CPQ CulturalDimensions

Culture is a group-level phenomenon, but itinfluences individuals’ perceptions, valuesand behavior, especially with respect to socialinteraction. In fact, much of the field of crosscultural management is itself based on thislatter premise. In spite of its wide use of theconcept of culture, the field has not yet

adopted a theory of culture conceptualizedand operationalized at the individual level ofanalysis. Important research by Triandis (e.g.Triandis, 1988; Kim et al., 1994; Triandis etal., 1995) and Schwartz (e.g. Schwartz, 1992,1994; Schwartz and Bilsky, 1990) has pro-vided some of the major elements of an

Articles

Cultural Dimensions at theIndividual Level of AnalysisThe Cultural Orientations Framework

ABSTRACT This article describes a theoretically-grounded framework of culturaldimensions conceptualized and operationalized at the individual level of analysis, based onthe work of anthropologists Kluckhohn and Strodtbeck. We present empirical data gatheredfrom five countries – Canada, Mexico, the Netherlands, Taiwan, and the United States – toassess the validity of the framework. We then use the results to explore how the culturalorientations framework can add insight and new perspectives to critical questions in crosscultural management research.

KEY WORDS • comparative • culture • individual level • measurement • theory

Copyright © 2002 SAGE Publications(London, Thousand Oaks, CA and New Delhi)

1470-5958 [200212]2:3;275–295;029108

CCM International Journal of

Cross CulturalManagement2002 Vol 2(3): 275–295

Martha L. MaznevskiInternational Institute for Management Development,

Lausanne, Switzerland

Joseph J. DiStefanoInternational Institute for Management Development,

Lausanne, Switzerland

Carolina B. GomezFlorida International University, USA

Niels G. NoorderhavenTilburg University, The Netherlands

Pei-Chuan WuNational University of Singapore, Singapore

01_articles2(3) 10/9/02 10:36 am Page 275

individual-level approach by developing theories of individualism and collectivism inthe case of the former, and personal values inthe case of the latter. However, culture incor-porates many dimensions beyond individual-ism and collectivism, and many aspects of anindividual’s characteristics beyond personalvalues. These other elements of culture maybe related to psychological and social phe-nomena in important ways, but those rela-tionships cannot be captured with currentapproaches.

This article presents an additional pieceof the ‘culture in individuals’ puzzle: a frame-work for analyzing cultural dimensions ofassumptions about interaction with othersand with the environment. The framework isbased on the anthropological work of Kluck-hohn and Strodtbeck (1961). In this articlewe describe the framework, placing it in thecontext of cross cultural research, and con-duct a preliminary assessment of its validityusing questionnaire data designed to test it.We present comparative data from five coun-tries, and identify patterns in the results with-in and between countries that provide newinsights into current cross cultural issues inorganizational research. We conclude byproviding suggestions for future research.

The Cultural OrientationsFramework

The cultural orientations framework waspresented by Kluckhohn and Strodtbeck intheir book Variations in Value Orientations (1961).The development of the theory, guided inpart by Parsons and Shils’ general theory ofaction (Parsons and Shils, 1951), took placeover 10 years through rigorous content analy-sis of a generation’s worth of field studiesfrom around the world. Kluckhohn andStrodtbeck believed that anthropology’s tra-ditional emphasis on whole cultures painteda rather static and simplified picture, andthat only by also studying variance withincultures could researchers understand cultur-

al change and complexity. They proposedthe existence of a limited set of questions,called ‘cultural orientations’, which eachsociety must answer to operate in an effectiveand cooperative way, and a limited set ofpossible answers for each question, called‘variations’. Orientations are among themost basic questions that people from all cul-tures ask themselves: How do I think aboutpeople? How do I see the world? How do Irelate to other people? How do I use time?(Adler, 1997).

In this scheme, culture is defined as thepattern of variations within a society, or,more specifically, as the pattern of deep-levelvalues and assumptions associated with soci-etal effectiveness, shared by an interactinggroup of people. Kluckhohn and Strodtbeckand their research associates identified a setof six basic cultural orientations with two orthree possible variations each. The six valueorientations answer the following specificquestions:

1 What is the nature of human beings: arethey good, evil or neutral?

2 What is our relationship to nature: arewe subjugated to nature, in harmonywith nature, or do we have mastery overit?

3 What is our relationship to other humanbeings: is it lineal (ordered positionwithin groups), collateral (primacy givento goals and welfare of groups), orindividualistic (primacy given to theindividual)?

4 What is our primary mode of activity: isour basic orientation one of being-in-becoming, doing or reflecting?

5 How do we view time: do we focus onthe past, present, or future?

6 How do we think about space: is it pub-lic, private, or mixed?

The framework is shown in Table 1 (for moredetail on management applications see alsoAdler, 1997; DiStefano and Maznevski, 2000;Lane et al., 2000).1

International Journal of Cross Cultural Management 2(3)276

01_articles2(3) 10/9/02 10:36 am 76

Kluckhohn and Strodtbeck conducted aninitial test of the framework in five cultures inthe Southwestern United States – a sampleselected to differ in societal organization butto be relatively constant in physical settingand, to a great extent, institutional context.

The field test demonstrated that the proposedorientations and variations discriminatedamong cultures and explained important patterns of individual behavior and otheroutcomes within cultures. This combinationof well-developed theory and rigorous initial

Maznevski et al.: Cultural Dimensions at the Individual Level of Analysis 277

Table 1 Cultural orientations and dimensions*

I. Nature of humans

Good/Evil: The basic nature of people is essentially good (lower score) or evil (higher score).Changeable/Unchangeable: The basic nature of humans is changeable (higher score) from good to

evil or vice versa, or not changeable (lower score).

II. Relationships among people

Individual: Our primary responsibility is to and for ourselves as individuals, and next for ourimmediate families.

Collective: Our primary responsibility is to and for a larger extended group of people, such as anextended family or society.

Hierarchical: Power and responsibility are naturally unequally distributed throughout society;those higher in the hierarchy have power over and responsibility for those lower.

III. Relation to broad environment

Mastery: We should control, direct and change the environment around us.Subjugation: We should not try to change the basic direction of the broader environment around

us, and we should allow ourselves to be influenced by a larger natural or supernatural element.

Harmony: We should strive to maintain a balance among the elements of the environment,including ourselves.

IV. Activity

Doing: People should continually engage in activity to accomplish tangible tasks.Thinking: People should consider all aspects of a situation carefully and rationally before

taking action.Being: People should be spontaneous, and do everything in its own time.

V. Time

Past: Our decision criteria should be guided mostly by tradition.Present: Our decision criteria should be guided mostly by immediate needs and circumstances.Future: Our decision criteria should be guided by predicted long term future needs and

circumstances.

VI. Space

Public: The space around someone belongs to everyone and may be used by everyone.Private: The space around someone belongs to that person and cannot be used by anyone else

without permission.

* Adapted from Kluckholn and Strodbeck (1961). See Lane et al. (2000) for a discussion on implications for international business in general, and Maznevski and Peterson (1997) for a discussion on implications for multicultural teams.

01_articles2(3) 10/9/02 10:36 am Page 277

testing resulted in a strong conceptual frame-work.

Kluckhohn and Strodtbeck’s (1961) frame-work and its roots influenced the selection ofdimensions of some frameworks commonlyused today, therefore, some overlaps areapparent. For example, elements of theRelationships orientation are echoed in Hof-stede’s individualism and power distanceconcepts (Hofstede, 1980), and Trompenaars’individualism–communitarianism, achieve-ment–ascription, and equality–hierarchy di-mensions (Trompenaars, 1993). The Environ-

ment orientation is related to Trompenaars’inner direction–outer direction dimension,and the Activity orientation encompasses hisanalysis–integration dimension and much ofthe meaning of the sequential time–synchro-nized time dimension (Trompenaars, 1993).Hall’s dimensions of time and space (Hall,1966, 1973) are captured in the Kluckhohnand Strodtbeck dimensions of the samename.

