re-Inquiry of Hofstede's Cultural Orr and Häuser

TABLÉ 3
Original Hofstede Items and Factor Groupings

Construct Item Hofstede Item

Label in
Present
Stud

Individualism and ici Have good working conditions (good lighting, adequate work s]iace, an attractive office, etc.)?
Collectivism Items icio I would not support my work group if I felt they were wrong.
icll If tbe group is slowing me down, it is better to leave and work a lone.
icl2 It is better to work in a group tban alone.
icl3 Groups make better decisions tban individuals.
icl4 I prefer to be responsible for my own decisions.
ici 3 Contributing to the group is tbe most important aspect of work.
ic2 Have considerable fi^edom to adopt your own approacb to tbe job?
ic3 Have a job tbat leaves sufficient time for your personal or family life?
ic4 Fully use your skills and abilities on tbe job?
ic5 Have a job on wbicb tbere is a great deal of day-to-day learning?
ic6 Competition among employees usually does more harm tban good.
ic7 Decisions made by individuals are usually of bigber quality tbaii decisions made by groups.
ic8 It is important to stick witb my work group, even tbrougb difficulties.
ic9 My personal accomplishment is more important than group success.

Items were used to icandmfl Have challenging work to do; work from which you can get a pimsonal sense of accomplishment.
measure both
Individualism and icandmf2 Having interesting work to do is just as important as having bigh earnings.
Masculinity (Because
items bad factor icandmD Most employees want to make a real contribution to tbe success of tbeir company.
loadings above O.S on
eacb construct)

Masculinity and mfl Live in an area desirable to you and your family?
femininity items mflO Have tbe security that you will not be transferred to a less desirable job?
mfll Work in a congenial and friendly atmospbere?
mfl2 A corporation sbould bave a major responsibility for tbe healtb and welfare of its employees and tbeir
immediate families.
mn3 A corporation sbould do as much as it can to help solve society's problems (poverty, discrimination,
pollution, etc.).
mn4 Most companies bave a genuine interest in the welfare of their umployees.
mfl5 The private life of an employee is properly a matter of direct concern to his company.
mfló It is important for me to have a job tbat provides opportunity for advancement.
mn7 It is important tbat I outperform otbers in tbe company.
mfl 8 It is important for me to have a job tbat provides an opportunity' for high earnings.
mfl9 It is important for me to work in a prestigious or successful company.
mf2 Have an opportunity for high earnings?
mf3 Work with people who cooperate well with one another?
mf4 Have the security tbat you will be able to work for your company as long as you want to?
mf5 Have an opportunity for advancement to bigber-level jobs?
mf6 Have a good working relationship with your manager?
mf7 Ciet the personal recognition you deserve when you do a good job?
mfS Have a job tbat allows you to make a real contribution to tbe success of your company?
mf9 Work in a company tbat is regarded in your coimtry as successlul?

Power Distance items pdl Employees lose respect for a manager wbo asks tbem for tbeir iidvice before be makes a fmal
decision.
pdlO For getting abead in industry, knowing influential people is usually more important tban ability.
pdll Even if an employee feels that he deserves a salary increase, he should not ask his manager for it.
pdl2 My superiors should make most decisions without consulting n le.
pdl3 It is improper to disagree with one's supervisor.
pdl4 I would never argue with my supervisor.
pdl5 I believe that those superiors who ask opinions too often of subordinates are weak or incompetent.
pdl6 I believe tbat superiors are entitled to special privileges.
pdl7 77iis question asks the respondent to circle his preferred manager type among three choices, from the
most consultative to the least consultative.
pd2 Employees sbould participate in tbe decisions made by managnment.
pd3 Company rules sbould not be broken; even wben tbe employee thinks it is in the company's best
interests.

Marketing Management Journal, Fall 2008

A Re-Inquiry of Hofstede's Cultural.... Orr and Häuser

TABLE 3
{continued)

Construct Item Label Hofstede item

in present
study
Power pd4 Employees should never express disagreement with their managers.
Distance Items pd5 Employees should always be told very clearly their duties and responsibilities, and how to perform
(continued) their jobs.
pd6 Most employees have an inherent dislike of work and will avoid it if they can.
pd7 A good manager gives his employees detailed and complete instructions as to how they should do
their jobs; he does not merely give general directions and depend on them to work out the details.
pd8 In general, the better managers in a company are those who have been with the company the longest
time.
pd9 There are few qualities in a man more admirable than dedication and loyalty to his company.

Uncertainty ual A good manager does not get too involved in the details of an employee's job; rather, these details are
Avoidance items left to the employee.
ua2 Staying with one company for a ling time is usually the best way to get ahead in business.
ua3 A large corporation is generally a more desirable place to work than a small company.
ua4 Companies should not change their policies and practices very often.
ua5 It is important for me to work for a company that provides high employment stability.
ua6 Clear and detailed rules/regulations are needed so workers know what is expected of them.
ua7 It is better to work in a well-defined job where the requirements and procedures are clear.

