The Relation Between Culture and Response StylesEvidence From 19 Countries

Abstract

The authors investigated at the country level the effects of four cultural orientations identified and studied by Hofstede on two commonly recognized response biases: extreme response style and acquiescent responding. Data are presented from approximately 18,000 survey questionnaires completed by employees in 19 nations on five continents (Australia, Belgium, Brazil, Czech Republic, Germany, Hungary, India, Japan, Malaysia, Portugal, Turkey, the United Kingdom, Mexico, the Philippines, Poland, Singapore, Hong Kong, France, and Italy). Hierarchical linear modeling was employed to examine the associations between person-level response styles and country-level cultural orientations. Consistent with theoretical expectations, power distance and masculinity were found to be positively and independently associated with extreme response style. Individualism, uncertainty avoidance, power distance, and masculinity were each found to be negatively associated with acquiescent response behavior. Further research is needed to identify how question characteristics might interact with cultural orientations to influence response behavior.

Full text

10.1177/0022022104272905JOURNAL OF CROSS-CULTURAL PSYCHOLOGYJohnson et al. / CULTURE AND RESPONSE STYLES
THE RELATION BETWEEN CULTURE AND RESPONSE STYLES
Evidence From 19 Countries
TIMOTHY JOHNSON
University of Illinois at Chicago
PATRICK KULESA
ISR LLC
YOUNG IK CHO
University of Illinois at Chicago
SHARON SHAVITT
University of Illinois at Urbana–Champaign
The authors investigatedat the country level the effects of four cultural orientations identified and studied by
Hofstede on two commonly recognized response biases: extreme response style and acquiescent respond-
ing. Data are presented from approximately 18,000 survey questionnaires completed by employees in 19
nations on five continents (Australia, Belgium, Brazil, Czech Republic, Germany, Hungary, India, Japan,
Malaysia, Portugal, Turkey, the United Kingdom, Mexico, the Philippines, Poland,Singapore,Hong Kong,
France, and Italy). Hierarchical linear modeling was employed to examine the associations between person-
level response styles and country-level cultural orientations. Consistent with theoretical expectations,power
distance and masculinity were found to be positively and independently associated with extreme response
style. Individualism, uncertainty avoidance, power distance, and masculinity were each found to be nega-
tively associated with acquiescent response behavior. Further research is needed to identify how question
characteristics might interact with cultural orientations to influence response behavior.
Keywords: acquiescence; extreme response style; culture-level; cross-cultural; method bias
Shortly after World War II, social researchers began to investigate a variety of potential
sources of measurement error in research instrumentation. Among these were systematic
variations in the styles used by respondents to answer survey questions. Two of these forms
of response style are extreme responding, or extreme response style (Cronbach, 1946), and
acquiescent, or yea-saying, behavior (Couch & Keniston, 1960). Although these mea-
surement artifacts have been recognized and studied now for more than 50 years, surpris-
ingly little is known with regard to their cultural origins. In this article, using diverse samples
of adults in 19 countries, we investigate the associations between several dimensions of
national-level culture and individual variability in acquiescent and extreme response styles.
Of specific interest in our comparisons are Hofstede’s (2001) original four cultural di-
mensions of power distance, uncertainty avoidance, individualism-collectivism, and
masculinity-femininity.
Extreme response style refers to a greater tendency of some respondents to select the end-
points of a response scale when answering questions. The implications of this measurement
artifact for cross-cultural research are serious, as group variations in extreme response pat-
264
AUTHORS’NOTE: An early version of this article was presented at the annual meeting of the Society for Industrial and Organiza-
tional Psychology, Orlando, Florida, April 11, 2003. The authors would like to gratefully acknowledge International Survey
Research, LLC, for making the data available for this article. Thank you also to Richard Campbell for his statistical advice.
JOURNAL OF CROSS-CULTURAL PSYCHOLOGY, Vol. 36 No. 2, March 2005 264-277
DOI: 10.1177/0022022104272905
© 2005 Sag e Pub licat ions

terns, when left unevaluated, may be misinterpreted as substantive differences in the con-
struct(s) being examined. Indeed, this concern is justified, as differences in extreme respond-
ing have been found between a variety of national and ethnic groups, including Caucasians,
African Americans, and Latinos in the United States (Bachman & O’Malley, 1984; Clarke,
2000; Hui & Triandis, 1989; Warnecke et al., 1997), between U.S. and Korean respondents
(Chun & Campbell, 1974; Lee & Green, 1991), Japanese and Americans (Stening & Everett,
1984; Zax & Takahashi, 1967), members of Northern and Southern European nations (van
Herk, Poortinga, & Verhallen, 2004), and between French and Australian samples (Clarke,
2000). The cultural qualities responsible for this variability, however, remain unknown.
Some have hypothesized that they reflect differences in emphasis on sincerity versus mod-
esty in social interactions (Marín & Marín, 1989).
A second response style of present interest is acquiescence bias, also known as agreement
bias. This is the tendency of some respondents to agree with questions, regardless of ques-
tion content. In the United States, differences have been found in acquiescent reporting
behavior among Caucasian, White, and Latino populations (Aday, Chiu, & Anderson, 1980;
Carr, 1971; Johnson et al., 1997; Marín & Marín, 1989), as well as in several cross-national
comparisons (Cunningham, Cunningham, & Green, 1977; Grimm & Church, 1999; Morris
& Pavett, 1992; Riordan & Vandenberg, 1994; Ross & Mirowsky, 1984; van Herk et al.,
2004). Similar to extreme responding, acquiescence bias, when uncontrolled, has the poten-
tial to introduce serious confounds that may threaten the validity of cross-cultural compari-
sons (Cheung & Rensvold, 2000). Welkenhuysen-Gybels, Billiet, and Cambre (2003), for
example, have documented the importance of controllingfor acquiescence bias when evalu-
ating construct equivalence across multiple groups. Cultural variability in acquiescent
behavior is often linked to the norm of agreeableness, which is known to vary significantly in
importance across cultures and nations (Hui & Triandis, 1983; Javeline, 1999).
Further emphasizing the importance of investigating measurement artifacts such as these
are findings of the temporal stability of acquiescence and extreme responding behaviors
across ethnic subgroups in the United States (Bachman & O’Malley, 1984). It is interesting
that all empirical comparisons of these two response style patterns, at the individual level,
have involved comparisons between no more than two or three national or ethnic subgroup
samples. Hence, our ability to generalize from this evidence remains limited. More impor-
tant, as mentioned earlier, the dimensions of respondent culture that drive these group dif-
ferences remain largely unexplored and unknown. Several recent international studies, for
example, have examined both extreme response styles and acquiescence but did not exam-
ine the potential effects of variability in cultural orientations on these response styles.
Baumgartner and Steenkamp (2001) explored response styles across 11 European Union
nations but did not address potential linkages between dimensions of culture and response
style. Several multinational surveys reported by van Herk (2000) also examined these
response styles in a number of European countries. Although cultural dimensions were not
examined, the author acknowledged the need for further research on this topic.
We are aware of three studies that have examined the association between response style
and cultural orientation. Chen, Shin-ying, and Stevenson (1995), employing an ad hoc mea-
sure of individualism in a study of high school students in Japan, Taiwan, Canada, and the
United States, reported an association between self-reported individualism and extreme
response behavior at the individual level within each country. In addition, recent research by
Smith (2004) explored evidence from multiple sources concerned with the cultural cor-
relates of acquiescent response styles. He found that national-level collectivism, power
distance, and uncertainty avoidance predicted acquiescence bias in response to personally
Johnson et al. / CULTURE AND RESPONSE STYLES 265

