Content uploaded by Gerhard Meisenberg
Author content
All content in this area was uploaded by Gerhard Meisenberg on Oct 30, 2018
Content may be subject to copyright.
This article was published in an Elsevier journal. The attached copy
is furnished to the author for non-commercial research and
education use, including for instruction at the author’s institution,
sharing with colleagues and providing to institution administration.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Are acquiescent and extreme response styles related
to low intelligence and education?
Gerhard Meisenberg
*
, Amandy Williams
Ross University School of Medicine, Picard Estate, Portsmouth, Dominica
Received 5 June 2007; received in revised form 8 January 2008; accepted 11 January 2008
Available online 3 March 2008
Abstract
Several lines of evidence suggest that acquiescent and extreme response styles are related to low educa-
tion or low cognitive ability. Using measures constructed from the World Values Survey, this hypothesis is
examined both in comparisons among individuals and comparisons among countries. At the individual
level, both acquiescent and extreme responding are positively related to age and negatively to education
and income in most world regions. Both response styles are most prevalent in the less developed countries.
At the country-level, extremity is best predicted by a low average IQ in the country, and acquiescence by a
high level of corruption.
Ó2008 Elsevier Ltd. All rights reserved.
Keywords: Acquiescent response style; Extreme response style; World values survey; IQ; Education
1. Introduction
Response styles are important for the interpretation of personality and attitude measures by
self-report. The most commonly observed response styles, also described as response biases, are
acquiescent and extreme responding. Acquiescence refers to (unthinking?) agreement with state-
ments, and extremity to a preferential use of the end points of the scale. Some authors attributed
0191-8869/$ - see front matter Ó2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.paid.2008.01.010
*
Corresponding author. +1 767 255 6227.
E-mail address: gmeisenberg@rossmed.edu.dm (G. Meisenberg).
www.elsevier.com/locate/paid
Available online at www.sciencedirect.com
Personality and Individual Differences 44 (2008) 1539–1550
Author's personal copy
country-level differences in response styles to cultural traditions. For example, Cheung and
Rensvold (2000) speculated that the low level of extreme response style (ERS) in Korea might
be due to social convention: ‘‘Members of low-ERS cultures may desire to appear modest and
nonjudgmental, whereas members of high-ERS cultures may wish to demonstrate sincerity and
conviction.”(Cheung & Rensvold, 2000, p. 189).
Other lines of evidence suggest that response styles are related to intelligence. Messick and
Frederiksen (1958) found low verbal ability associated with acquiescent responding on a modified
form of the California F-scale. A more recent study found that on a Caribbean island both acqui-
escent and extreme responding are associated with low intellectual ability (Meisenberg, Lawless,
Lambert, & Newton, 2006, pp. 400–401).
Other observations suggest a negative relationship of intellectual ability or achievement with
response styles in comparisons between countries and cultural groups. In the United States, re-
sponses of Blacks and Hispanics are more extreme than those of non-Hispanic Whites (Bachman
& O’Malley, 1984; Hui & Triandis, 1989; Marin, Gamba, & Marin, 1992). Smith (2004) found
that European countries consistently scored lowest on measures of acquiescence derived from
six different surveys. The highest-scoring countries were those with lesser levels of economic devel-
opment, such as Panama, Nigeria, the Philippines, Pakistan and Bangladesh.
If low intellectual development directly favors extreme or acquiescent responding, we can pre-
dict an independent negative association of these response styles with education and intelligence
both in comparisons among individuals and in comparisons among countries. The present study
tests these predictions with response style measures derived from the World Values Survey.
2. Methods
2.1. The World Values Survey
The 2000 wave of the World Values Survey included 118,520 respondents in 80 countries. This
is an omnibus survey with more than 200 questions, but only some questions were asked in all or
nearly all countries. The raw data are available (on CD) with Inglehart, Basa
´n
˜ez, Dı
´ez-Medrano,
Halman, and Luijkx (2004). This survey is useful for the present purpose because of the great vari-
ety of questions asked and the large sample sizes. Its main weakness is the poor representativeness
of the country samples, and especially the oversampling of educated individuals in many of the
less developed countries.