Trompenaars (1997) noted, they doubt that Another criticism is that culture cannot be best
American IBM managers serving in foreign expressed in a mathematical language, as
countries are much different than American Hofstede does. This point can be best
IBM managers in America. summarized by a story given by Hamden-
Tumer and Trompenaars (1997). A leamed
Inclusive with the sampling issue is the matter researcher diced a piece of cheese with a
of the original sample size. While the sample kitchen gadget and then wrote a leamed
eventually grew to quite some size, the original dissertation on the cubic nature of cheese. We
constructs were derived from very few workers. get out of factor analysis what we put into it,
According to Hair and colleagues (1998), the nothing more, and nothing less. As the previous
minimum sample size is five observations per section described, Hofstede's determinations
variable to be analyzed. However, ten were haphazard at best.
observations per variable are better, and some
even recommend 20 per variable. Therefore, Another problem with Hofstede's work is that
Hofstede's results were sample specific and the study did not begin as a cultural study. It
they took advantage of random correlations. initially began as a work satisfaction study.
Some samples were as low as 231 for 146 Hofstede was a brilliant researcher who noticed
items. the dimensions as they developed. However,
the original survey was not designed for its
Hamden-Tumer and Trompenaars (1997) bring final purpose. The survey was refined and
up another important criticism of Hofstede's changed several times to make the necessary
work. "Are cultural categories linear and adjustments. Thus, the dimensions were derived
exclusive?" Hamden-Tumer and Trompenaars empirically, rather than theoretically.
(1997) do not feel that if you are an
individualist you cannot be a coUectivist. Finally, Hofstede (1980) specifically identified
Perhaps some people tend to be very the ecological fallacy that exists with his work.
individualistic at work, but family oriented and The ecological fallacy can be defined as
collectivistic at home. "confusion between within-system and
Marketing Management Journal, Fall 2008 8
A Re-Inquiry of Hofstede's Cultural.... Orr and Häuser

ecological correlations." Similarly, he readily the difficulties with Hofstede's research.

admits that within-culture variations can be as Trompenaars' book (1993) has become a harsh
great as if not greater than between-culture debate between Hofstede and himself (for full
variations. This is an important observation to details, see, Hamden-Tumer and Trompenaars
make. While it is helpful to understand that the 1997; Hofstede 1996; Hofstede 1997).
majority of Chinese citizens tend to be very
collective, marketers sell to the individual, THE METHODOLOGY
mangers recruit, train and hire the individual, OF THE, RE-INQUIRY
and psychologists, economists, and all social
scientists are concerned with the individual and Even with all the rejüication studies that exist in
groups existing in their society. the literature, very few have been exact
replications, using Hofstede's actual, original
The above mentioned limitations have been items. In fact, prior to 2001, Hofstede did not
examined and identified in numerous make these original items readily available.
replication studies over the last four decades. Thus, the purpose of this study was to produce
Sendergaard (1994) located 61 replications of a very similar replication study. A further aim
Hofstede's research. Full confinhation of of this study was to analyze the dimensions
Hofstede's dimensions was found in only four using multiple statistical techniques in order to
studies (Hoppe 1990; Shackelton and Ali examine the items and constructs as thoroughly
1990). Partial confirmation was found in as possible. In order to start this process,
another fifteen studies (e.g., Ashkanini 1984; Hofstede's original instrument was obtained
Chow et al. 1991; Forss 1989; Huo and Randall from Culture's Consequences. The survey was
1991; Lowe 1994; Maldonado 1983; Pooyan administered in two different forms. The first
1984; Westwood and Everett 1987; Yeh 1988). study was an attemjit to assess face validity. In
Lowe's (1994) study is particularly interesting the second study, convergent and discriminant
because he used IBM employees from Hong validity was assessed. Finally, the data is tested
Kong and the United Kingdom for his sample. through confirmatory factor analysis utilizing
Lowe was not able to find differences between structural equations tnodeling.
the two countries for Hofstede's uncertainty
avoidance dimension (S0ndergaard 1994). In order to ascertain the effects of the
instrument on a non homogeneous sample (e.g.,
After Sondergaard's (1994) study, other authors samples from more than one company) a
have critiqued and replicated Hofstede's work number of different samples were used. This
and applied his dimensions to various contexts was also done to increase the probability of
(e.g., Fernandez, Carlson, Stepina, and variance from other sources, such as gender,
Nicholson 1997; Kelleher 2000; Marshall 1997; while minimizing the confounding effects of
Naumov and Puffer 2000; Robertson and such factors as occupational status (e.g., white
Hofftnan 2000; Smith, Dugan, Peterson, and collar) within one over-arching corporate
Leung 1998; Sopachitwattana 2000; culture environment.
Trompenaars 1993; Trompenaars 1997; Van
Oudenhoven 2001; Van Oudenhoven, Study One
Mechelse, and de Dreu 1998; Verbeke 2000;
Yeh and Lawrence 1995). Most of these studies In study one, Hofstede's instrument, wilh no
have come to very similar conclusions as the modifications, was administered to two samples
ones prior to 1994. No known rigorous study (refer to Table 3). The first sample included
using Hofstede's exact instrument has found 123 undergraduate students and the second
complete confirmation to Hofstede's work. In sample included 65 graduate students, and the
fact, Trompenaars (1993) examined Hofstede's third sample coniained 13 marketing and
dimensions and arrived at his own dimensions, miinagement faculty. The undergraduate class
which, according to Trompenaars, overcome took the survey after a whole class period
Marketing Management Joumat, Fall 2008
A Re-Inquiry of Hofstede's Cultural Orr and Häuser