relevant survey items. van Hemert, van de Vijver, Poortinga, and Georgas (2002) also inves-
tigated national-level correlates of the Eysenck Lie Scale, which is believed to reflect confor-
mity and, hence, acquiescence. They found individualism to be negatively associated and
power distance to be positively associated with this measure. These studies highlight the im-
portant role that cultural orientation may play in understanding response styles and suggest
promising directions for additional research.
THE RELATION BETWEEN
CULTURAL DIMENSIONS AND RESPONSE STYLES
Despite extensive work on the cultural dimensions that Hofstede (2001) originally pro-
posed, there is little theoretical guidance concerning the form that associations between
these cultural dimensions and response styles might take. We nonetheless offer tentative
hypotheses concerning potential relationships between culture and each form of response
style. These hypotheses are based on the following assumptions about culturally relevant
norms and motives associated with the response styles under investigation. First, we argue
that an extreme response style is in line with a motivation to achieve clarity, precision, and
decisiveness in one’s explicit verbal statements, whereas a middling response style is in line
with norms for ambiguity, flexibility, and modesty in one’s verbal statements. Second, we
suggest that acquiescence is a submissive response style that conveys agreeableness and def-
erence to hierarchy, especially in contexts in which interpersonal or group harmony is impor-
tant (Javeline, 1999). Such response behavior has been documented in several cultural
groups. Among Latinos, this pattern of social interaction is known as simpatìa (Triandis,
Marín, Lisansky, & Betancourt, 1984). A similar pattern exists among East Asian popula-
tions, where it has been referred to as a “courtesy bias” (Deutscher, 1973; Niikura, 1999).
Cultures high in power distance tend to be more authoritarian societies where conformity
is stressed and submissiveness is common (Hofstede, 2001). One manner in which confor-
mity might be expressed is via deferential, or acquiescent, behavior. National-level analyses
of data reported by Gordon (1976; cited in Hofstede, 2001, p. 509) have found a positive
independent effect of power distance on measures of conformity across 17 countries, sug-
gesting a linkage between this cultural dimension and deference. Hofstede (2001, p. 96) has
also interpreted research by Williams, Satterwhite, and Saiz (1998) as evidence of greater
acquiescent behavior among persons in high power distance cultures. Several national-level
measures of acquiescence have additionally been found to be positively correlated with
Hofstede’s measure of power distance (Smith, 2004; van Hemert et al., 2002). We thus
expect to find that persons in societies that are high in power distance will be more likely to
exhibit acquiescent response behavior when completing questionnaires. The large power
differentials of a high power distance culture may also demand or foster greater decisiveness
and definitiveness in communications, which we hypothesize will promote extreme
response style behavior. Conversely, persons in low power distance cultures, which are simi-
lar to horizontal cultures in their emphasis on equality in status (Chen et al., 2001; Triandis &
Gelfand, 1998), may be more likely to emphasize modesty as a value (Nelson & Shavitt,
2002). Thus, low power distance may be associated with a more middling response style.
Persons embedded in masculine cultures may also be more likely to endorse extreme
responses on questionnaires. Perhaps some of the better known features of masculine cul-
tures are emphases on assertiveness and on decisive and daring behavior (Hofstede, 1998,
2001, p. 298). These qualities may encourage respondents within such cultures to select the
266 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