2.2. Measures for acquiescence and extremity
A measure of acquiescent response style was constructed from 22 items in agree–disagree for-
mat that had been asked in a majority of the countries, for example: ‘‘The government should re-
duce environmental pollution, but it should not cost me any money: strongly agree/agree/
disagree/strongly disagree.”Two of the 22 questions had 2 answer choices, two had 3 choices,
eight had 4 choices, and ten had 5 choices. Positively correlated questions with similar content
were either avoided or averaged into a single score. In two cases, two highly correlated items with
similar content were averaged and given the weight of a single item, and in one case, three items
1540 G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550
Author's personal copy
were averaged. Whenever possible, questions with opposite content (e.g., one pro-democracy item
and one anti-democracy item) were included to reduce contamination of the bias measure with
substantive item content. The questions were scored starting with zero for the extreme disagree
response, and in increments of 1 for each successive option. Thus the maximum score was 1
for a 2-choice question and 4 for a 5-choice question. The sum of all items was the acquiescence
score. This measure is weighted heavily toward the 4-choice items and especially the 5-choice
items. However, when the items with 2 or 3 choices, 4 choices, and 5 choices were summed into
three separate acquiescence measures, all three were positively correlated at the individual level (2/
3 choices and 4 choices: r= 0.055; 2/3 choices and 5 choices: r= 0.231; 4 choices and 5 choices:
r= 0.248; Nis between 44,487 and 55,124) and the country-level (2/3 choices and 4 choices:
r= 0.264; 2/3 choices and 5 choices: r= 0.686; 4 choices and 5 choices: r= 0.411; N= 50 in each
case).
The measure for extreme response style was computed from 18 questions: 12 with 4 answer
choices, and 6 with 10 answer choices. Example: ‘‘How much confidence do you have in labor
unions? A great deal/quite a lot/not very much/none at all.”None of the questions was in
agree–disagree format. Highly correlated questions and those with very asymmetric responses
(including most of the religious items) were rejected. The 4-choice questions got a score of 1
for the two extreme choices, and zero for the two intermediate choices. The 10-choice items got
a score of 2 for the extreme choices, and zero for the 8 intermediate choices.
Since most questions had not been asked in all countries, and individual respondents had miss-
ing data for questions that had been asked, the acquiescence and extremity scores for these cases
had to be extrapolated by linear regression. In all, an acquiescence score was calculated for
109,728 respondents, and an extremity score for 80,809 respondents.
2.3. Country characteristics
The following country-level data were used:
1. Economic wealth was measured as the logarithm of gross domestic product (GDP) adjusted
for purchasing power, average for the years 1990–2005. Data were from the World Develop-
ment Indicators of the World Bank. In country-level regression models, lgIncome was
allowed to substitute for lgGDP whenever this improved the model fit. lgIncome was calcu-
lated from GDP and the average self-reported relative wealth of the survey respondents. This
variable was used because attitudinal and behavioral outcomes can depend both on the aver-
age wealth in the country, and the personal wealth of the respondents.
2. The average educational level in the country was operationalized by averaging the standard-
ized scores of 4 measures: (1) Average years of schooling of adults from the Barro-Lee dataset,
average for 1990–2000. (2) The school life expectancy (1999–2003 average), from UNESCO.
(3) The combined primary, secondary and tertiary school enrollment ratio in 2002 from the
2005 Human Development Report of the United Nations. (4) The arcsine-transformed aver-
ages of the adult literacy rates in 1990 and 2002 from the 2005 Human Development Report.
In regression models, the average self-reported education of the survey respondents (averaged
from length of schooling and highest educational degree) was allowed to substitute for the
country-level measure whenever this improved the model fit. The correlation of this sam-
G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550 1541
Author's personal copy
ple-level measure with the composite country-level measure was only 0.36, which shows the
poor representativeness of the samples interviewed in the World Values Survey.
3. Average country IQs are from Lynn and Vanhanen (2006) and Lynn (2006), who provide
measured scores for 114 countries and estimates for 79 additional countries. The Lynn &
Vanhanen IQs correlate with the composite measure of education at r= 0.750, and with
lgGDP at r= 0.653. The correlation between education and lgGDP is r= 0.800
(N= 186). An alternative measure of ‘‘intellectual competence”was formed by averaging
the Lynn & Vanhanen IQs with scores on international school achievement tests in mathe-
matics and science from TIMSS (Third International Mathematics and Science Study) and
PISA (Program for International Student Assessment), with missing data extrapolated. Mea-
sured IQ and school achievement correlate with r= 0.919 in those 57 countries for which
both data are available (Lynn, Meisenberg, Mikk, & Williams, 2007).