dedicated to teaching Hofstede's cultural students at a large mid-south university. Of the

dimensions. All samples were given the survey 161 respondents, 58 percent were America and
with thorough definitions of each of Hofstede's 41.6 percent were Far Eastern. Respondents
four cultural dimensions at the top of the were required to have full time work experience
instrument. The subjects were then instructed to in order to participate in the study. This criteria
ascertain which dimension each survey item was added to the survey so that all respondents
was attempting to measure by circling IC could identify with the work-related questions
(individualism/collectivism), MF that appear on Hofstede's instrument.
(masculinity/femininity), UA (uncertainty Respondents had work experience ranging from
avoidance), or PD (power distance) after each 1 to 40 years and were between 22 and 59 years
item. of age. A second sample of non-student adults
was taken in America (N = 233). Demographic
As can be seen in Table 4, the task proved to be characteristics of this sample were very similar
quite difficult. The graduate students performed to sample one, except 97.4 percent of the
better than the faculty, followed by the sample was American. Finally, a third sample
undergraduate students, ascertaining correctly was gathered from non-student adults in
64.62 percent, 44.25 percent, and 32.25 percent Australia (N = 210). Due to the widespread
of the time respectively. For a face validity cultural differences that exist in Australia, this
assessment, the percentages should be much sample was extremely diverse. The sample
higher, whether the sample is common workers contained people with varying nationalities
or especially trained researchers and academics representing a total of 21 countries.
such as those used in this study. The
individualism items were easiest to identify Data were analyzed in many different ways in
while the masculinity items were the hardest to order to demonstrate that the factor structure
classify. does not hold in any circumstance. Reliabilities
were assessed for each of the four factors with
Study Two the three combined samples and with the
samples separately. Likewise, factor analyzes
In study two, Hofstede's exact survey with no were run with each sample separately and
modifications was given to three different combined. Exploratory factor analyses were
samples. Two out of the three samples were in used to remain consistent to Hofstede's original
America. The first sample consisted of graduate methodology and many different techniques
were utilized. Varimax and oblique rotations

TABLE 4
Face Validity Assessments: Average Number of Times
Ascertained Correctly Across Dimension

Faculty Graduate Undergraduate Average of all

sample sample sample three samples

Masculinity items 28% 40.73% 15% 27.91%

Individualism items 55% 78.24% 46% 59.75%
Power Distance items 41% 60.94% 33% 44.98%
Uncertainty avoidance items 53% 78.57% 35% 55.52%
Average 44.25% 64.62% 32.25% 47.04%

Marketing Management Journal, Fall 2008 10

A Re-Inquiry of Hofstede's Cultural.... Orr and Häuser

were tried to remain true to Hofstede. Likewise, components, explainable variance dropped to
factors were extracted by examining around 30 percent.
eigenvalues greater than one and by "forcing"
the solution to only four factors. Additionally, These statistics seem discouraging, but it is
even though Hofstede did not use confirmatory even more discouraging to analyze what items
factor analysis, this study attempts to do so. All are loading with which construct. In all
analyses will be discussed subsequently. analyses, items had no pattern as to which
construct they loaded with. This is the case for
Reliabilities all samples separately and together, with
varimax and oblique rotation, and with Eigen
Reliabilities were examined subsequent to values greater than one or forced factors. Due
performing the factor analyses. The results are to space constraints, all 32 factor structures
presented in Table 5. Some reliabilities were cannot be presented. Explanation or
extremely low (.3405) and some are relatively classification of each component is not
high (.8131). Curiously, the data shows no possible. The components' items have no
consistent pattern across the samples. In other pattern or similarities. There was absolutely no
words, masculinity has the only adequate theoretical and empirical structure which
reliabilities in sample two, but is the second emerged.
lowest in sample one, both of which were
American samples. Confirmatory Fador Analysis

Exploratory Factor Analyses Next, structural equations modeling was used to

analyze the data because of its ability to be used
As mentioned, factor analyses were examined as a confirmatory technique, instead of as an
in many possible ways. Data are presented for exploratory technique, even though Hofstede
each sample separately and then are presented did not originally employ this technique. Since
by combined samples. As can be seen in the relationships and constructs had already
Table 6, in all samples, when the number of been established by Hofstede, the model just
components is not forced, the instrument needed to be analyzed to assess the adequacy of
explained around 70 percent of the variance. the model. The three previous mentioned
However, rotation could not account for samples were utilized to test the model.
coverage many times, and when it could, the
number of factors ranged between 15 and 20. The results of the model showed an inadequate
When the data were forced into four fit. The chi-squared/df ratio was equal to 5.345

TABLES
ReUabilities (Study 2)

Sample 1 Sample 2 (US non- Sample 3 All samples

(US graduate students) student adults) (Australian) together

Masculinity .5254 .7593 .8131 .7574

Individualism .4816 .3703 .6527 .4920

Power Distance .6974 .3510 .7003 .7417

Uncertainty .6330 .3405 .6845 .5655

Avoidance

11 Marketing Management Journal, Fall 2008

A Re-Inquiry of Hofstede's Cultural Orr and Häuser

TABLEÓ
Exploratory Factor Analyses (Study 2)
Sarnple 1 Sample 2 Sample 3 All samples together
(US graduate students) (US non-student adults) (Australian)