strongest available choices for representing their opinions. In contrast, more feminine cul-
tures emphasize modesty (Hofstede, 1998), which may be reflected at the individual level by
personal preferences for more middling and less extreme response styles. These same quali-
ties suggest that persons in masculine cultures may be less likely to display acquiescent
response behaviors. Supporting evidence comes from a secondary analysis of data first re-
ported by Bass and Burger (1979), in which Hofstede (2001) found national-level measures
of assertiveness to be independently associated with masculinity across 12 nations (p. 267).
However, recent work by van Hemert et al. (2002) and Smith (2004) both reported the
absence of a relationship between national-level measures of Hofstede’s femininity dimen-
sion and several direct indicators of acquiescence. It is thus with some caution that we
hypothesize that extreme response styles will be more common, and acquiescence less
common, among persons in masculine cultures.
Individualistic cultural orientations may also be associated with these response artifacts.
Persons from nations with individualistic cultures seek to achieve clarity in their explicit ver-
bal statements (Hall, 1976; Triandis, 1995) because they are less concerned with the conse-
quences of expressing strong opinions. Therefore, extreme response styles may be more
common among persons from individualist countries. Conversely, collectivism is associated
with a greater emphasis on interpersonal harmony and with less emphasis on individual
opinions (Chen et al., 2001; Hofstede, 2001). Ambiguity in communication is adaptive in
these cultural contexts. Thus, a middling response style should better fit the cultural norms
and imperatives of persons living in collectivist cultures. As mentioned earlier, Chen et al.
(1995) have reported findings consistent with an association between individualism, mea-
sured at the person level, and extreme response behavior. For the same reasons, response
acquiescence may be less common among persons in individualistic cultures because main-
taining harmony and conveying agreeableness and deference are less emphasized in these
cultural contexts. Persons in collectivist cultures, in contrast, may be “more sensitive to the
social pressures emanating from the questionnaire” (Hofstede, 2001, p. 218). A cross-
cultural meta-analysis of Asch’s (1956) classic conformity experiments produced evidence
consistent with the expectation that conformity is less common in individualistic societies
(Bond & Smith, 1996). In addition, many studies that have compared national and sub-
national groups (Aday et al., 1980; Grimm & Church, 1999; Johnson et al., 1997; Marín,
Gamba, & Marín, 1992; Ross & Mirowsky, 1984; van Herk et al., 2004) have reported evi-
dence supportive of the hypothesis that acquiescent response behaviors may be more com-
mon in collectivistic societies. National-level correlations between cultural orientations
and acquiescence reported by Smith (2004), van Hemert et al. (2002), and Hofstede (2001,
p. 509) also support this hypothesis. Consequently, we anticipate that persons within indi-
vidualistic cultures will be less likely to exhibit acquiescent response behavior and more
likely to demonstrate extreme response behavior.
According to Hofstede (2001), societies higher in uncertainty avoidance have many rules
and have little tolerance for ambiguity. Research has suggested that individuals’ extreme
responding is a reflection of intolerance of ambiguity (Hamilton, 1968). The endpoints of a
measurement scale may often be interpreted by respondents as being more definitive and
clear than are scale midpoints, which are more likely to be subject to qualifications and mul-
tiple interpretations by respondents. We thus hypothesize that extreme responding will be
more common in cultures that emphasize uncertainty avoidance.
The nature of the relation between uncertainty avoidance and acquiescent response style
is somewhat less clear. On one hand, one might argue that providing acquiescent responses
Johnson et al. / CULTURE AND RESPONSE STYLES 267

across multiple items in a questionnaire is inherently ambiguous, as a respondent’s true opin-
ions or feelings cannot be discerned from such data. Persons embedded within cultural
frameworks less tolerant of ambiguity may thus have less inclination to provide ambiguous,
or acquiescent, responses. On the other hand, for any given item, if there is any uncertainty
about how to respond, agreeing with the item offers a less ambiguous resolution than does a
middling (“unsure”) response. That is, similar to the reasoning concerning the link between
uncertainty avoidance and extreme response style, acquiescent behavior may be a mecha-
nism for enhancing the certainty of one’s responses to individual items. If that is the case,
acquiescent response behavior may be more common within cultures that are high in uncer-
tainty avoidance. Consistent with this interpretation, Smith’s (2004) research suggests a pos-
itive association between uncertainty avoidance and acquiescent response behavior. These
findings, however, were based on measures from the Global Leadership and Organizational
Behavior Effectiveness (GLOBE) project (House, Hanges, Javidan, Dorfman, & Gupta,
2004), rather than the Hofstede measure of uncertainty avoidance, which was not signifi-
cantly associated with acquiescence in that study. Hofstede’s uncertainty avoidance measure
was also unassociated with the measure of acquiescence employed by van Hemert et al.
(2002). Our data afford an opportunity to further explore the nature of this and other
relationships between cultural dimensions and response behaviors.
METHOD
DATA SOURCE
The data to be analyzed were originally collected as part of employee surveys conducted
by ISR LLC, between 1992 and 2002. A total of 20,270 self-administered surveys was avail-
able for analysis. A total of 19 countries on five continents is represented in these data.
Europe is somewhat overrepresented, with data contributed from 9 countries. Seven Asian
countries are additionally represented, along with 1 country each from North and South
America, plus Australia. The specific countries included are as follows: Australia, Belgium,
Brazil, the Czech Republic, France, Germany, Hong Kong, Hungary, India, Italy, Japan,
Malaysia, Mexico, the Philippines, Poland, Portugal, Singapore, Turkey, and the United
Kingdom. The number of respondents within each of these countries is reported in Table 1.
All respondents completed a common core set of questions using 5-point Likert-type
response scales. As these were employee surveys, the questionnaire content was largely
restricted to questions concerned with employee satisfaction and work environment (the
specific wording of sample questions examined is available from the authors). Although the
precise number of questions varied somewhat across countries, the average number of items
included in each questionnaire was approximately 120. Most items were closed-ended and
employed 5-point Likert-type response formats. The predominant response format was the
following: disagree,tend to disagree,unsure,tend to agree,agree. Two less commonly
employed response formats included (a) very poor,poor,adequate,good,very good, and (b)
very dissatisfied,dissatisfied,neither,satisfied,very satisfied. Only items using the disagree-
to-agree format were included in these analyses. Questions were worded in a similar manner
and direction across all countries.1Additional information was available with regard to sev-
eral person-level characteristics, including age, gender, and length of employment.
268 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

269
TABLE 1
Country-Level Values for Hofstede’s Four Dimensions and Per Capita Gross National Product (GNP)
Country Total N= 20,270 Power Distance Uncertainty Avoidance Individualism Masculinity GNP Per Capitaa
Australia 459 low medium high medium medium
Belgiumb995 medium high high low high
Brazil 1,092 high medium medium low medium
Czech Republicb749 medium medium medium medium medium
France 1,028 medium high high low high
Germanyb1,751 low medium medium high high
Hong Kongb502 medium low low medium high
Hungaryb750 low medium high high medium
India 1,525 high low medium medium low
Italy 1,054 low medium high high medium
Japanb726 low high medium high high
Malaysia 500 high low low low low
Mexico 1,420 high medium low high low
Philippines 508 high low low medium low
Polandb788 medium high medium medium low
Portugalb591 medium high low low medium
Singaporeb519 high low low low high
Turkeyb1,000 medium high medium low low
United Kingdom 4,313 low low high high medium
a. Based on the World Bank estimates of 1998 GNP per capita published at http://www.worldbank.org/depweb/english/modules/economic/gnp/datanot.html
b. Nations included in the analysis of acquiescence.