However, three key assumptions underly-ing Kluckhohn and Strodtbeck’s approachdistinguish it from the others and make thecultural orientations framework particularlywell suited for research in cross cultural management. First, individuals are clearlyidentified as the ‘holders’ of the preferencesfor variations, and the cultural pattern isdefined by patterns among individuals’ pref-erences. Consistent with the framework,researchers can make hypotheses and testthem at the individual level of analysis, aggre-gate measures to develop descriptions of cul-tures, and examine variance both within andbetween cultures. Second, all dimensions arepresumed to be found in all societies, but eachsociety is proposed to exhibit, at the aggregatelevel, a defining rank order of elements with-in each orientation. This assumption allowsresearchers to analyze the dynamics withincultures as well as identify major aggregatetrends. Variations in patterns within culturesare assumed to be inevitable and even neces-sary for societies to function effectively as a

whole and to change and adapt over time.Third, the dimensions are proposed to beconceptually independent, even within orien-tations. For example, Relationships–individual

is independent from Activity–being, but so alsois Relationships–individual from collective, andActivity–doing from being. This aspect allowsanalysis of culture in more depth than is possible with a less complex, bipolar frame-work. It does not preclude individuals fromagreeing with two variations within an orien-tation. In fact, such complexity may allowresearchers a better understanding of culturaldifferences. For example by allowing indi-viduals the ability to agree with two variationswithin an orientation, we may discover thatthose who agree with both variations act dif-ferently from those who agree with only one.

The cultural orientations framework com-plements research by Triandis and colleagueson individualism/collectivism and allocen-trism/idiocentrism (e.g. Kim et al., 1994;Triandis, 1972, 1988; Triandis et al., 1995)and by Schwartz on values (Schwartz, 1992,1994; Schwartz and Bilsky, 1990; Schwartzand Sagiv, 1995). Certainly the assumptionsbehind Triandis’s work and dimensionsthemselves fit squarely into the cultural orien-tations framework, articulating in more detailtwo of the Relationships variations. The othercultural orientation dimensions, then, can beseen as adding to Triandis’s approach.

The cultural orientations framework isalso complementary to Schwartz’s and Hof-stede’s values approach in two importantrespects. First, values are beliefs that relate todesirable end states or behaviors and as suchtranscend specific situations and guide selec-tion or evaluation of behavior and events(Schwartz, 1992; Rokeach, 1973). Despitethis definition of values, both Schwartz’s andHofstede’s research incorporates how a focalindividual would like the world to work withassumptions about how the world really does

work. The cultural orientations frameworksimilarly focuses on how individuals believethe world should work and an individual’s


01_articles2(3) 10/9/02 10:36 am Page 278

assumptions about how the world works. Suchassumptions are naturally a reflection of theculture to which that individual belongs; assuch the individual is reporting on his or herculture. It is important to note that theseassumptions are typically not questioned, nor are they even normally discussed. Forexample, in a society dominated by theActivity–doing, most people would know that,‘of course’, humans’ basic nature is toachieve and keep busy, and perhaps if theability to achieve and work towards goalswere removed from people, they and theirsociety would suffer from severe trauma andbreakdown (see Marx’s Communist Manifesto inCowling, 1998).

Second, values generally focus on thevalue-holder him- or herself. Values aremotivators of individual behavior. For exam-ple, someone who values ‘helpful’ is moti-vated to be helpful, derives satisfaction frombeing helpful, and fulfills a personal needwhen being helpful. Cultural orientationsalso comprise the individuals’ assumptionsabout social organization and elements out-side the individual. The orientations serve asperceptual filters that screen information andpotential choices about behavior (Erez andEarley, 1993), and provide scripts for socialinteraction (Maznevski and Peterson, 1997).

The cultural orientations frameworkthrough an understanding of values will helpus understand individual motivations, andwill illuminate many elements of individualbehavior alone and in social settings, withinand across cultures (Schwartz, 1992, 1994;Schwartz and Bilsky, 1990; Schwartz andSagiv, 1995). In addition, the cultural orien-tations will provide an understanding ofsocial behavior patterns, organized systemsand decision-making.

Given the above review of literature andour comparison between the cultural orienta-tion and other approaches, we expect to findseveral important results when testing thecultural orientations framework with datafrom five countries. First, we expect the

variations within and across orientations tobe conceptually independent. Second, weexpect there to be variations within orienta-tions within each country included in thestudy. Finally, we expect to find between-country differences on the variations withinorientations. These differences should be inline with the research that has been con-ducted on cross cultural values differences(Hofstede, 1980; Ronen and Shenkar, 1985),but should also add richness to prior expla-nations. These results will provide validity tothe cultural orientations framework.

Methodology

To gain some initial empirical evidence forevaluating the utility of the cultural orienta-tions framework in cross cultural research,we measured the orientations and their vari-ations in five countries selected to be similarand different on cultural characteristicsaccording to previous research (e.g. Hof-stede, 1980): Canada, Mexico, the Nether-lands, Taiwan, and the United States. Wethen examined within-country characteristicsand between-country comparisons to identifypatterns, comparing them with previousresearch.

Instrument

We used the Cultural Perspectives Question-naire, version 4 (CPQ4: Maznevski et al.,1997) to measure 11 variations of four cultur-al orientations (Relationships, Environment,Nature of Humans, Activity). The instrumentconsisted of 79 single-sentence statementsand asked the respondent to record his or herstrength of agreement with each, on a scalefrom ‘1’ (strongly disagree) to ‘7’ (stronglyagree). Variations were measured withbetween five to eight items. Appendix 1shows sample items and the number of itemsfor each variation. To reduce response biasfrom proximity of similar items, items foreach variation were randomly distributedthroughout the questionnaire. The question-


01_articles2(3) 10/9/02 10:36 am Page 279

naire also asked a wide variety of demo-graphic questions, including country of birth,countries of residence (and how long), culturemost closely identified with, gender, age,occupation, and amount of formal education.

All respondents were surveyed in theirnative language: the Canadian and US sub-jects completed an English-language version,the Mexican subjects a Spanish version, theNetherlands subjects a Dutch version, andthe Taiwanese subjects a Chinese version. All translations were conducted using trans-lation–back translation procedures, withchecks from bilingual speakers for dialectappropriateness and subtle meanings of items(Brislin, 1980).

Sample and QuestionnaireAdministration

Respondents were either practicing business-people or business students at the seniorundergraduate or graduate level with signifi-cant working experience. Further demo-graphics are summarized in Table 2. Weadministered the questionnaires in controlledsettings as part of other research projects or during business/management educationsessions. To avoid making the assumptionthat culture and country are synonymous,

the questionnaire asks respondents severalquestions regarding their cultural identity.For each cultural response, respondents arealso asked to rate how typical their viewswere for that culture. We combined thesemeasures to select samples for this analysis,such that each country group in this study iscomposed only of individuals from that coun-try sample who identify themselves with thatcountry’s culture more closely than they dowith any other culture.

Data Analysis and Results

We conducted three sets of analyses. First, toexamine the properties of the framework weundertook factor analyses using principalcomponents analysis and structural equationmodeling, and correlation analyses. Next, weconducted paired t-tests between pairs ofwithin-orientation variations, within coun-tries. Finally, we conducted a MultivariateAnalysis of Variance (MANOVA) and subse-quent Univariate Analyses of Variance(ANOVAs) to test for differences amongcountries.

Only one demographic dimension – gen-der – affected scores systematically withinsamples. Gender did not affect patterns ofcorrelations among items; that is, males and


Table 2 Sample demographics

% Undergrad Average Averagebusiness % MBA % Working age years full

Total % Male students students full time categorya time workb

Canada 333 70.3 30.1 17.9 52.0 26–30 6.7Mexico 182 79.7 0 0 100.0 26–30 9.1Netherlands 164 73.2 100.0 0 0 < 25 N/ATaiwan 424 33.7 0 3.9 96.1 31–35 8.9United States 498 52.0 36.2 28.7 35.1 26–30 10.3

a Age was measured in 10 categories, where 1 was ‘25 or less’, 2 was ‘26–30’, 3 was ‘31–35’, etc., with 10 being ‘66or over’.b Average years full time work for respondents who were MBA students or currently working full time.