Number Variance Number of Variance Number of Variance Number of Variance

of factors explained factors explained factors explained factors explained

Varimax 19 70% Rotation 69.6% Rotation 69.2% 15 60%

rotation/ Eigen could not could not
values greater converge converge in
than one in 100 100
iterations iterations

Oblique rotation/ Rotation 70% 20 69.6% Rotation 69.2% 15 60%

Eigen values could not could not
greater than one converge converge in
in 100 100
iterations iterations
Varimax 4 31.7% 4 30% 4 37% 4 36%
rotation/ Four
"forced"
components
Oblique rotation/ 4 30% 4 30% 4 37% 4 36%
Four "forced"
components

(8493 with 1589 d.f.). Ratios under three The results of the variance explained by each
indicate an acceptable fit (Carmines and Mclver construct had even worse results. The highest
1981). The Root Mean Square Error of variance extracted was for the uncertainty
approximation (RMSEA) by Browne and avoidance construct (16.54 percent), followed
Cudeck (1993), which is a goodness of fit by power distance (10.33 percent), then by
measure that accounts for model complexity, individualism (10.33 percent), and then by
was 0.85. Browne and Cudeck (1993) state that masculinity having the lowest total variance
RMSEA values of about .05 or less indicate a extracted (7.97 percent). According to Hair et
close fit of a model in relation to the degree of al. (1998), guidelines suggest that the variance
fi"eedom. Likewise, the normed fit index extracted value should exceed .50 for a
(Bentler and Bonnett 1980) was 0.887, which construct. None of the constructs had value
should be above 0.90, which indicates and above this percentage.
acceptable level of fit (Hair, Anderson, Tatham
and Black 1998). Results of these analyses are In regards to factor loadings, some items have
presented in Tables 7 and 8. high factor loadings and are significant.
However, a larger number have very low
The reliabilities and the variances explained by loadings and are not significant. Amazingly, not
each of the latent contructs are presented in one single power distance item is significant.
Table 8. The highest reliability was for the While masculinity, individualism, and
masculinity construct (a = 0.73), followed by uncertainty avoidance may be adequate
power distance (a = 0.58), then by Uncertainty constructs (by these criteria only) once a few
avoidance (a = 0.40), and then with imnecessary items are eliminated, power
individualism having the lowest reliability (a = distance has clear empirical problems.
0.33). According to Hair et al. (1998), the
indicator reliabilities should exceed .50, which Additionally the correlations among latent
roughly corresponds to a standardized loading constructs are provided in Table 9. The most
of 0.70. striking correlation is that of masculinity and

Marketing Management Joumal, Fall 2008 12

A Re-Inquiry of Hofstede's Cultural.., Orr and Häuser

TABLE 7
CFA Results
Masculinity Individualism Power Distance Uncertainty Avoidance T-value
Reliability = 0.73 Reliability = 0.3 Reliability = 0.58 Reliability = 0.40
Variance Extracted = Variance Extracted = Variance Extracted = Variance Extracted =
7.97% 9.15% 10.33% 16.54%
0.55 12.212***
mf2_l 0.50 10.099***
mf3_l 0.66 12.224***
mf4_l 0.64 11.968***
0.58 11.241***
mf6_ 0.66 12.190***
mf7_ 0.62 11.652***
mf8_ 0.64 11.998***
mf9_ 0.57 11.031***
mflO 0.66 12.216***
mm 0.70 12.682***
-0.10 -2.282*
-0.02 -0.503
mfl4 0.04 0.832
mfl5 -0.08 -1.878
-0.04 -0.991
-0.15 -3.455***
-0.13 -2.97**
-0.05 -1.101
id 0.64 13.368***
ic2 0.55 11.858***
ic3. 0.54 11.767***
ic4. 0.63 13.366***
ic5. 0.58 12.329***
ic6 0.08 1.825
ic7. -0.03 -0.757
ic8. -0.05 -1.262
ic9_ 0.05 1.150
iclO_ 0.02 0.418
icll_ -0.01 -0.245
icl2_ -0.04 -0.906
icl3_ -0.09 -2.043*
icl4_ -0.07 -1.728
icl5_ -0.22 -5.047***
pdl_ 0.05 0.545
pd2_ -0.09 -0.998
pd3_ 0.30 1.12
pd4_ 0.75 1.136
pd5_ 0.19 1.101
pd6_ 0.52 1.133
pd7_l -0.10 •1.013
pd8_l 0.33 1.126
pd9_l 0.17 1.091
pdIO_ -0.07 .0.909
pdll_ 0.29 1.121
pdl2_ 0.55 1.133
pdl3_ 0.46 1.132
pdl4_ 0.79 1.136
pdl5_ 0.32 1.125
pdl6_ 0.62 1.134
pdl7_ -0.73 -1.135

13 Marketing Management Journal, Fall 2008

A Re-Inquiry of Hofstede's Cultural.... Orr and Häuser

TABLE 7
CFA Results (continued)
Masculinity Individualism Power Distance Uncertainty Avoidance T-value
Reliability = 0.73 Reliability = 0.3 Reliability = 0.58 Reliability = 0.40
Variance Extracted = Variance Extracted = Variance Extracted = Variance Extracted =
7.97% 9.15% 10.33% 16.54%
ua7_ 0.82 16.436***
ua6_ 0.11 2.493*
ua5 0.77 15.673***
ua4_ 0.45 10.032***
ua3 -0.15 -3.386***
ua2 -0.01 -0.105
ual_ -0.08 -1.862
*p < .05
**P<.01
***p < .001
Items listed in Table 4.