MEASURES
Country-level scores for power distance, uncertainty avoidance, individualism-collectivism,
and masculinity-femininity were abstracted from Hofstede’s (2001) results and appended to
the data set. As there was a wide distribution of values for each dimension (approximate
range = 0 to 100), and a relatively small number of countries available for analysis, these
scores were trichotomized into high-medium-low ordinal values for purposes of analysis.
This rescaling had the effect of truncating the outliers within the distributions of each mea-
sure, a conservative process that might contribute to attenuated relationships. It also assured
that the number of cases exceeded the distribution of each measure. The values assigned to
each country for each dimension are presented in Table 1. These values served as the primary
independent variables for these analyses.
Measures of extreme response style and acquiescence were constructed from the ques-
tions available in the core survey instrument. These were employed as the dependent vari-
ables of interest. To develop an index of extreme response style, 61 questions sharing a simi-
lar set of response options (e.g., ranging from 1 = disagree to 5 = agree) were recoded such
that selection of either endpoint received a code of 1 and the middle values received a code of
0. These items were then summed to form an extreme response style measure with a hypoth-
esized range of 0 to 61. Higher values on this index were reflective of more extreme respond-
ing. The psychometric qualities of this measure were excellent. The pooled alpha reliability
coefficient was .96. Within-country alpha coefficients ranged from .92 to .96.
A measure of acquiescence was developed by selecting a separate set of 18 items, half of
which were worded in a positive direction and half of which were worded negatively.
Responses to these items were available only for 10 of the 19 nations included in the data file.
Hence, all analyses of acquiescence are restricted to this subset of 10 nations, which are iden-
tified in Table 1. The 18 items included 9 pairs of items that dealt with similar content but
were worded in opposite directions. For example, pairs of questions included the following:
“For the work I do, I am fairly paid,” and “For the work I do, I am very much underpaid.” A
second pair of questions included, “I am optimistic about the future of my company,” and “I
am frequently worried about the future of my company.” Using the coding scheme devel-
oped by Winkler, Kanouse, and Ware (1982), values of 1 were assigned to each question pair
for which respondents agreed or tended to agree with both items; otherwise, a value of 0 was
assigned. These items were then summed to create an index that ranged from 0 to 9, with
higher values representing greater levels of acquiescence. The pooled alpha reliability coef-
ficient for this measure was .43 and ranged within countries from .36 to .55.
The correlation between our measures of acquiescence and extreme response style was
low (r= .05), suggesting that these response style indicators are largely independent of one
another. The acquiescence measure, however, was strongly correlated with Hofstede’s
acquiescence indicator (see Hofstede, 2001, p. 484) when averaged at the national level (r=
.96) for the seven nations included in both data sets. Our acquiescence measure was also
associated, but less strongly (r= .37, n= 9 countries), with the national-level measure of
acquiescence employed by Smith (2004). These findings suggest that the measure has good
construct validity.
Other individual-level measures included in these analyses were gender, age, and length
of employment. Each was used as a control variable to adjust for potential variations in sam-
ple composition across countries. Some evidence is available to suggest that age and gender,
in particular, may be associated with these response artifacts (Hamilton, 1968; Johnson et al.,
270 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

1997). Summary information for all variables measured at the person level is supplied in
Table 2.
ANALYSES
Hierarchical linear models (Raudenbush, Bryk, Cheong, & Congdon, 2000) were esti-
mated to examine the independent effects of national-level predictors of Hofstede’s (2001)
four cultural dimensions on extreme response style and acquiescent response behavior, con-
trolling for the effects of the person-level predictors and taking into account the random vari-
ations across nations. Person-level predictors included in each model were gender, age, and
length of current employment. An additional covariate, measured at the country level, was
also included in these models: gross national product (GNP) per capita. This variable was
added to adjust for known associations between national affluence and cultural orientations
(Hofstede, 2001; Triandis, 1995; van Hemert et al., 2002). As with our other country-level
indicators, per capita GNP was trichotomized for these analyses. Values assigned to each
country are also presented in Table 1.
RESULTS
Table 3 presents hierarchical linear models that examine the predictors of extreme
responding. As presented in the upper panel of the table (i.e., fixed effect model), two of four
country-level indicators of cultural orientations were found to be independently associated
with extreme response style. Consistent with two of our hypotheses, power distance and
masculinity were each positively related to extreme responding behavior. However, uncer-
tainty avoidance and individualism were not independently associated with extreme re-
sponse style. Per capita GNP was also not associated with extreme responding. None of the
Johnson et al. / CULTURE AND RESPONSE STYLES 271
TABLE 2
Summary Information for Individual-Level Variables
Extreme Response Acquiescence
Variable N M (SD) N M (SD)
Total 18,307 24.87 (16.19) 8,079 5.28 (1.80)
Gender
Male 10,681 25.37 (16.61) 4,742 5.33 (1.80)
Female 7,626 24.17 (15.55) 3,337 5.20 (1.81)
Age
Younger than 25 3,044 26.59 (16.04) 1,006 5.19 (1.87)
25 to 34 years 5,850 24.79 (16.10) 2,482 5.18 (1.76)
35 to 44 years 5,044 24.52 (16.61) 2,286 5.35 (1.79)
45 or older 4,370 24.16 (15.84) 2,307 5.34 (1.82)
Tenure/length of current employment
Less than 2 years 4,337 25.59 (16.40) 1,563 5.12 (1.88)
2 to less than 5 years 4,029 24.97 (16.24) 1,754 5.25 (1.73)
5 to less than 10 years 3,647 24.20 (16.05) 1,682 5.23 (1.81)
10 to less than 20 years 2,318 24.49 (16.22) 1,807 5.37 (1.77)
20 or more years 2,401 25.00 (15.79) 1,266 5.43 (1.81)