01_articles2(3) 10/9/02 10:36 am Page 280

females from each sample had the same pat-tern of correlations between items. Thereforewe did not control for gender when con-ducting analyses regarding within-personrelationships among the items. Gender wasrelated, though, to differences in mean scoreson some of the variations in some samples.2

When comparing mean scores within andacross countries we controlled for the gendereffect by using subsamples with the samenumber of male and female respondents.

Factor Analysis and InternalConsistency of Measures

To examine the structure of the data andcompare it to the model, we separately testedeach orientation with its respective variations(e.g. all Relationships items together).3 We firstconducted a principal components analysisin each individual country and in an aggre-gated sample with the same number ofrespondents from each country. We exploredthe factor structure within each orientation,and eliminated a total of 13 items (no morethan two from any variation) that did notload clearly on the intended factor. To assessthe theoretical framework, we conductedconfirmatory factor analysis with the revisedset of items for each orientation using LISREL8 (Jöreskog and Sörbom, 1993). Weevaluated the model in the aggregated bal-anced sample, then within each country sample. The results of the LISREL analysesare presented in Table 3, with results for thefull model in the first half of the table andresults for the revised model (reduced set ofitems) in the second half of the table. Eachanalysis reports the χ2 statistic for the full andnull models, the root mean square residual,and goodness of fit and adjusted goodness offit indices.

The standard for acceptable goodness offit and adjusted goodness of fit indices forestablished research is .95 or higher, while.90 is acceptable for early or exploratoryresearch (Hair et al., 1992). The models hadbetter psychometric properties when tested

with the aggregated sample than within theindividual countries. This is probably due tothe increased overall variance and samplesize when all countries are included, and thefact that items loaded with different strengthsin the various country samples. The goodnessof fit indices (GFI) for the aggregated data set ranged from .91 to .93. In our within-country analyses, the 20 GFI ranged from.84 to .94. We concluded that the results met or approximated the standards for earlyresearch although they did not reach thestandards for established research. We pro-ceeded to check for patterns in the data, rec-ognizing that for some variations the amountof measurement error may attenuate theresults.

To examine relationships among varia-tions, we calculated correlations for theaggregated sample and each of the five country samples, using the revised set ofitems for each variation. The correlations arereported in Table 4, which highlights within-orientation correlations.

The relationships among variations differwidely among the five samples. Even wherethe direction and significance of correlationsare the same, such as a positive correlationbetween Activity–doing and thinking andbetween Environment–mastery and harmony in allsamples, the magnitude of the correlationsvaries greatly. This observation reinforcesthe importance of adopting a more multi-dimensional and individual-level approach tocross cultural analysis to obtain a more com-prehensive picture of culture.

Analyses of Variance

Calculation of scores for means compar-isons Previous research has demonstrateda significant country effect on respondents’tendency to use different parts of the scale(Leung and Bond, 1989; Mullen, 1995; Singh,1995). The established procedure for remov-ing bias associated with scale response is within-person standardization (Leung andBond, 1989). We standardized the data with-


01_articles2(3) 10/9/02 10:36 am Page 281


Tab

le 3

LIS

RE

L an

alys

is o

f var

iatio

ns b

y or

ient

atio

n

Full

Mod

el –

All

Item

s R

evis

ed M

odel

– R

educ

ed S

et o

f Ite

ms

All

Can

ada

Mex

ico

Net

her

Tai

wan

USA

All

Can

ada

Mex

ico

Net

her

Tai

wan

USA

N =

825

N =

333

N =

182

N =

164

N =

424

N =

498

N =

825

N =

333

N =

182

N =

164

N =

424

N =

498

Rel

atio

nshi

psχ2

Ful

l71

1.91

460.

9833

8.96

281.

8663

9.80

487.

2164

9.47

421.

2328

3.15

273.

0675

5.27

771.

98df

206

206

206

206

206

206

186

186

186

186

186

186

p0.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

00χ2

Nul

l20

18.2

110

98.2

052

3.88

507.

5014

27.9

613

46.5

219

19.5

010

52.6

451

0.59

498.

7015

25.2

914

88.2

6df

231

231

231

231

231

231

210

210

210

210

210

210

Roo

t Mea

n Sq

Res

idua

l0.

060.

070.

080.

080.

080.

060.

060.

070.

080.

080.

080.

08G

oodn

ess

of F

it In

dex

0.92

0.89

0.85

0.87

0.87

0.92

0.92

0.89

0.87

0.87

0.86

0.87

Adj

Goo

dnes

s of

Fit

Inde

x0.

900.

860.

820.

840.

850.

900.

900.

870.

840.

840.

830.

84

Env

iron

men

tχ2

Ful

l84

7.42

392.

7126

4.04

357.

6052

2.72

523.

2171

9.58

350.

8519

0.73

336.

6271

2.06

793.

89df

186

186

186

186

186

186

167

167

167

167

167

167

P0.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

00χ2

Nul

l27

36.4

795

4.16

608.

6771

4.37

1514

.47

1419

.76

2598

.11

907.

5551

3.09

693.

3714

53.1

417

43.6

1D

f21

021

021

021

021

021

019

013

613

613

613

613

6R

oot M

ean

Sq R

esid

ual

0.08

0.07

0.07

0.09

0.07

0.07

0.08

0.07

0.07

0.09

0.08

0.08

Goo

dnes

s of

Fit

Inde

x0.

900.

900.

880.

840.

900.

900.

910.

900.

910.

840.

860.

86A

dj G

oodn

ess

of F

it In

dex

0.88

0.87

0.85

0.80

0.87

0.88

0.89

0.88

0.88

0.79

0.82

0.82

Act

ivity

χ2 F

ull

1185

.31

931.

1848

2.74

538.

7876

3.14

887.

9262

9.16

436.

3622

4.40

220.

6564

3.79

614.

01df

272

272

272

272

272

272

149

149

149

149

149

149

p0.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

000.

00χ2

Nul

l45

83.5

918

81.4

276

4.78

1004

.44

2061

.71

2317

.85

3890

.04

1290

.89

733.

0765

5.68

1828

.88

1894

.59

df30

030

030

030

030

030

017

117

117

117

117

117

1R

oot M

ean

Sq R

esid

ual

0.06

0.09

0.08

0.09

0.07

0.07

.06

0.08

0.07

0.08

0.07

0.08

Goo

dnes

s of

Fit

Inde

x0.

900.

830.

830.

810.

870.

87.9

30.

870.

890.

880.

860.

88A

dj G

oodn

ess

of F

it In

dex

0.88

0.79

0.80

0.77

0.84

0.85

.90

0.83

0.86

0.84

0.82

0.85

Hum

an N

atur

eχ2

Ful

l35

1.29

143.

7074

.34

111.

2319

2.00

194.

9831

6.36

110.

0074

.38

88.3

119

5.34

187.

93D

f43

4343

4343

4334

3434

3434

34P

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

χ2 N

ull

1423

.83

773.

1019

8.24

453.

3253

1.31

77.3

013

76.7

371

4.13

216.

4542

0.00

605.

7496

2.32

df55

5555

5555

5545

4545

4545

45R

oot M

ean

Sq R

esid

ual

0.08

0.08

0.08

0.09

0.08

0.07

0.08

0.07

0.09

0.09

0.09

0.07

Goo

dnes

s of

Fit

Inde

x0.

930.

930.

930.

900.

930.

930.

930.

940.

930.

910.

920.