4. A measure of corruption was obtained by reversing the scores of the Corruption Perception
Index of Transparency International, averaged over the years 1999-2005. The correlation of
corruption with lgGDP is 0.800 (N= 158).
5. Measures for ‘‘freedom and democracy”included the scores of political freedom (political
rights + civil liberties) from Freedom House, 1988–2005 average; and Vanhanen’s democ-
racy index, 1990–2004 average, from the Finnish Social Science Data Archive. These highly
correlated measures (r= 0.847, N= 179), and the average of both, were used alternatively.
Details about internet-based data sources are available from the author on request.
6. Communism was scored alternatively in three ways: categorically as Communist history, cat-
egorically as ex-Communist, and as years under Communist rule.
Measures of corruption, freedom/democracy and Communist history were included because
these measures were shown previously to be related to personality and attitudinal outcomes (Mei-
senberg, 2004, in press). The major cultural regions of the world were defined similar to Inglehart
et al. (2004). However, Uruguay was assigned to Latin America rather than Catholic Europe;
Greece and Israel were lumped with Catholic Europe into a ‘‘Catholic Europe (+ Mediterra-
nean)”category; the Muslim countries from Morocco to Pakistan formed the ‘‘Middle East”; In-
dia, Bangladesh, the Philippines and Indonesia were ‘‘South (+ Southeast) Asia”; and China and
Vietnam were labeled ‘‘Asian Communist.”‘‘Africa”includes sub-Saharan Africa only, repre-
sented by Nigeria, South Africa, Tanzania, Uganda and Zimbabwe.
3. Results
3.1. Differences between world regions
Extremity and acquiescence were positively correlated, with r= 0.241 at the individual level
(N= 79,053 respondents) and r= 0.601 at the country-level (N= 79 countries). Table 1 shows
that relative to individual differences within countries, differences between world regions are sub-
stantial: up to 1.52 within-country standard deviations for acquiescence and 1.33 standard devi-
ations for extremity. Both response biases are more prominent in economically less developed
regions such as Africa, South Asia and the Middle East.
1542 G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550
Author's personal copy
3.2. Correlates at the individual level
Tables 2 and 3 show the results of regression models in which the response styles were predicted
jointly by sex, age, education and income, with country-level effects controlled. Acquiescence is
greater in older, poorer and less educated people except in the countries of South and Southeast
Asia. Also extreme responding tends to be associated with older age, low education and low in-
come, but in this case maleness is the most consistent predictor.
3.3. Correlates at the country-level
Table 4 shows that at the country-level, acquiescence is most closely associated with high cor-
ruption and somewhat less with low lgGDP and low education. Extreme response style shows a
Table 1
The levels of response bias in the major ‘‘cultural provinces”of the world
Region Acquiescence Extremity
Protestant Europe .857 .468
Catholic Europe .326 .331
English-speaking .739 .355
Ex-Communist .108 .169
Middle East .487 .559
Latin America .128 .417
South Asia .658 .352
East Asia .213 .693
Asian Communist .173 .037
Africa .475 .632
The scores are standardized to a world mean of zero and an average within-country standard deviation of 1.
Table 2
Relationship of acquiescent response style with four demographic predictors
Region NSex Age Education Income
Protestant Europe 6799 .019 .219
***
.164
***
.034
**
Catholic Europe 10,168 .015 .177
***
.110
***
.054
***
English-speaking 7465 .021 .106
***
.153
***
.086
***
Ex-Communist 26,068 .020
**
.120
***
.035
***
.011
Middle East 13,476 .005 .046
***
.024
**
.051
***
Latin America 12,747 .012 .052
***
.041
***
.043
***
South Asia 5200 .008 .007 .004 .044
**
East Asia 4433 .020 .148
***
.036
*
.032
*
Asian Communist 1795 .062
**
.111
***
.026 .054
*
Africa 6879 .003 .076
***
.022 .096
***
Partial correlations with acquiescent response style are shown, with the other three predictors held constant.