TABLE 8
CFA Results: Correlations Among Latent Constructs

Uncertainty
Masculinity Individualism Power Distance Avoidance
Masculinity 1.000
Individualism 0.987 1.000
Power Distance -0.218 -0.093 1.000
Uncertainty Avoidance -0.012 -0.059 -0.651 1.000

individualism (0.987). Perhaps this is because high in uncertainty avoidance want clear rules
Hofstede used the same items to measure each and regulations so that there is less uncertainty
construct. Hofstede states, "...reversing the to deal with.
sign of the scores (for the items for
individualism), I have called this dimension DISCUSSION
'Masculinity'" (Hofstede 1984, p 189). Items
that load negatively on a construct are not a This study has attempted to empirically assess
separate construct, just the opposite "pole" of the validity of Hofstede's cultural dimensions.
that construct. Power distance and Uncertainty Substantial questions have arisen in this
avoidance also have a high correlation between analysis as to the reliability and validity of
them (-0.651). Many of the power distance Hofstede's methodology and instrumentation.
items relate to the supervisor's responsibility to In fact, the analysis discussed in this re-inquiry
establish clear rules and regulations (e.g., does seem to reify many of the limitations of
"Employees should always be told very clearly Hofstede's work discussed earlier.
their duties and responsibilities, and how to
perform their jobs" and "A good manager gives First, the samples used in this analysis were
his employees detailed and complete comprised of diverse individuals who were not
instructions as to how they should do their jobs; part of an over-arching collectivity, namely
he does not merely give general directions and IBM employees. Thus the impact of a
depend on them to work out the details"). corporate culture was not in play here. Not
Conceptually, one can see how these items only does this minimize the corporate-wide
correlation with uncertainty avoidance. People
Marketing Management Journal, Fall 2008 14
A Re-Inquiry of Hofstede's C u l t u r a l . . . . Orr and Häuser

socialization effects as to how one should think assigned to an entire group or in this case
and act within that corporate culture, it culture. This analysis discussed supports the
minimizes the possibility of socially desirable need to understand differences at the societal
respondents by the respondents. level. Currently, this lack of understanding of
the causality between the individual level and
Second, limitations around the lack of the the socio-structural level precludes any clear
mutual exclusivity of the dimensions also indication of what is actually causing
surfaced in this re-inquiry. This analysis found differences within and across Hofstede's
significant overlap within and across many of dimensions.
the constructs. Thus, with the samples used
herein, there were no clearly identifiable factors Finally, the impact of the Internet and relatively
supporting the instrumentation and, to a large seamless global communications on a society's
degree, the methodology in its current state. cultural stance cannot be over-estimated in the
early 21^ centuiy. Short of extreme
Third, and probably most importantly, the governmental contol, it is nearly impossible
analysis questions the relevance of the original for individuals in any culture to not have access
dimensions and their meaning to 21^' century to and be infiuencod by information on other
businesses and individuals. What do the cultures, attitudes, and behaviors. Future
dimensions mean to individuals within and research on Hofstede's cultural dimensions
across different cultures? What about sub- must investigate the impact of global
cultural differences that exist in many countries communications on cultural dimensions and
and regions? Most importantly, what effect do individual responses to them.
traditional social institutions have on the
dimensions defined by Hofstede? For example, CONCLUSIONS
strong religious dogma and practices in a
culture will most likely strongly impact an Hofstede's seminal work has been the
individual's perception of individualism as well benchmark for cultural analysis for the last
as culturally sanctioned definitions of three decades. However, it has been subject to
masculine and feminine roles. Similarly, legal, criticism on both tlie theoretical and empirical
economic, and educational institutions within levels. The intent of this investigation was to
the given social structure will dramatically test Hofstede's constructs with a non-
influence how one responds to Hofstede's homogenous population (i.e., individuals that
dimensions. were not from one company only) in order to
ascertain the validity and reliability of the
This analysis also tends to support a number of measures. In order to do so a number of
Hofstede's critics as to the applicability of the samples drawn fi-om American, Far Eastern
four (or five) dimensions. In general, the Asians, and Australians were used. Analyses
dimensions only attempt to measure cultural were performed into the face validity,
differences at the individual level and are convergent, and discriminant validity of
therefore psychologically reductionistic. Cross- Hofstede's constructs. Likewise, Hofstede's
cultural analysis requires an understanding the factors were subjected to exploratory and
impact of the socialization and other confirmatory factor analysis where they
sociological factors that brought these about. performed poorly. This investigation has
As a matter of fact, Hofstede has long concluded that Hofstede's factors overlap
contended that an ecological fallacy is significantly and do not share a common factor
contained in the cultural dimensions. If structure within or between cultures.
anything, the current analysis suggests that a
reverse ecological fallacy may be the case While it is outside the scope of this
where individual characteristics are being investigation, it appears that Hofstede's