93A

dj G

oodn

ess

of F

it In

dex

0.89

0.89

0.89

0.84

0.89

0.90

0.89

0.90

0.88

0.85

0.87

0.89

01_articles2(3) 10/9/02 10:36 am Page 282


Table 4 Correlations among Variations a,b,c

Indiv Coll Hier Har Mast Subj G/E Chge Do Be Thk

Individual –.11 .36** .10 .40** .05 .34** .03 .21** .33** .26**

.19** .20** .28** .27** .20** .34** .17** .26** .35** .26**–.08 .18** –.01 .21** .15** .39** .04 .24** .17** .22**

Collective .03 .10 .45** .17* .03 –.13 –.14 .20** .14 .09–.16** .35** .51** .43** .10* .12* .27** .50** .21** .48**–.01 .04 .35** .21** .03 –.09* .14** .25** .06 .21**

Hierarchy .23** .22** .08 .36** .16* .34** –.03 .37** .19* .31**

.28** –.01 .22** .25** .24** .19** .24** .34** .25** .35**

.12 .25** –.09* .16** .23** .36** .01 .23** .10* .26**

Harmony .11** .42** .07* .27** –.07 –.06 .02 .11 .16* .05

.03 .29** –.12* .49** –.04 .12* .30** .54** .15** .54**

.00 .23** .01 .32** .07 –.07 .18** .21** .14** .30**

Mastery .23** .24** .10** .40** –.21** .14 .10 .34** .16* .36**

.28** .13** .19** .14* –.02 .14** .32** .52** .25** .49**

.09 .18* .11 .47** –.12** .11* .20** .39** .00 .42**

Subjugation .16** .20** .45** .09* –.11** .23** –.19* .01 .28** .04

.17** .10 .28** .08 –.17** .23** .09 –.08 .38** –.07

.14 .20** .43** –.12 .01 .32** .01 .04 .27** .12**

Good/Evil .34** .13** .52** .13** .11** .50** .09 .38** .16* .17*

.38** –.03 .55** –.05 .13** .36** .17** .17** .26** .11*

.14 .12 .44** .08 .06 .56 .06 .18** .11* .25**

Changeable .09* .18** .21** .24** .20** .21** .26** .04 .02 –.02–.10 .10 .15** –.02 –.03 .21** .12* .33** .14** .35**

.09 .13 .12 .21** .15* .10 .27** .10* .01 .14**

Doing .27** .35** .29** .37** .45** .17** .33** .25** –.07 .35**

.33** .22** .34** .18** .37** .05 .25** .04 .07 .63**

.02 .33** .11 .28** .25** .15* .17* .18* –.10* .51**

Being .23** .15** .24** .12** .10** .27** .28** .12** .00 .06

.21** .10 .15** .12* .05 .21** .27** .11* .02 .08

.13 .18* .28** .07 .19* .36** .40** .18* .05 –.05

Thinking .22** .34** .33** .43** .43** .24** .36** .32** .62** .05

.22** .17** .29** .27** .24** .13* .31** .06 .40** .07–.03 .25** .09 .47** .39** –.01 .13 .19** .41** –.04

a Correlations in the bottom left triangle are in the following order: aggregated sample (N = 825, 165 from each

country); Canada (N = 333); Mexico (N = 182). Correlations in the top right triangle are in the following order:

Netherlands (N = 165); Taiwan (N = 428); United States (N = 498).b Statistical significance: *p < .05; **p < .01c Correlations within solid-lined boxes are within-Cultural Orientation correlations.

Act

ivity

Hum

an N

atur

eE

nvir

onm

ent

Rel

atio

nshi

ps

01_articles2(3) 10/9/02 10:36 am Page 283


Tab

le 5

Test

s of

with

in-c

ount

ry d

iffer

ence

s

Can

ada

Mex

ico

Net

herl

ands

Tai

wan

USA

N =

190

N =

70

N =

86

N =

286

N =

472

Mea

n a

With

inM

ean

aW

ithin

Mea

n a

With

inM

ean

aW

ithin

Mea

n a

With

in(S

D)

Ori

ent’n

b(S

D)

Ori

ent’n

b(S

D)

Ori

ent’n

b(S

D)

Ori

ent’n

b(S

D)

Ori

ent’n

b

Rel

atio

nshi

ps

Indi

vidu

alis

m.2

77I=

C.1

09I<

C.1

23I<

C.0

59I<

C.2

79I=

C(.3

66)

I>H

(.372

)I>

H(.2

95)

I>H

(.304

)I>

H(.3

64)

I>H

Col

lect

ive

.326

C=

I.3

06C

>I,

H.2

80C

>I,

H.3

39C

>I,

H.2

97C

=I

(.384

)C

>H

(.285

)(.3

41)

(.293

)(.3

80)

C>

HH

iera

rchi

cal

– .52

0H

<I,

C– .

309

H<

I,C

– .29

6H

<I,

C– .

345

H<

I,C

– .48

6H

<I,

C(.3

27)

(.332

)(.2

90)

(.285

)(.3

51)

Env

iron

men

t

Har

mon

y.5

54H

=M

.512

H=

M.5

67H

>M

,S.5

91H

>M

,S.5

15H

<M

(.221

)H

>S

(.211

)H

>S

(.271

)(.2

37)

(.246

)H

>S

Mas

tery

.550

M=

H.4

92M

=H

.326

M<

H.3

10M

<H

.560

M>

H,S

(.290

)M

>S

(.234

)M

>S

(.326

)M

>S

(.275

)M

>S

(.274

)Su

bjug

atio

n– .

908

S<H

,M– .

886

S<H

,M– .

753

S<H

,M– .

767

S<M

,H– .

878

H<

M(.2

38)

(.270

)(.3

56)

(.289

)(.2

39)

H>

S

Act

ivity

Doi

ng.1

81D

>B

,T.2

75D

>B

.117

D>

B.3

24D

>B

.183

D>

B(.3

20)

(.239

)D

<T

(.394

)D

=T

(.251

)D

<T

(.322

)D

=T

Bei

ng– .

370

B<

D,T

– .65

1B

<D

,T– .

369

B<

D,T

– .61

1B

<D

,T– .

442

B<

D,T

(.471

)(.4

16)

(.432

)(.4

60)

(.514

)T

hink

ing

.083

T<

D.4

95T

>D

,B.0

95T

=D

.474

T>

D,B

.156

T=

D(.4

26)

T>

B(.2

53)

(.353

)T

>B

(.261

)(.3

51)

T>

B

aE

ach

coun

try

sam

ple

has

an e

qual

num

ber

of m

ale

and

fem

ale

resp

onde

nts.

Dat

a ar

e st

anda

rdiz

ed w

ithin

per

son,

with

in o

rien

tatio

n (se

e te

xt fo

r de

tails

), an

d ar

e th

eref

ore

stan

dard

sco

res

with

mea

n =

0 a

nd s

tand

ard

devi

atio

n =

1 w

ithin

eac

h or

ient

atio

n.b

Sign

ifica

nce

test

ed u

sing

pai

red

t–te

sts

of p

airs

of v

aria

tions

with

in e

ach

orie

ntat

ion.

Diff

eren

ces

note

d ar

e si

gnifi

cant

at p

< .0

5. I

nitia

ls de

note

var

iatio

ns w

ithin

the

sam

eor

ient

atio

n.

01_articles2(3) 10/9/02 10:36 am Page 284

in person and within orientation, such thateach individual’s scores for the items of thetwo or three scales within each orientationhave a mean of 0 and a standard deviation of 1.4

Comparisons of means To test the pre-ferred rankings of variations within orienta-tions within each culture, we conducted pairedt-tests for each pair of variations within eachof the Relationships, Environment, and Activity

orientations. Within all orientations of allsamples, patterns of preferences were statisti-cally significant. These results are shown inTable 5 and will be discussed in the next section.

To test for between-country differences,we first conducted a MANOVA, whichresulted in an approximate F of 19.31(Pillais), p < .0001. Univariate F tests showedsignificant differences between countries forall variations except for Relationships–collective.We continued with post hoc comparisonsusing Dunnett’s T, which is appropriatewhen variances are unequal (Levene statisticsshowed that the variances were unequal forall variations except Environment–harmony andHuman Nature–good/evil). Results for between-country comparisons are shown in Table 6and will be discussed below.

Discussion

Our results lead us to two major conclusions.First it appears that the items developed torepresent the cultural orientations frame-work demonstrate reasonable, although notideal, internal consistency measures. Second,the within-country comparisons and thebetween-country comparisons appear toshed light on cross cultural research in sever-al aspects. Specifically, in this discussion wehighlight three important aspects: the com-parison of our results with patterns in previ-ous research; an examination of the patternsobserved in light of the convergence/diver-gence debate; and an exploration of possible

cultural transition and regional variation asreflected in the data.

Comparison with EstablishedFindings

Well-established broad findings have beenpublished on only a few elements addressedby this framework and sample. Here we willaddress country clusters, individualism–col-lectivism, hierarchy (power distance), andcharacteristics of Chinese culture.