*
p< 0.05.
**
p< 0.01.
***
p< 0.001.
G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550 1543
Author's personal copy
different pattern, being most strongly associated with IQ, less with education, and even less with
lgGDP.
The relationships of the response styles with country-level characteristics were explored further.
Regression models were started with IQ, education, lgGDP, corruption, the average of political
freedom and democracy, and Communist history. Nonlinear effects of the predictors were mod-
eled by the inclusion of quadratic terms. During development of the models, alternative measures
of the same construct were explored to improve the fit, and non-predictors were dropped. Tables 5
and 6 show the best-fitting models.
Acquiescence is predicted independently by high corruption and to a lesser extent by low edu-
cation. Only the average education of the sample, but not the average education in the country
predicts the outcome. This reinforces the observation that low education predicts high acquies-
cence at the individual level within countries (Table 2). However, measures of intelligence do
not independently predict acquiescent responding at the country-level. This is not due to collin-
Table 3
Relationship of extreme response style with four demographic predictors
Region NSex Age Education Income
Protestant Europe 5615 .047
***
.065
***
.056
***
.042
**
Catholic Europe 8028 .071
***
.057
***
.018 .042
***
English-speaking 6529 .053
***
.065
***
.106
***
.021
Ex-Communist 17,953 .039
***
.017
*
.007 .017
*
Middle East 8076 .075
***
.011 .009 .024
*
Latin America 10,726 .038
***
.048
***
.067
***
.035
***
South Asia 4017 .050
**
.014 .041
*
.096
***
East Asia 3421 .056
**
.009 .076
***
.021
Asian Communist 1334 .057
*
.016 .026 .021
Africa 5790 .072
***
.038
**
.018 .027
*
Partial correlations with extreme response style are shown, with the other three predictors held constant. Partial
correlations of extreme response style with demographic predictors.
*
p< 0.05.
**
p< 0.01.
***
p< 0.001.
Table 4
The correlations of acquiescent and extreme response styles with country characteristics (Pearson’s r)
Acquiescence Extremity N
Extremity .601 79
IQ .551 .783 79
Education .676 .702 79
lgGDP .688 .630 79
Corruption .735 .567 78
Political freedom .574 .453 79
Democracy .543 .455 78
All correlations are significant at the p< 0.001 level.
1544 G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550
Author's personal copy
earity between intelligence and education, because IQ is not an independent predictor even in
models that do not include education (data not shown).
Table 6 shows that extreme responding is related mainly to low intelligence. This confirms the
impression from Table 4, and also the pattern in Table 1 showing that extreme responding is most
prevalent in the world region with the lowest average IQ (sub-Saharan Africa, average IQ 67
according to Lynn, 2006), and least prevalent in the region with the highest IQ (East Asia, average
IQ 105). The effect of education is curvilinear, reducing extreme responding at low levels of edu-
cation but raising it among the most educated samples. Among the economic and political indi-
cators, only corruption has a significant effect.
The model in Table 6 contains the average intelligence in the country, but the average educa-
tional level of the interviewed sample. This can lead to sampling artifacts because the intelligence
of the interviewed samples is affected by the average education of the sample relative to the aver-
age education in the country. Therefore the average IQs of the samples were estimated by adjust-
ing the country IQ according to the extent to which different educational levels were over-or
undersampled in the World Values Survey. When this sample-level measure was used instead
of the country-level IQ, the linear bias-reducing effect of education vanished. Thus the bias-reduc-
ing effect of education in the model of Table 6 can be attributed to greater intellectual ability of
the more highly educated samples.
Table 6
The best-fitting model for extreme response style
Predictor st. bp
Intelligence .530 <.001
Intelligence
2
.114 .093
Education .203 .003
Education
2
.186 .004
lgInc
2
.129 .080
Corruption .262 .010
Ex-Communist .101 .210
Intelligence is a composite of IQ and school achievement. Education and the logarithm of income (lgInc) are sample
characteristics, the other predictors are country characteristics. N= 78 countries, R
2
= 0.745.
Table 5
The best-fitting model for acquiescent response style
Predictor st. bp
Education .190 .016
Education
2
.080 .281
lgGDP .236 .109
Corruption .513 .002
Political freedom
2
.138 .095
Years Communist .100 .245
Education, average education of the interviewed sample. The other variables are country-level characteristics. N=78
countries; R
2
= 0.632.