person-level background characteristics were found to be predictive of extreme response
behavior. However, random variation across nations in the effects of each of these variables
remained significant, indicating that the effects of gender, age, and employment tenure were
varied across nations. Gender differences in extreme response behavior, for example, varied
significantly across nations, as did age and employment differences.
A second model that examined the effects of cultural dimensions on acquiescence is also
presented in Table 3. As discussed earlier, the acquiescence measure was only available for a
subset of the nations included in this data file. Consequently, this model is based on answers
from 8,062 respondents in 10 countries only. All four of the Hofstede dimensions were
found to be independently associated with this response style. Consistent with our hypothe-
ses, persons from individualistic countries were less likely to engage in acquiescent respond-
ing. In addition, persons in nations rated highly on uncertainty avoidance were less likely to
provide acquiescent responses. Respondents embedded in more masculine societies were
also less likely to respond in an acquiescent manner. Contrary to our expectations, persons in
high power distance countries were less likely to provide acquiescent responses. Length of
current employment was also found to be positively associated with acquiescent response
behavior. Background characteristics did not have significant fixed effects on acquiescent
response behavior, but the effects of gender and length of employment were found to be sig-
nificantly varied across countries. National GNP was also found to be negatively associated
with acquiescence. That is, the lower the affluence of the country, the more individuals in the
272 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY
TABLE 3
Hierarchical Linear Models Evaluating Effects of Individual- and
National-Level Measures on Extreme and Acquiescent Responding
Extreme Response Acquiescence
N= 18,308 (19 nations) N= 8,062 (10 nations)
Fixed effect Coefficient (SE) t-ratio Coefficient (SE) t-ratio
Intercept –10.06 (15.62) –.66 9.85 (1.01) 9.74***
Person-level predictors
Gender (male) –.26 (.28) –.95 .08 (.06) 1.31
Age .14 (.17) .81 .04 (.03) 1.43
Tenure (length of current employment) .31 (.18) 1.68 .09 (.04) 2.50*
Country-level predictors
Power distance 9.63 (2.71) 3.55** –.94 (.24) –3.92*
Uncertainty avoidance .59 (1.18) .50 –.32 (.11) –2.88*
Individualism 3.41 (2.40) 1.42 –.35 (.09) –3.99*
Masculinity 5.80 (1.73) 3.36** –.62 (.16) –3.92*
GNP per capita –2.30 (1.84) –1.55 –.30 (.08) –4.03*
Variance Variance
Random Effect Component (df) 2Component (df) 2
Person-level variance 179.09 3.10
Country-level (between countries)
Average level 41.52 (13) 169.72*** .08 (4) 19.60**
Effect of gender .74 (18) 35.32** .02 (9) 20.99*
Effect of age .36 (18) 35.22** .00 (9) 7.82
Effect of tenure .50 (18) 85.46*** .01 (9) 35.77***
*p< .05. ** p< .01. *** p< .001.

country tended to manifest an acquiescent response style. This is consistent with previous
findings reported by van Hemert et al. (2002).2
DISCUSSION
These findings expand our knowledge of the relation between culture and response
biases. In particular, evidence is provided that two commonly observed response styles
covary in a systematic manner with established dimensions of national culture across a broad
set of nations. The associations between culture and extreme responding behavior were con-
sistent with an expanded view of extreme response style behavior that recognizes the effects
of cultural dimensions on the response behaviors of individuals. Specifically, we suggested
that an extreme response style serves the goals of achieving clarity, precision, and decisive-
ness in one’s explicit verbal statements, characteristics that are valued in masculine and high
power distance cultures. Consistent with this, results also indicated that persons in cultures
with high masculinity and with high power distance were more likely to select extreme re-
sponse options on a questionnaire.
Acquiescent response style was found to have associations with each of Hofstede’s cul-
tural orientations. In particular, respondents from more individualistic nations were less
likely to provide acquiescent answers, a finding supportive of Hofstede’s (2001) observation
that conformity is less common in highly individualistic societies. Acquiescence was also
greater among respondents within less uncertainty avoidant countries. This finding is less
clear. As was discussed earlier, it is not immediately obvious whether one would expect to
find a positive or negative association between acquiescence and uncertainty avoidance. Our
analyses, however, indicated an inverse relationship between national-level uncertainty
avoidance and individual-level acquiescent behavior, a finding consistent with the expecta-
tion that acquiescence is a response style less common within cultures that reject ambiguity
and uncertainty. Also consistent with our hypothesis, acquiescent reporting behavior was
lower among individuals in more masculine cultures. Because countries high in masculinity
value decisive, assertive, and daring action (Hofstede, 1998, 2001), it would appear that
acquiescence is a communication pattern less likely to be favored or acceptable.
Of additional interest is the negative association between national power distance and
acquiescence among individual respondents, a finding that is contrary to our original hy-
pothesis. It is notable that, although acquiescence is an individual-level behavior, the avail-
able evidence suggesting a positive association with power distance is based on national-
level ecological correlations (Hofstede, 2001, p. 96; Smith, 2004; van Hemert et al., 2002).
Indeed, the direction of this correlation was confirmed when we examined the correlation
between power distance and mean acquiescence scores among the 10 countries for which we
have acquiescence data. That correlation was r= .21, a finding consistent with these previous
studies. Our data thus suggest that power distance and acquiescence are positively correlated
at the national level but negatively associated with one another in cross-level (i.e., HLM)
analyses where acquiescence is examined as an individual behavior. We believe that, for the
purposes of this study, the cross-level analyses provide a more appropriate representation of
these data as each variable is examined at the level of analysis for which we wish to make
inferences. That is, acquiescence is examined as an individual-level behavior that is influ-
enced by national-level characteristics, which are represented by the Hofstede dimensions.
Our contradictory findings from these various levels of analysis serve as a reminder that
Johnson et al. / CULTURE AND RESPONSE STYLES 273