Country clusters Ronen and Shenkar(1985) identified eight clusters of country cul-tures, based on a meta-analysis of previousattitudinal and perceptual research. Althoughthe Netherlands was not included in theiranalysis, the other four countries we sampledwere included, with Canada and the UnitedStates being in the Anglo cluster, Mexico inthe Latin American one, and Taiwan in theFar Eastern one. In their Smallest SpaceAnalysis, Latin American and Far Easterncultures were located next to each other,implying relatively close cultural proximitywhen compared with Anglo or Germanic cul-tures. Our Canadian and American samplesdiffered on only one of the 11 dimensions, andtheir within-country patterns were highlysimilar. The Mexican and Taiwanese samplesdiffered on three of the 11 variations, andtheir relative preferences for Relationships andActivity were similar, but were not similar forthe Environment. On the other hand, the twoAnglo countries differed from at least one ofMexico or Taiwan on nine variations. Thiscorrespondence with Ronen and Shenkar’s(1985) meta-analysis provides some prelimi-nary support for the validity of the culturalorientations framework.

Individualism–collectivism patterns Consistent with previous research (e.g. Hof-stede, 1980; O’Grady and Lane, 1996), ourdata show that the average or typical respon-dents of the United States and Canadian


01_articles2(3) 10/9/02 10:36 am Page 285


Tab

le 6

Test

s of

bet

wee

n-co

untr

y di

ffere

nces

Can

ada

Mex

ico

Net

herl

ands

Tai

wan

USA

N =

190

N =

70

N =

86

N =

286

N =

472

Mea

n a

Cou

ntry

Mea

n a

Cou

ntry

Mea

n a

Cou

ntry

Mea

n a

Cou

ntry

Mea

n a

Cou

ntry

(SD

)D

iffsb

(SD

)D

iffsb

(SD

)D

iffsb

(SD

)D

iffsb

(SD

)D

iffsb

Rel

atio

nshi

ps

Indi

vidu

alis

m.2

77.1

09.1

23.0

59.2

79(.3

66)

C>

M,N

,T(.3

72)

M<

C,U

(.295

)N

<C

,U(.3

04)

T<

C,U

(.364

)U

>M

,N,T

Col

lect

ive

.326

.306

.280

.339

.297

(.384

)(.2

85)

(.341

)(.2

93)

(.380

)H

iera

rchi

cal

– .52

0– .

309

– .29

6– .

345

– .48

6(.3

27)

C<

M,N

,T(.3

32)

M>

C,U

(.290

)N

>C

,U(.2

85)

T>

C,U

(.351

)U

<M

,N,T

Env

iron

men

t

Har

mon

y.5

54.5

12.5

67.5

91.5

15(.2

21)

(.211

)(.2

71)

(.237

)T

>U

(.246

)U

<T

Mas

tery

.550

.492

.326

.310

.560

(.290

)C

>N

,T(.2

34)

M>

N,T

(.326

)N

<C

,M,U

(.275

)T

<C

,M,U

(.274

)U

>N

,T

Subj

ugat

ion

– .90

8– .

886

– .75

3– .

767

– .87

8(.2

38)

C<

N,T

(.270

)M

<T

(.356

)N

>C

,U(.2

89)

T>

C,M

,U(.2

39)

U<

N,T

Hum

an N

atur

e

Cha

ngea

ble

1.62

61.

794

2.04

81.

873

1.82

3(.7

35)

C<

N,T

,U(.4

86)

(.730

)N

>C

(.769

)T

>C

(.607

)U

>C

Goo

d/E

vil

– .25

2– .

215

– .24

1– .

040

– .24

6(.2

75)

C<

T(.2

75)

M<

T(.2

80)

N<

T(.2

67)

T>

C,M

,N,U

(.268

)U

<T

Act

ivity

Doi

ng.1

81.2

75.1

17.3

24.1

83(.3

20)

C<

M,T

(.239

)M

>C

,N,U

(.394

)N

<M

,T(.2

51)

T>

C,N

,U(.3

22)

U<

M,T

Bei

ng– .

370

– .65

1– .

369

– .61

1– .

442

(.471

)C

>M

,T(.4

16)

M<

C,N

,U(.4

32)

N>

M,T

(.460

)T

<C

,N,U

(.514

)U

>M

,T

Thi

nkin

g.0

83.4

95.0

95.4

74.1

56(.4

26)

C<

M,T

(.253

)M

>C

,N,U

(.353

)N

<M

,T(.2

61)

T>

C,N

,U(.3

51)

U<

M,T

aE

ach

coun

try

sam

ple

has

an e

qual

num

ber

of m

ale

and

fem

ale

resp

onde

nts.

Dat

a ar

e st

anda

rdiz

ed w

ithin

per

son,

with

in o

rien

tatio

n (se

e te

xt fo

r de

tails

), an

d ar

e th

eref

ore

stan

dard

sco

res

with

mea

n =

0 a

nd s

tand

ard

devi

atio

n =

1 w

ithin

eac

h or

ient

atio

n.b

Sign

ifica

nce

test

ed u

sing

Dun

nett

’s T

for

post

hoc

com

pari

sons

whe

n va

rian

ces

are

uneq

ual.

Diff

eren

ces

note

d ar

e si

gnifi

cant

at p

< .0

5. I

nitia

ls de

note

cou

ntri

es.

01_articles2(3) 10/9/02 10:36 am Page 286

samples are among the most individualistic inthe study, while the average respondentsfrom Mexico and Taiwan are among theleast individualistic. However, we can gainmuch more information by separating indi-vidualism from collectivism, rather than con-ceptualizing them as two ends of a bipolarconstruct (see also Triandis et al., 1995). Forexample, the respondents from Mexico andTaiwan preferred collectivism over individual-ism, while in the other countries the two vari-ations were equally preferred by the respon-dents. Respondents from the five countriesdid not differ from each other statistically ontheir levels of collectivism, which may be dueto measurement error and/or high varianceamong individuals. But it also opens us to therealization that individuals from countriessuch as Canada, the Netherlands, and theUnited States may not be ‘pro-individual-ism/anti-collectivism’ to the same extent that individuals from other countries, such as Mexico and Taiwan, are simply ‘non-individualist’. We note in this regard that thedomestic management practice and researchliterature in the United States is replete withstudies on teams (Cohen and Bailey, 1997),while in the international arena the countryis assumed to be the least team-oriented inthe world. It may be that the preference forcollectivism is not as low in the United Statesas international research has previously sup-posed.

Hierarchy This dimension is similar indefinition to Hofstede’s dimension of powerdistance. Consistent with his results (1980),Canadian and American respondents in thisstudy preferred hierarchy less than Mexicanand Taiwanese ones did. However, whileHofstede found Netherlands to be low onpower distance, compared with our othersamples the Dutch respondents were thehighest. It may be that this group of Dutchbusiness students represents a specific sub-culture within the Netherlands. As with theindividualism/collectivism dimensions, this

demonstrates a potential strength of the cultural orientations perspective. Since theframework conceptualizes cultural variationat an individual (as well as group) level ofanalysis, it allows researchers to propose andmeasure finer grained pictures, includingsmaller subcultures within larger cultures.Further research on this subculture would besuggested from our results.

East Asian cultures East Asian cultureshave often been characterized as very differ-ent from those in the West (e.g. ChineseCulture Connection, 1987), with a stronginfluence from the religions and philosophiesof Confucianism, Taoism and Buddhism(Chew and Putti, 1995; Redding, 1993). Interms of the cultural orientations framework,these foundations predict dominant prefer-ences for hierarchy and collectivism for Relation-

ships, harmony with a secondary emphasis onmastery concerning the Environment, and athinking mode of Activity for these cultures(Bond and Hwang, 1986; Hwang, 1985; Oh,1991; Yau, 1988). Not completely as pre-dicted, the Taiwan sample showed a firstpreference for collectivism in terms of Relation-

ships, with hierarchy as the lowest preference;however, their preference for hierarchy washigher than that of respondents from Canadaand the United States, suggesting a relativeacceptance. The Taiwanese sample did showa definite and strong preference of harmony

over mastery over subjugation, and for thinking

over other types of Activity modes. This pat-tern of results fits the descriptions of EastAsian philosophy and religion very well.