IS Marketing Management Journal, Fall 2008

A Re-Inquiry of Hofstede's Cultural Orr and Häuser

theoretical constructs need to be thoroughly re- both validity and reliability issues with the
examined within the context of early 21^' original constructs and instrumentation, it is not
century cross-cultural attitudes and patterns of our intent to denigrate Hofstede's original
behavior. Cross-cultural relationships (positive conceptualization. Instead, we recommend that
and negative) have changed dramatically over additional research be undertaken to build on
the past quarter of a century, be they political, Hofstede's cross-cultural dimensions to better
economic, or from a business perspective. adapt them to the 21*' century global
Worldwide political systems, such as environment. As such, each construct should
communism, have dramatically lost their be thoroughly re-examined and both be
influence since Hofstede first posited his theoretically and operationally redefined within
cultural dimensions. Free market economies contemporary cross-cultural and business
have taken a foothold (to varying degrees) in environments.
many cultures while businesses have become
more global in their reach and influence. Hofstede's work has contributed significantly
to the foundations of cross-cultural analysis and
At the same time, the changes mentioned above understanding. But like any good model, it
have exacerbated within society changes in needs to constantly be re-examined, re-defined,
many regions of the world. As political and and adapted to the current envirorunent.
economic systems decayed, long constrained Therefore, ñiture studies should attempt to
cultural and sub-cultural differences have re- build on, strengthen, and adapt what has been
emerged, in a number of cases, to the point learned from Hofstede's seminal work.
where the country has been divided into a Perhaps as we have moved well into this new
number of smaller cultural or historically millennium very different cross-cultural values,
"tribal" based enclaves. With this in mind, attitudes, and behaviors exist.
Hofstede's original dimensions may be
inaccurate, or at the least, outdated in defining REFERENCES
contemporary cultural differences be they
within or across different cultures. Ashkanini, Mohammed G. A. (1984), A Cross-
Cultural Perspective on Work Related Value,
With regard to research implications, empirical Unpublished doctoral dissertation.
establishment of convergent, discriminant, and International University, San Diego.
nomological validity for the cultural Baehr, Melany E. (1954), "A Factorial Study of
dimensions are of first importance. If the the SRA Employee Inventory," Personnel
constructs are not defined empirically, then Psychology, 7, 319-336.
they cannot be measured. Likewise, if Baumgartel, Howard and Thomas Hill (1982),
Hofstede's dimensions cannot be "Geert Hofstede: Culture's Consequences:
operationalized, then they cannot be correlated International Differences in Work-Related
with other concepts or used in other studies to Values," Personal Psychology, 35(1),
have practical significance. The study of cross- 192-196.
cultural values is simply too important in this Bennett, Milton (1996), Better Together Than
time of globalization. Hofstede's dimensions , A-P-A-R-T: Intercultural Communication: An
are not reasonable empirically. Although there Overview, [videorecording]. Newtonville,
may be a conceptual gold mine imdemeath it Mass.: Intercultural Resource Corporation.
all, a theory is worthless to investigators if it Bentler, Peter M. and Douglas G. Bonnett
cannot be operationalized. (1980), "Significance Tests and Goodness of
Fit in the Analysis of Covariance Structures,"
In conclusion, the purpose of this re-inquiry Psychological Bulletin, 88, 588-606.
was to re-examine Hofstede's original
methodology to test the validity with a number
of diverse samples. While this study found
Marketing Management Journal, Fall 2008 16
A Re-Inquiry of Hofstede's Cultural.... Orr and Häuser