caution should be exercised when examining individual behavior with aggregated data
(Robinson, 1950).
Another important direction for further research will be to examine the degree to which
potential relationships between cultural orientations and response behaviors may be modi-
fied by question characteristics. Certain question-response formats, for example, might
amplify or suppress some of the hypothesized relationships discussed in this article. The
relationship between uncertainty avoidance and extreme responding behavior serves as a
case in point. Providing labels for the endpoints but not the midpoints of a response scale, a
common practice in many countries, may be more problematic for respondents with an intol-
erance for ambiguity and may encourage them to select extreme response options, which are
more clearly labeled. The effects of uncertainty avoidance on respondent reporting behavior
may thus be stronger for some question forms than for others and might possibly account for
the null findings observed here with regard to this hypothesized relationship. Recent
research by Wong, Rindfleish, and Burroughs (2003) has also identified cultural variability
in the applicability of reverse-worded Likert-type questions, which they find to be problem-
atic when administered to East Asian, but not Western, populations. Ironically, the use of
reverse-worded items is a common procedure for reducing acquiescence bias (Watson,
1992; Wong et al., 2003). Consequently, further research designed to more systematically
evaluate how culture interacts with question design features seems essential.
Although not a focus of this research, we note that the negative association in this study
between GNP and acquiescence is consistent with the findings of van Hemert et al. (2002),
who reported a negative correlation of –.67 (p< .01) between GNP and the Eysenck Lie
Scale. As discussed earlier, van Hemert’s analysis was based on national-level measures of
acquiescence, whereas our study reflects a multilevel correlation between an individual-
level measure of acquiescence and a national-level GNP measure. The consistent findings
across these two studies that reflect differing levels of analysis suggest that further research
concerned with the relationship between culture and acquiescence will need to account for
national wealth.
There are several limitations to be considered in evaluating this research. First, the mea-
sures of acquiescent and extreme response behaviors were developed post hoc from survey
questions originally collected for other purposes, primarily to evaluate employee satisfac-
tion and attitudes. Although these measures have good psychometric properties, they are
nonetheless unique to this data set and the generalizability of our findings cannot be
assumed. The national-level measures of Hofstede’s (2001) four important cultural dimen-
sions may also be questioned, as they were initially developed on the basis of survey data col-
lected more than 30 years ago. Although Hofstede’s analyses were careful only to retain
dimensions that he felt could be replicated, the age of these measures remains a source of
concern, given the dynamic nature of culture (Chen et al., 2001; Hermans & Kempen, 1998;
Zhang & Shavitt, 2003). We also note that Hofstede’s measures are correlated and hence
cannot be interpreted as independent of one another in these analyses. It is nonetheless note-
worthy that these measures continue to demonstrate their relevance in analyses such as
those presented here. The effects of other cultural dimensions, such as those described by
Schwartz (1994) and Smith, Dugan, and Trompenaars (1996), should also be explored.
Finally, we note that our employee samples may not be representative of the nations within
which they were sampled, making cross-national comparisons problematic. Nonetheless, it
is important to note that our sample also represents an important strength. Most significant,
for assessing extreme response style, it incorporates large samples from 19 nations that
274 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

reflect considerable heterogeneity across each of the cultural dimensions of interest. By way
of contrast, most of the literature concerned with extreme response and acquiescence arti-
facts is based on comparisons across two or, at most, three nations or ethnic subgroups.
To our knowledge, there have been few attempts to link Hofstede’s exhaustively studied
national dimensions of culture with the response styles examined in this study. This research
consequently makes an important contribution to the small but expanding body of research
concerned with investigating and understanding the mechanisms by which culture influ-
ences the collection of questionnaire data (Harkness, van de Vijver, & Mohler, 2003). Under-
standing how culture influences individual response behaviors remains a prerequisite for the
sound practice of cross-cultural research.
NOTES
1. Prior analyses ofthesedata by thesecondauthor foundthatwithin each country, a55-item subsetofthe survey
established construct equivalence across nations. Specifically, an exploratory factor analysis of data from all 19
countries combined revealed 12 underlying factors related to work environment and satisfaction. Confirmatory fac-
tor analysis models conducted separately by country showed that this 12-factorsolution fit the data from each coun-
try adequately, as assessed by standard measures of model fit (i.e., GFI, AGFI, and CFI). Details of these analyses
can be obtained from the authors.
2. Two reanalyses ofthese data were conducted.The first investigated the utility of employing the originalmet-
rics of the Hofstede measures, rather than the collapsed versions depicted in Table 1. As described earlier, these mea-
sures had each been collapsed for use as level two variables in this study’s hierarchical analyses because the range of
each greatly exceeded the number of available observations. When the original Hofstede values are nonetheless
employed, only GNP was found to be associated with extreme response behavior, and only powerdistance, individ-
ualism, and GNP were found to be associated with acquiescent response styles. Also, it was noted that three of
the nations included in these analyses, the Czech Republic, Hungary, and Poland, were not included as part of
Hofstede’s (1980) original study. Rather, the data used to estimate cultural scores for these nations were derived
from unmatched sources other than the original IBM data (see Exhibit A5.3 in Hofstede, 2001, p. 502). A second
reanalysis of the models presented in Table3 was conducted with these three countries excluded. The revised model
for extreme response style produced identical findings. The revisedacquiescence model, however, did not converge
due to the small number of degrees of freedom (and number of countries)available (n= 7) for this revised analysis.
REFERENCES
Aday, L. A., Chiu, G. Y., & Anderson, R. (1980).Methodological issues in health care surveys of the Spanish heri-
tage population. American Journal of Public Health,70, 367-374.
Asch, S. E. (1956). Studies of independence and conformity: A minority of one against a unanimous majority. Psy-
chological Monographs,70(9, Whole No. 416).
Bachman, J. G., & O’Malley, P. M. (1984). Yea-saying, nay-saying and going to extremes: Black-white differences
in response styles. Public Opinion Quarterly,48, 491-509.
Bass, B. M., & Burger,P. C. (1979). Assessment of managers: An international comparison. New York: Free Press.
Baumgartner, H., & Steenkamp,J. B. (2001). Response styles in marketing research: A cross-national investigation.
Journal of Marketing Research,38, 143-156.
Bond, R., & Smith, P. B. (1996). Culture and conformity: A meta-analysis of studies using Asch’s (1952, 1956) line
judgment task. Psychological Bulletin,119, 111-137.
Carr, L. G. (1971). The Srole items and acquiescence. American Sociological Review,36, 287-292.
Chen, C., Shin-ying, L., & Stevenson, H. W. (1995). Response style and cross-cultural comparisons of rating scales
among East Asian and North American students. Psychological Science,6, 170-175.
Chen, X. P., Triandis, H. C., Kim, U., Carnevale,P.J., de Dreu, C., Gelfand, M. J., et al. (2001). Culture and deception
in business negotiations:A multilevel analysis. International Journal of Cross-CulturalManagement,1, 73-90.
Cheung, G. W., & Rensvold, R. B. (2000). Assessing extreme and acquiescence response sets in cross-cultural
research using structural equations modeling. Journal of Cross-Cultural Psychology,31, 187-212.
Johnson et al. / CULTURE AND RESPONSE STYLES 275