The fact that these results are, for themost part, consistent with past findings usingalternative frameworks or other ways tounderstand culture lends support to thevalidity of the cultural orientations frame-work. The latter framework, though, enablesus to examine many more dimensions and tocompare results within and between culturesto gain more insight into the complexities ofthe phenomenon.


01_articles2(3) 10/9/02 10:36 am Page 287

Cultural Patterns:Industrialization orCrossvergence?

The rankings of elements within culturesdemonstrated a very striking pattern. Foreach of the three orientations for which vari-ations are ranked (Relationships, Environment,and Activity), the same orientation was by farthe lowest for the average respondent in allcountry samples (hierarchy, subjugation, andbeing, respectively). This pattern was not evi-dent among Kluckhohn and Strodtbeck’s(1961) US samples, nor did they suggest itwould be found across other cultures. Thissimilarity implies a possible convergenceamong the individuals in these cultures, andwe suggest the convergence is associated witha global industrialization culture. The busi-ness environment is changing more andmore rapidly, with competitors entering,leaving, and changing industries. Researchon organizational structure shows thatstrongly hierarchical organizations find it difficult to adapt to change (e.g. Burns andStalker, 1961). Perhaps a culture that prefershierarchy to any other form of organizing isnot able to adapt to changes well enough tocompete in business among industrializedeconomies. Furthermore, a culture that pre-fers subjugation may not be able to lead in anyof these industry changes, and thus may notbe as strongly industrialized. Global businessrequires coordination of logistical and meet-ing schedules among many people in differ-ent locations. A culture that prefers being toother modes of Activity may not be able tointeract reliably with the rest of the globaleconomy on a daily basis, and may not beable to support a growing industrial econ-omy.

Although on average respondents in allcountries ranked hierarchy, subjugation and being

as lowest, there were still significant differ-ences among respondents in different coun-tries with respect to the strength of these vari-ations. For example, although the Mexicans,

Dutch, and Taiwanese respondents in oursamples ranked hierarchy lower than both indi-

vidualism and collectivism, all three preferredhierarchy more than either the Canadians orAmericans did. This pattern would predictthat even though these groups may agree onsome common priorities, organizations andindividuals from the different cultures maystill encounter culture-based conflicts whentrying to work together.

Moreover, for these three orientations,different combinations of ordering of theother two variations within the orientationwere found in the samples. For example, withrespect to Activity, the average respondentfrom Canada preferred doing to thinking; theaverage respondents from Mexico andTaiwan preferred thinking to doing; and theaverage Dutch and American respondentsranked the two equally. This complex butconsistent pattern of results offers a new perspective to the convergence/divergencedebate. We suggest that this ‘global industri-alization pattern’ is an area of convergencefor cultures of industrialized countries; spe-cifically, cultures converge on the dimensionsof low hierarchy, low subjugation, and low being.This agreement of priorities facilitates theconduct of global coordination. Divergencestill occurs on other dimensions, and the disagreement on these priorities can lead toboth conflicts on how organizations shouldbe managed, and opportunities for synergyand innovation. Complex combinations ofconvergence and divergence, labeled ‘cross-vergence’ by Ralston and his colleagues(Ralston et al., 1993, 1997), certainly offer afruitful avenue for further investigation intocultural similarities and differences and theirimplications.

Mexico: Cultural Transition orRegional Variation?

In light of this pattern of industrialization, avery interesting picture arises with the resultsconcerning Mexico. Previous research andcharacterizations of Mexican culture have


01_articles2(3) 10/9/02 10:36 am Page 288

generally portrayed hierarchical and collective

Relationships (Condon, 1985; Davis, 1969;deForest, 1994; Hofstede, 1980; Kras, 1989;Rodríguez Estrada and Ramírez-Buendia,1992; Zurcher et al., 1965), a subjugation ori-entation to the Environment (Kluckhohn andStrodtbeck, 1961; Kras, 1989; RodríguezEstrada and Ramírez-Buendia, 1992), andbeing and possibly thinking modes of Activity

(Adler, 1997; Hall and Hall, 1990; Kluck-hohn and Strodtbeck, 1961; Kras, 1989).The Mexican sample in this study preferredhierarchy least and collectivism most; mastery overthe Environment equally to harmony, with bothover subjugation; and thinking over doing, withbeing as the least preferred mode of Activity.Some of these results are consistent with theprevious research. For example, althoughhierarchy is the least preferred Relationship, theMexican sample showed a stronger prefer-ence for hierarchy than did respondents fromCanada and the United States. A Canadianor American interacting in Mexico may find the Mexicans comparatively favorabletowards hierarchy. But given the compositeresults, it would hardly be right to character-ize this Mexican sample as hierarchy oriented.

A possible explanation for these resultslies in appreciating cultural change andregionalism. Most of the prior research onthe Mexican culture is either decades old, oris based on general knowledge drawn fromthe culture of southern Mexico. Mexico hasbecome more industrialized in the past twodecades, the effect of which has been notedin cultural shifts in studies conducted inMexico (Alduncin Abatia, 1993; Kras, 1989).For example, Mexicans appear to have amore proactive approach towards life(Alduncin Abatia, 1993; Schwartz, 1994)which may be reflected in the high scores ona doing orientation as well as a mastery orienta-tion. Furthermore, the northern regions have led this trend for quite some time(Rodríguez Estrada and Ramírez-Buendia,1992). Researchers have attributed the morecompetitive culture with a stronger work

ethic to both the extreme weather as well asthe US influence. The data for this studywere collected from business employees inthe northern region, and perhaps they reflectthe industrialized culture to which Mexico ischanging.

Summary Data related to these threeissues were selected to illustrate the potentialutility to cross cultural research of incorpo-rating a framework of cultural orientationsthat is theoretically grounded, compre-hensive, and allows conceptualization andmeasurement at the individual level of analy-sis. Osland and Bird (2000) describe theimportance of paying attention to and pro-cessing cultural paradoxes; that is, incidentsor patterns that are counter-predicted by anunderstanding of the overall cultural dimen-sions. The issues above show how the cul-tural orientations framework helps us devel-op research that accepts these cultural para-doxes and analyzes them within the scope ofthe overall culture. There are many otherissues that could be addressed using thesedata; however, these demonstrate a broadspectrum of applications of the framework.

It is important to at least mention some of the possibilities the cultural orientationsframework provides in terms of research onindividual behavior. First and foremost, itprovides individual-level ideas and scales ofindividualism, collectivism, hierarchy, mas-tery, subjugation, hierarchy and so forth.DiStefano and Maznevski (2000) and Lane etal. (2000) provide examples of how the orientations are related to work behavior.Some research has already used these scalessuccessfully in measuring cultural beliefs atthe individual level and using such measure-ment to predict organizational attitudes. Forexample, Kirkman and Shapiro (2001) studied the relationship between collectivismand resistance to teams. In a study presentedin Brockner et al. (2001), the researchers usedthe hierarchy measure as a way to measurepower distance at the individual level and


01_articles2(3) 10/9/02 10:36 am Page 289

better understand the relationship betweenparticipation and organizational commit-ment cross culturally. The variations withinthe orientations lend themselves to testingmany relationships only theorized about atpresent (Adler, 1997). For example, we cannow test the effect that a mastery versus aharmony or subjugation orientation has onbusiness planning and decision-making.Similarly, we can test the effect a manager’sbelief on the nature of humans has on management style. Yet another possiblestudy could look at the effect of an indi-vidual’s activity orientation on their prefer-ence for certain rewards. Such researchcould go a long way in beginning to test moti-vation theories cross culturally. The possibleresearch questions are infinite, with scalesthat measure cultural values at the individuallevel.

Limitations and Suggestionsfor Future Research

The contribution of this research is limited byseveral factors. First, although the Kluckhohnand Strodtbeck framework incorporates sixorientations with a combined total of 16 vari-ations, only four of the orientations (11 of thevariations) were measured here. The orienta-tions of Time and Space, which we did notmeasure, play a large role in cultural normsand values, as demonstrated by Hall’s work(Hall, 1966, 1973), so they clearly deservefuture research attention. Moreover, otherbasic dimensions not identified in Kluckhohnand Strodtbeck’s research may be found.Further research on cultural differences couldprovide more tests of the cultural orientationsframework itself, and elaborate on its dimen-sions.