Berry, John W. (1997), "Preface." In Berry, J. Galton, Francis [1883] (1928). Inquiries into
W., Poortinga, Y. H. and Pandey, J. (EA), Human Faculty and its Development.
Handbook of Cross-Cultural Psychology, London: Dent.
Volume 1: Theory and Method. Needham Graves, Desmond (1986), Corporate Culture
Heights, MA: Allyn and Bacon. Diagnosis and Change: Auditing and
Browne, Michael W. and Robert Cudeck Changing the Culture of Organizations.
(1993), "Alternative Ways of Assessing London: Frances I'inter.
Model Fit," In Testing Structural Equation Hair, Joseph F., Ronald E. Anderson, Rolph L.
Models, Kenneth A. Bollen and Scott J. Tathan and William C. Black (1998),
Long, eds. Newbury Park, CA: Sage. Multivariate Data Analysis: (5* ed.). Upper
Carmines, Edward G. and John P. Mclver Saddle River, NJ: Prentice Hall.
(1981), "Analyzing Models with Unobserved Hamden-Turner, Charles and Fons
Variables," In Social Measurement, Current Trompenaars (1997), "Response to Geert
Issues, George W. Bohmstedt and Edgar F. Hofstede," International Journal of
Borgatta, (Ed), Beverly Hills: Sage. Intercultural Relations, 21(1) 149-159.
Chinese Culture Connection (1987), "Chinese Herzberg, Fredrick, Bernard Mausner, O.
Values and the Search for Culture-Free Peterson and Dora F. Capwell (1957), Job
Dimensions of Culture," Journal of Cross- Attitudes: Review of Research and Opinion.
Cultural Psychology, 18, 143-164. Pittsburgh, PA: Psychological Service of
Chow, Chee W., Michael D. Shields and Yoke Pittsburg.
Kai Chan (1991), "The Effects of Hinrichs, J. R. (19(í8), "A Replicated Study of
Management Controls and National Culture Job Satisfaction Dimensions," Personnel
on Management Performance," Accounting, Psychology. 21(4), 479-503.
Organizations, and Society, 209-226. Hofstede, Geert, A. L. Kraut and S. H.
De Cieri, Helen and Peter J. Dowling (1995), Simonetti (1971), Development of a Cote
"Cross-Cultural Issues in Organizational Attitude Study Questionnaire for
Behavior," In Cooper, C. L. and D. M. International Use, Corte Madera, CA: Select
Rousseau (Ed), Trends in Organizational Press.
Behavior. West Sussex, England: John Wiley (1975), The Stability of Attitude Survey
and Sons. Questions: In Particular Those Dealing with
Etzioni, Amitai (1975), A Comparative Work Goals, Unpublished paper, Brussels:
Analysis of Complex Organizations (Revised European Institute for Advanced Studies in
ed.). New York: The Free Press. Management.
Fernandez, Denise Rotondo, Dawn S. Carlson, (1980), Culture's Consequences:
Llee P. Stepina and Joel D. Nicholson, International Differences in Work Related
(1997), "Hofstede's Country Classification Values, Thousand Oaks, CA: Sage
25 Years Later," Journal of Social Publications, Inc.
Psychology, \37(^l), 43-54. (1983), "National Cultures in Four
Forss, Kim (1989), Comparative Management Dimensions: A Research-Based Theory of
in Public Administration: A Pilot Study of Cultural Differences Among Nations,"
Values and Attitudes in Two Organizations, International Studies of Management and
The Central Bureau of Statistics, The Organization, 13,46-75.
National Institute for Civil Service Training (1988), "The Confucius Connection: From
and Development and the National Tax Cultural Roots to Economic Growth,"
Board. Organizational Dynamics, 16,4-22.
French, John R. P. and Bertram Raven (1959),
"The Bases of Social Power," In D.
Cartwright (Ed), Studies in Social Power.
Ann Arbor Michigan: ISR.

17 Marketing Management Journal, Fall 2008

A Re-Inquiry of Hofstede's Cultural Orr and Häuser

(1994), Forward in U. Kim, Harry C. Kelleher, B. R. (2000), "A Cross-Cultural

Triandis, C. Kagitqibasi, Sang-Chin and G. Study of Work-Related Values: An
Yoon, (Ed), Individualism and Collectivism: Investigation Across Three Countries Using
Theory, Method and Applications, (pp. ix- Hofstede's Cultural Indices," Dissertation
xiii). Thousand Oaks, CA: Sage Publications, Abstracts International Section A:
Inc. Humanities and Social Sciences, 60(7-A),
(1996), "Riding the Waves of Commerce: 2555.
A Test of Trompenaars' 'Model' of National Lowe, S. (1994), "Hermes Revisited," Paper for
Cultural Differences," International Journal the ISPAB Conference, Bangkok.
of Intercultural Relations, 20(2), 189-198. Maldonado, E. M. (1983), "Mexican-American
(1997), "Riding the Waves: A Rejoinder," Participation in Law Enforcement in
International Journal of Intercultural California: Law Enforcement the Impact of
Relations, 21(2), 287-290. Cultural Related Value Differences,"
(2001), Culture's Consequences: Unpublished doctoral dissertation, Claremont
International Differences in Work Related Graduate School, Claremont.
Values, 2001 edition. Thousand Oaks, CA: Marshall, R. (1997), "Variances in Levels of
Sage Publications, Inc. Individualism across Two Cultures," Journal
Hoppe, Michael H. (1990), "Testing the of Cross-Cultural Psychology, 28(4),
Construct Validity of Geert Hofstede's 490-496.
Masculinity-Femininity Dimension on Elites Martinsons, M. G. and R. I. Westwood (1997),
from Nineteen Countries," Paper for Xlth "Management Information Systems in the
International Congress of Cross-Cultural Chinese Business Culture: An Explanatory
Psychology, Liege. Theory," Information and Management, 32
Hsu, Francis L. K. (1971), "Psychosocial (5), 215-229.
Homeostasis and Jen: Conceptual Tools for Merker, S. (1982), "Geert Hofstede: 'Culture's
Advancing Psychological Anthropology," Consequences'," Behavioral Science. 27 (2),
American Anthropologist, 11, 23-44. 195-197.
Huo, Pau Y. and Donna M. Randall, (1991), Naumov, Alexander I. and Shelia M. Puffer
"Exploring Subcultural Differences in (2000), "Measuring Russiem Culture Using
Hofstede's Value Survey: The Case of the Hofstede's Dimensions," Applied
Chinese," Asia Pacific Journal of Psychology: An International Review, 49(4),
Management, 892, 159-173. 709-718.
Institute for Research on Intercultural Parsons, Talcott and Edward A. Shils (1951),
Cooperation (2001), "Europeans in the 2000 Toward a General Theory of Actions,
Social Science Citation Index," Cambridge, Massachusetts: Harvard
http://cwis.kub.nl/~fsw 2/iric/index2.htm. University Press.
International Association of Cross-Cultural Pooyan. Abdullah (1984), "Acculturation,
Psychology (2007), "New IACCP Honorary Acculturative Stress, and their Relationships
Fellows," http://www.iaccp.org/news_ to Both Work and Non-Work Outcomes,"
page.php?thisID=441 andnewstype=news. Unpublished doctoral dissertation, Dallas:
Jahoda, G. and B. Krewer (1997), "History of University of Texas.
Cross-Cultural and Cultural Psychology," In Redding, S. Gordon (1994), "Comparative
J. W. Berry, Y. H. Poortinga, and J. Pandey, Management Theory: Jungle, Zoo, or Fossil
(Ed), Handbook of Cross-Cultural Bed?" Organization Studies, 15(3), 232-359.
Psychology, Volume 1: Theory and Method. Robertson, Christopher J. and James J.
Needham Heights, MA: Allyn and Bacon. Hof&nan (2000), "How Different Are We?
Kahle, Lynn R. (1983), Social Values and An Investigation of Confucian Values in the
Social Change: Adaptation to Life in United States," Journal of Managerial Issues,
America, New York: Praeger. 12, 34-47.