Chun, K. T., & Campbell, J. B. (1974). Extreme response style in cross-cultural research: A reminder. Journal of
Cross-Cultural Psychology,5, 465-480.
Clarke, I. (2000). Extreme response style in cross-cultural research: An empirical investigation. Journal of Social
Behavior and Personality,15, 137-152.
Couch, A. S., & Keniston, K. (1960). Yeasayers and naysayers: Agreeing response set as a personality variable.
Journal of Abnormal and Social Psychology,60, 151-174.
Cronbach, L. J. (1946). Response sets and test design. Educational and Psychological Measurement,6, 475-494.
Cunningham, W., Cunningham, I. C. M., & Green, R. T. (1977). The ipsative process to reduce response set bias.
Public Opinion Quarterly,41, 379-394.
Deutscher, I. (1973). Asking questions: Linguistic comparability. In D. Warwick & S. Osherson (Eds.), Compara-
tive research methods (pp. 163-186). Englewood Cliffs, NJ: Prentice Hall.
Gordon, L. V. (1976). Survey of interpersonal values, revised manual. Chicago: Science Research Associates.
Grimm, S. D., & Church, A. T.(1999). A cross-cultural study of response biases in personality measures. Journal of
Research in Personality,33, 415-441.
Hall, E. T. (1976). Beyond culture. Oxford, UK: Anchor.
Hamilton, D. L. (1968). Personality attributes associated with extreme response style. Psychological Bulletin,69,
192-203.
Harkness, J. A., van de Vijver, F. J. R., & Mohler, P. P. (2003). Cross-culturalsurvey methods. New York:John Wiley
& Sons.
Hermans, H. J. M., & Kempen, H. J. G. (1998). Movingcultures: The perilous problems of cultural dichotomies in a
globalizing society. American Psychologist,53, 1111-1120.
Hofstede, G. (1980). Culture’s consequences. Beverly Hills, CA: Sage.
Hofstede, G. (1998). Masculinity and femininity: The taboo dimension of national cultures. Thousand Oaks, CA:
Sage.
Hofstede, G. (2001). Culture’s consequences (2nd ed.). Thousand Oaks, CA: Sage.
House, R. J., Hanges, P. J., Javidan, M., Dorfman, P. W., & Gupta, V. (2004). Culture, leadership, and organizations:
The GLOBE study of 62 societies. Thousand Oaks, CA: Sage.
Hui, C. H., & Triandis, H. C. (1983). Multistrategy approach to cross-culturalresearch: The case of locus of control.
Journal of Cross-Cultural Psychology,14, 65-83.
Hui, C. H., & Triandis, H. C. (1989). Effects of culture and response format on extreme response style. Journal of
Cross-Cultural Psychology,20, 296-309.
Javeline, D. (1999). Responseeffects in polite cultures: A test of acquiescence in Kazakhstan. Public Opinion Quar-
terly,63, 1-28.
Johnson, T. P., O’Rourke, D., Chavez, N., Sudman, S., Warnecke, R., Lacey, L., et al. (1997). Social cognition and
responses to survey questions among culturally diverse populations. In L. Lyberg, P. Biemer, M. Collins, E. de
Leeuuw,C. Dippo, N. Schwarz, & D. Trewin (Eds.), Survey measurementand process quality (pp. 87-113). New
York: John Wiley & Sons.
Lee, C., & Green, R. T. (1991). Cross-cultural examination of the Fishbein behavioral intentions model. Journal of
International Business Studies,22, 289-305.
Marín, G., Gamba, R. J., & Marín, B. V. (1992). Extreme response style and acquiescence among Hispanics: The
role of acculturation and education. Journal of Cross-Cultural Psychology,23, 498-509.
Marín, G., & Marín, B. V. (1989). Research with Hispanic populations. Newbury Park, CA: Sage.
Morris, T., & Pavett, C. M. (1992). Management style and productivity in two cultures. Journal of International
Business Studies,23, 169-179.
Nelson, M. R., & Shavitt, S. (2002). Horizontal and vertical individualism and achievement values: A multimethod
examination of Denmark and the United States. Journal of Cross-Cultural Psychology,33, 439-458.
Niikura, R. (1999). Assertiveness among Japanese, Malaysian, Filipino, and U.S. white-collar workers. Journal of
Social Psychology,139, 690-699.
Raudenbush, S., Bryk, A., Cheong, Y.F., & Congdon, R. (2000). HLM 5: Hierarchical linear and nonlinear model-
ing. Chicago: Scientific Software International.
Riordan, C. M., & Vandenberg, R. J. (1994). A central question in cross-cultural research: Do employeesof different
cultures interpret work-related measures in an equivalent manner? Journal of Management,20, 643-671.
Robinson, W.S. (1950). Ecological correlations and the behavior of individuals. American Sociological Review,15,
351-357.
Ross, C. E., & Mirowsky, J. (1984). Socially-desirable response and acquiescence in a cross-cultural surveyof men-
tal health. Journal of Health and Social Behavior,25, 189-197.
Schwartz, S. H. (1994). Beyond individualism/collectivism: New cultural dimension of values. In U. Kim, H. C.
Triandis, Ç. Kagitçibasi, S. C. Choi, & G. Yoon (Eds.), Individualism and collectivism: Theory, method, and
applications (pp. 85-119). Thousand Oaks, CA: Sage.
Smith, P. B. (2004). Acquiescent response bias as an aspect of cultural communication style. Journal of Cross-
Cultural Psychology,35, 50-61.
Smith, P. B., Dugan, S., & Trompenaars, F. (1996). National culture and the values of organizational employees: A
dimensional analysis across 43 nations. Journal of Cross-Cultural Psychology,27, 231-264.
276 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