Second, the measures demonstrated adequate psychometric properties for earlystages of research, especially given thebreadth and depth of the constructs beingstudied, but they are clearly not ideal. Thelarge amount of measurement error con-

tributes to a decrease in the power of statisti-cal comparisons, and our results may haveprovided sharper distinctions if we had hadbetter measures. Research is needed to develop these measures further. In addition,issues such as cross cultural equivalence ofmeasures must be addressed with greatersophistication in the field of internationalmanagement. For example, while Mullen(1995) and Singh (1995) assert that a con-struct is only equivalent in two samples if the same items load on to the same factor significantly, Cheung and Rensvold (1997)propose that different items will inevitablyload on to the same factor in different cul-tures, and that measures should not beassumed to be incomparable due to non-equivalence. The progress currently beingmade in methodology will contribute greatlyto the development of better measures of cultural dimensions.

Third, to obtain more valid measures theitems were contextualized by business situa-tions and the samples here were limited topeople involved in business. The dimensionstheoretically should apply more generally to the broader population, but culture isimpossible to articulate without a context.The importance of appropriate context,including matching the measures with thesample and context to be studied, providesanother possible explanation of the counter-intuitive results from the Netherlands. Thesample of undergraduate students, who onaverage had not worked at all in a full-timecapacity, may actually be representative ofNetherlands culture. However, the instrument

might not have provided statements thatrelated to a cultural context in which theyhad experienced Dutch cultural patterns,and thus it may not have provided a validmeasure of cultural orientations for them.Measures of the framework’s dimensionsshould be adapted for different contexts so the cultural orientations of other sub-populations can be measured and compared.

Finally, further research should incorpo-


01_articles2(3) 10/9/02 10:36 am Page 290

rate other variables of interest to cross cul-tural researchers, such as perceptions, beliefs,values, and behaviors. The fact that the cul-tural orientations and their variations can beconceptualized and measured at the indi-vidual level of analysis should provide morecomprehensive and conclusive studies ofcross cultural phenomena. As this study hasdemonstrated, the framework provides inter-esting new interpretations and raises newquestions regarding issues relevant to crosscultural research.

Notes The authors gratefully acknowledge fundingfrom the McIntire Foundation and Ivey BusinessSchool.

1 Note that, in Table 1 and throughout the restof this article, we have revised Kluckhohnand Strodtbeck’s original labels for thevariations in the Relationships and Activityorientations to make them more consistentwith typical cross cultural managementusage. More specifically, we have renamedtheir lineal as hierarchical, collateral as collective,being-in-becoming as being, and reflecting asthinking.

2 MANOVA of gender’s main effect resultedin F = 3.75, p = .000; in Canada, theNetherlands, and the US males scored lowerthan females on Environment–harmony; inCanada males scored higher than females onEnvironment-mastery, in the Netherlands malesscored lower than females on HumanNature–changeable; in the United States malesscored higher than females onActivity–thinking.

3 We conducted principal components andstructural equation modeling analyses withall 11 variations simultaneously, but theresults were not interpretable.

4 Within-person standardization is usuallycalculated with respect to the wholeinstrument (Leung and Bond, 1989).However, if these data are standardized withrespect to the instrument as a whole, thescores for one orientation affect the scores foranother, reducing the validity of cross-country comparisons at the orientation level.For example, the strength of an individual’sresponses to Environment statements will

affect the eventual standard scores of his orher Relationship statements, making theRelationship scores difficult to compare ontheir own with the Relationship scores fromothers. We therefore standardized the datawithin person and within orientation.

ReferencesAdler, N.J. (1997) International Dimensions of

Organizational Behavior, 3rd edn. Cincinnatti:South-Western.

Alduncin Abatia, E. (1993) Los Valores de losMexicanos: En Busca de una Esencia. Mexico:Fomento Cultural Banamex.

Bond, M.H. and Hwang, K.-K. (1986) ‘TheSocial Psychology of Chinese People’, inM.H. Bond (ed.) The Psychology of the ChinesePeople, pp. 213–66. Hong Kong: OxfordUniversity Press.

Brislin, R.W. (1980) ‘Translation and ContentAnalysis of Oral and Written Materials’, inH.C. Triandis and J.W. Berry (eds) Handbookof Cross-cultural Psychology, vol 2 (Methodology),pp. 389–444. Boston: Allyn & Bacon.

Brockner, J., Ackerman, G., Greenberg, J.,Gelfand, M.J., Francesco, A.M., Chen, Z.X,Leung, K., Bierbrauer, G., Gómez, C.,Kirkman, B.L. and Shapiro, D.L. (2001)‘Culture and Procedural Justice: TheModerating Influence of Power Distance onReactions to Voice’, Journal of ExperimentalSocial Psychology 37: 300–15.

Burns, T. and Stalker, G.M. (1961) TheManagement of Innovation. London: TavistockPublications.

Cheung, G.W. and Rensvold, R.B. (1997) ‘Cross-cultural Comparisons Using NonequivalentMeasurement Items’, paper presented atAcademy of Management Annual Meeting.

Chew, I.K.H. and Putti, J. (1995) ‘Relationshipof Work-related Values of Singaporean andJapanese Managers in Singapore’, HumanRelations 48(10): 1149–70.

Chinese Culture Connection (1987) ‘ChineseValues and the Search for Culture-freeDimensions of Culture’, Journal of Cross-Cultural Psychology 18(2): 143–64.

Cohen, S.G. and Bailey, D.E. (1997) ‘WhatMakes Teams Work: Group EffectivenessResearch from the Shop Floor to theExecutive Suite’, Journal of Management 23(3):239–90.

Condon, J.C. (1985) Good Neighbors. Yarmouth,ME: Intercultural Press.


01_articles2(3) 10/9/02 10:36 am Page 291

Cowling, M. (ed.) (1998) The Communist Manifesto:New Interpretations. Edinburgh: EdinburghUniversity Press.

Davis, S. (1969) ‘U.S. versus Latin America:Business and Culture’, Harvard Business Review88–98.

deForest, M.E. (1994) ‘Thinking of a Plant inMexico?’, Academy of Management Executive 8(1):33–40.

DiStefano, J.J. and Maznevski, M.L. (2000)‘Creating Value with Diverse Teams inGlobal Management’, Organizational Dynamics29(1): 45–63.

Erez, M. and Earley, P.C. (1993) Culture, Self-identity, and Work. Oxford: Oxford UniversityPress.

Hair, J.F. Jr, Anderson, R.E., Tatham, R.L. andBlack, W.C. (1992) Multivariate Data Analysis.New York: Macmillan Publishing Company.

Hall, E.T. (1966) The Hidden Dimension. GardenCity, NY: Doubleday.

Hall, E.T. (1973) The Silent Language. GardenCity, NY: Anchor Press.

Hall, E.T. and Hall, M.R. (1990) UnderstandingCultural Differences. Yarmouth, MA:Intercultural Press.

Hofstede, G. (1980) Culture’s Consequences:International Differences in Work-related Values.Newbury Park: Sage.

Hwang, K.-K. (1985) ‘Face and Favour: TheChinese Power Game’, American Journal ofSociology 92: 944–74.

Jöreskog, K.G. and Sörbom, D. (1993) LISREL8. Mooresville, IL: Scientific Software.

Kim, U., Triandis, H.C., Kagitcibasi, C., Choi,S.-C. and Yoon, G. (eds) (1994) Individualismand Collectivism: Theory, Method, and Applications.Thousand Oaks, CA: Sage Publications.

Kirkman, B. and Shapiro, D. (2001) ‘The Impactof Cultural Values on Job Satisfaction andOrganizational Commitment in Self-managing Work Teams: The Mediating Roleof Employee Resistance’, Academy ofManagement Journal 44: 557–69.

Kluckhohn, F. and Strodtbeck, F. (1961)Variations in Value Orientations. Evanston, IL:Row, Peterson.

Kras, E.S. (1989) Management in Two Cultures:Bridging the Gap between U.S. and MexicanManagers. Yarmouth, ME: Intercultural Press.