Marketing Management Journal, Fall 2008 18

A Re-Inquiry of Hofstede's Cultural.... Orr and Häuser

Rokeach, Milton (1973), The Nature of Human Verbeke, Willem (2000), "A Revision of
Values, New York: The Free Press. Hofstede et al.'s (1990) Organizational
S0ndergaard, Mikael (1994), "Research Note: Practices Scale," Journal of Organizational
Hofstede's Consequences: A Study of Behavior, 21(5), 587-602.
Reviews, Citations and Replications," Vroom, Victor H. (1964), Work and
Organization Studies, 15(3), 447-456. Motivation, New York, Wiley.
Schwartz, Shalom H. and W. Bilsky (1987), Warner, Malcolm (1981), "Culture's
"Toward a Universal Psychological Structure Consequences," Journal of General
of Human Values," Journal of Personality Management. 1 (1), 75-78.
and Social Psychology, 53(3), 550-562. Westwood, Robert G. and James E. Everett
Shackelton, Viv J. and Abbas H. Ali (1990), (1987), "Culture's Consequences: A
"Work Related Values of Managers: A Test Methodology for Comparative Management
of Hofstede's Model," Journal of Cross- Studies in Southeast Asia," Asia Pacific
Cultural Psychology, 21(1), 109-118. Journal of Management, 4 (3), 187-202.
Smith, Peter B., Shaun Dugan, Ark Peterson Wherry, Robert J. (1954), "An Orthogonal Re-
and Wok Leung (1998), "Individualism: Rotation of the Baehr and Ash Studies of the
Collectivism and the Handling of SRA Employees Inventory," Personnel
Disagreement. A 23 Country Study," Psychology, 1, 365-380.
International Journal of Intercultural Yeh, Ryh-Song (1988), "Values of American,
Relations, 22(3), 351-361. Japanese, and Taiwanese Managers in
Sopachitwattana, P. (2000), "An Examination Taiwan: A Test of Hofstede's Framework,"
of the Effects of Participative Management 48* Aimual Meeting of the Academy of
and Cultural Values on Organizational Management, Anaheim. Academy of
Commitment for American and Thai Management Best papers Proceedings 1988.
Managers," Dissertation Abstracts Yeh, Ryh-Song and John J. Lawrence (1995),
International Section A: Humanities and "Individualism aiid Confucian Djmamism: A
Social Sciences, 61(3), 1075. Note on Hofstede's Cultural Root to
T0nnies, Ferdinand (1963) [1887], Community Economic Growth," Journal of International
and Society, New York: Harper and Row. Business Studies, 26(3), 655-670.
Triandis, Harry C. (1982), "Culture's Yoo, Boonghee aiid Naveen Donthu (2002),
Consequences," Human Organization, 41 (1), "Books in Review," Journal of Marketing
86-90. Research, (39) August, 388-389.
Trompenaars, Fons (1993), Riding the Waves of
Culture: Understanding Diversity in Global
Business, New York: Irwin.
Tucker, Ledyard R. and Charles Lewis (1973),
"The Reliability Coefficient for Maximum
Likelihood Factor Analysis," Psychometrika,
38, 1-10.
Van Oudenhoven, Jan Pieter (2001), "Do
Organizations Reflect National Cultures? A
10-Nation Study," International Journal of
Intercultural Relations. 21(1), 149-159.
, L. Mechelse, and C. K. W. de Dreu
(1998), "Managerial Conflict Management in
Five European Countries: The Importance of
Power Distance, Uncertainty Avoidance, and
Masculinity," Applied Psychology: An
International Review, 47(3), 439-455.

19 Marketing Management Journal, Fall 2008