Stening, B. W., & Everett, J. E. (1984). Response styles in a cross-cultural managerial study. Journal of Social Psy-
chology,122, 151-156.
Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview Press.
Triandis, H. C., & Gelfand, M. J. (1998). Converging measurement of horizontal and vertical individualism and col-
lectivism. Journal of Personality and Social Psychology,74, 118-128.
Triandis, H. C., Marín, G., Lisansky, J., & Betancourt, H. (1984). Simpatìa as a cultural script for Hispanics. Journal
of Personality and Social Psychology,47, 1363-1375.
van Hemert, D. A., van de Vijver, F. J. R., Poortinga, Y. H., & Georgas, J. (2002). Structural and functional equiva-
lence of the Eysenck Personality Questionnaire within and between countries. Personality and Individual Dif-
ferences,33, 1229-1249.
van Herk, H. (2000). Equivalence in a cross-national context: Methodological & empirical issues in marketing
research. Unpublished doctoral dissertation, Tilburg University, the Netherlands.
van Herk, H., Poortinga, Y. H., & Verhallen, T. M. M. (2004). Response styles in rating scales: Evidence of method
bias in data from six EU countries. Journal of Cross-Cultural Psychology,35, 346-360.
Warnecke, R. B., Johnson, T. P.,Chavez, N., Sudman, S., O’Rourke, D., Lacey, L., et al. (1997). Improving question
wording in surveys of culturally diverse populations. Annals of Epidemiology,7, 334-342.
Watson, D. (1992). Correcting for acquiescent response bias in the absence of a balanced scale: An application to
class consciousness. Sociological Methods & Research,21, 52-88.
Welkenhuysen-Gybels, J., Billiet, J., & Cambre, B. (2003). Adjustment for acquiescence in the assessment of the
construct equivalence of Likert-type score items. Journal of Cross-Cultural Psychology,34, 702-722.
Williams, J. E., Satterwhite, R. C., & Saiz, J. L. (1998). The importance of psychological traits: A cross-cultural
study. New York: Plenum.
Winkler, J. D., Kanouse, D. E., & Ware, J. E. (1982). Controlling for acquiescence response set in scale develop-
ment. Journal of Applied Psychology,67, 555-561.
Wong,N., Rindfleish, A., & Burroughs, J. E. (2003). Do reverse-worded items confound measures in cross-cultural
consumer research? The case of the Material Values Scale. Journal of Consumer Research,30, 72-91.
Zax, M., & Takahashi, S. (1967). Cultural influences on response style: Comparisons of Japanese and American col-
lege students. Journal of Social Psychology,71, 3-10.
Zhang, J., & Shavitt, S. (2003). Cultural values in advertisements to the Chinese X-generation. Journal of Advertis-
ing,32, 23-33.
Timothy Johnsonis the director of the Survey Research Laboratory and a professor of public administration
in the College of Urban Planning and Public Affairs at the University of Illinois at Chicago. His main areas
of research interest include cross-cultural sources of measurement error in survey research and the social
epidemiology of substance abuse.
Patrick Kulesa received his Ph.D. in social psychology from Northwestern University. He is the global
researchdirector at ISR LLC, a global organization research firm specializing in employee and management
opinion surveys. ISR LLC’s research focuses on cross-national differences in employee opinion, the links
between organization cultureand company performance, and the effect of demographic factors (e.g., gender,
age, management level) on employee commitment and behavior.
Young Ik Cho is a research assistant professor at the Survey Research Laboratory in the College of Urban
Planning and Public Affairs at the University of Illinois at Chicago, from which he earned his Ph.D. in soci-
ology. His expertiseis in the analysis of secondarydata gathered with a complex survey design. He has pub-
lished extensively on substance use and abuse among minorities, and his current research interests include
the cross-cultural study on health, especially substance use and abuse among immigrants.
Sharon Shavitt receivedher Ph.D. in social psychology from the Ohio State University. She is an IBE Distin-
guished Professor of Business Administration, a research professor at the Survey Research Laboratory, and
a professor in the Department of Psychology at the University of Illinois at Urbana–Champaign. Her
researchinterests are at the intersection of social psychology and consumer behavior, focusing in particular
on cross-cultural factors affecting consumer persuasion and survey responding.
Johnson et al. / CULTURE AND RESPONSE STYLES 277