Lane, H.W., DiStefano, J.J. and Maznevski,M.L. (2000) International Management Behavior,4th edn. Cambridge, MA: Blackwell.

Leung, K. and Bond, M.H. (1989) ‘On theEmpirical Identification of Dimensions for

Cross-cultural Comparisons’, Journal of Cross-Cultural Psychology 20(2): 133–51.

Maznevski, M.L., DiStefano, J.J., Gomez, C.B.,Noorderhaven, N.G. and Wu, P. (1997) ‘TheCultural Orientations Framework andInternational Management Research’, paperpresented at Academy of InternationalBusiness Annual Meeting.

Maznevski, M.L. and Peterson, M.F. (1997)‘Societal Values, Social Interpretation, andMultinational Teams’, in C.S. Granrose andS. Oskamp (eds) Cross-cultural Work Groups, pp.61–89. Thousand Oaks, CA: Sage.

Mullen, M.R. (1995) ‘Diagnosing MeasurementEquivalence in Cross-national Research’,Journal of International Business Studies 26(3):573–96.

O’Grady, S. and Lane, H.W. (1996) ‘ThePsychic Distance Paradox’, Journal ofInternational Business Studies 27(2): 309–33.

Oh, T.K. (1991) ‘Understanding ManagerialValues and Behaviour among the Gang ofFour: South Korea, Taiwan, Singapore andHong Kong’, Journal of Management Development10(2): 46–56.

Osland, J. and Bird, A. (2000) ‘BeyondSophisticated Stereotyping: Cultural SenseMaking in Context’, Academy of ManagementExecutive 14(1).

Parsons, T. and Shils, E.A. (1951) Toward aGeneral Theory of Action. Cambridge, MA:Harvard University Press.

Ralston, D.A., Gustafson, D.J., Cheung, F.M.and Terpstra, R.H. (1993) ‘Differences inManagerial Values: A Study of U.S., HongKong and PRC Managers’, Journal ofInternational Business Studies 24(2): 249–75.

Ralston, D.A., Holt, D.H., Terpstra, R.H. andKai-Cheng, Y. (1997) ‘The Impact ofNational Culture and Economic Ideology onManagerial Work Values: A Study of theUnited States, Russia, Japan, and China’,Journal of International Business Studies 28(1):177–207.

Redding, S.G. (1993) The Spirit of ChineseCapitalism. New York: Walter de Gruyter &Co.

Rodríguez Estrada, M. and Ramírez-Buendia, P.(1992) Psicología del Mexicano en el Trabajo.Mexico: McGraw-Hill.

Rokeach, J. (1973) The Nature of Human Values.New York: Free Press.

Ronen, S. and Shenkar, O. (1985) ‘ClusteringCountries on Attitudinal Dimensions: AReview of and Synthesis’, Academy ofManagement Review 10(3): 435–54.


01_articles2(3) 10/9/02 10:36 am Page 292

Schwartz, S.H. (1992) ‘Universals in the Contentand Structure of Values: TheoreticalAdvances and Empirical Tests in 20Countries’, Advances in Experimental SocialPsychology 25: 1–65.

Schwartz, S.H. (1994) ‘BeyondIndividualism/Collectivism: New CulturalDimensions of Values’, in U. Kim, H.C.Triandis, C. Kagitcibasi, S.-C. Choi and G.Yoon (eds) Individualism and Collectivism: Theory,Method, and Applications, pp. 85–119. ThousandOaks, CA: Sage Publications.

Schwartz, S.H. and Bilsky, W. (1990) ‘Toward aTheory of the Universal Content andStructure of Values: Extensions and Cross-cultural Replications’, Journal of Personality andSocial Psychology 58(5): 878–91.

Schwartz, S.H. and Sagiv, L. (1995) ‘IdentifyingCulture-specifics in the Content and Structureof Values’, Journal of Cross-Cultural Psychology26(1): 92–116.

Singh, J. (1995) ‘Measurement Issues in Cross-national Research’, Journal of InternationalBusiness Studies 26(3): 597–619.

Triandis, H.C. (1972) Analysis of Subjective Culture.New York: Wiley.

Triandis, H.C. (1988) ‘Collectivism v.Individualism: A Reconceptualisation of aBasic Concept in Cross-cultural SocialPsychology’, in G.K. Verma and C. Bagley(eds) Cross-cultural Studies of Personality, Attitudes,and Cognition. New York: St Martin’s Press.

Triandis, H.C., Chan, D.K.-S., Bhawuk, D.P.S.,Iwao, S. and Sinha, J.B.P. (1995)‘Multimethod Probes of Allocentrism andIdiocentrism’, International Journal of Psychology30(4): 461–80.

Trompenaars, F. (1993) Riding the Waves of Culture:Understanding Cultural Diversity in Business. Avon:The Bath Press.

Yau, O.H.M. (1988) ‘Chinese Cultural Values:Their Dimensions and MarketingImplications’, European Journal of Marketing22(5): 44–57.

Zurcher, L.A., Meadow, A. and Zurcher, S.(1965) ‘Value Orientation, Role Conflict andAlienation from Work: A Cross-culturalStudy’, American Sociological Review 30: 539–48.


Appendix 1 CPQ4 Sample Items

No. of Scale items Sample items

Relationships

Individualism 7 People tend to think of themselves first, before they think of others.Society works best when each person serves his or her own interests.

Collectivism 8 Good team members subordinate their own goals and thoughts to those of theteam.Society works best when people willingly make sacrifices for the good of everyone.

Hierarchical 7 A hierarchy of authority is the best form of organization.People at lower levels in organizations should carry out the requests of peopleat higher levels without question.

Environment

Subjugation 7 People should not try to change the paths their lives are destined to take.It’s best to leave problem situations alone to see if they work out on their own.

Mastery 7 With enough knowledge and resources, any poor-performing business can beturned around.Good performance comes from taking control of one’s business.

Harmony 7 All living things are equal and deserve the same care and consideration.It is our responsibility to conserve the balance of elements in our environment.

continues

01_articles2(3) 10/9/02 10:36 am Page 293

MARTHA L. MAZNEVSKI is at IMD –International Institute for ManagementDevelopment, Chemin de Bellerive 23, PO Box915, CH-1001 Lausanne, Switzerland.[email: [email protected]]

JOSEPH J. DISTEFANO is at IMD –International Institute for ManagementDevelopment, Chemin de Bellerive 23, PO Box915, CH-1001 Lausanne, Switzerland.[email: [email protected]]

CAROLINA B. GOMEZ is in the Departmentof Management and International Business,College of Business Administration, Florida

International University, University Park, Miami,FL 33199, USA.[email: [email protected]]

NIELS G. NOORDERHAVEN is in the Facultyof Economics and Institute for Research onIntercultural Cooperation, Tilburg University,PO Box 90153, 5000 LE Tilburg, TheNetherlands.[email: [email protected]]

PEI-CHUAN WU is in the NUS BusinessSchool, National University of Singapore,Business Link, Republic of Singapore 117591.[email: [email protected]]


Appendix 1 CPQ4 Sample Items (cont.)

No. of Scale items Sample items

Activity

Doing 10 It’s human nature to place more importance on work than on other activities.Accomplishing a great deal of work is more rewarding than spending time inleisure.

Being 7 People should take time to enjoy all aspects of life, even if it means not gettingwork done.You shouldn’t worry about working when you don’t feel like it.

Thinking 8 Even if it takes more time, business decisions should always be made based onanalysis, not intuition.The outcomes of a business decision can be predicted accurately by a logicalanalysis of that decision.

Human Nature

Good/Evil 6 If supervisors don’t always check when workers come and go, workers willprobably lie about how many hours they work.You should be suspicious of everybody.

Changeable 5 Anyone’s basic nature can change.If someone is essentially a good person now, she or he will probably always begood. (Reverse Scored)

01_articles2(3) 10/9/02 10:36 am Page 294


01_articles2(3) 10/9/02 10:36 am Page 295

01_articles2(3) 10/9/02 10:36 am Page 296

CPQ CulturalDimensions

Documents

Transcript of CPQ CulturalDimensions