G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550 1545
Author's personal copy
3.4. Region-specific effects
To investigate the possibility that region-specific cultural factors are responsible for differences
in response styles, the world regions were dummy-coded and were added, one at a time, to the
regression models in Tables 5 and 6. The only significant effect was a higher-than-expected level
of extreme responding in Latin America (p= 0.007). The reasons for this effect are uncertain.
3.5. Alternative measures of response styles
If the computed scores represent true acquiescence and extremity rather than the incidental ef-
fects of substantive question content, then alternative measures of these constructs should not
only correlate with each other but also have the same external correlates. Therefore two alterna-
tive measures were formed from subsets of the questions that had been used for the computation
of the original acquiescence and extremity scores. For acquiescence, alternating items in the se-
quence in which they appeared in the survey were computed into two separate measures, except
for two cases in which the sequence was changed to avoid highly correlated items. For extremity,
one measure was formed from the twelve 4-choice items and another from the six 10-choice items.
The correlation between the two acquiescence measures was 0.375 at the individual level
(N= 95,408 respondents) and 0.708 at the country-level (N= 77 countries). Both correlated most
highly with corruption (positive), education and lgGDP (both negative). For both measures, the
results of regression models were not substantially different from those in Table 5.
The correlation between the two extremity measures was 0.261 at the individual level
(N= 84,432 respondents) and 0.602 at the country-level (N= 78 countries). Both correlated most
strongly with measures of intelligence, and either IQ or the composite of IQ and school achieve-
ment was the predominant predictor in the best-fitting regression models.
3.6. Generality of response styles
To determine the importance of response styles for country-level scores on those questions that
do not fit the narrow definitions used in selecting items for the bias measures, answers to such
questions were predicted by a model that included IQ, education, lgGDP, corruption, freedom/
democracy, Communist history, and either extreme response style or acquiescent response style.
The analysis was limited to questions that had been asked in at least 70 countries and that had not
been used in the computation of the bias measures.
Extreme response style predicted responses on only one of 46 2-choice questions with p< 0.05,
which is at chance level. For another 46 questions with at least 3 choices and the opportunity for
extreme responding, extreme response style predicted responses to 3 questions with
0.01 < p< 0.05, and 2 questions with p< 0.01. Thus extreme response style is unimportant for
aggregate scores on 2-choice questions and affects only few questions with more than two choices.
As expected, when one end of the scale had far higher responses than the other, extreme response
style shifted aggregate responses towards the more popular end of the scale.
Results were very different for acquiescence. For 2 out of 92 questions that did not have the
classical agree–disagree format, acquiescence influenced the score independent of the other predic-
tors with p< 0.001; for 5 additional questions the significance level was p< 0.01; and for 16 addi-
1546 G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550
Author's personal copy
tional questions it was p< 0.05. Thus acquiescence affects responses on many questions that are
not asked with agree–disagree options. Some observations are:
1. Acquiescence biases responses away from negatively worded choices. Example: With which
statement do you tend to agree: A. Regardless of what the qualities and faults of one’s parents
are, one must always love and respect them/One does not have the duty to respect and love par-
ents who have not earned it by their behavior and attitudes (p-value for acquiescence:
p< 0.001).
2. When the choice is between ‘‘good”and ‘‘bad”, acquiescence biases answers toward ‘‘good”.
Example: Please tell me... whether it would be a good thing, a bad thing or don’t you mind?
More emphasis on family life: good/don’t mind/bad (p< 0.001). This confirms and extends
early observations that ‘‘true”and ‘‘yes”responses are associated with ‘‘agree”responses
at the level of individuals (Couch & Keniston, 1960).
3. Acquiescence is greatest for agreement with contents of a conventional, conformist, moralist
or authoritarian nature. Example: How much confidence do you have in the police?: A great
deal/quite a lot/not very much/none at all (p= 0.003), or: Homosexuality is: never justifiable
– (10-choice scale) – always justifiable (p= 0.003). In the last example, acquiescers actually
prefer the negatively worded choice (‘‘never”justifiable). Interestingly, the likelihood of
belonging to a religious denomination is higher in country samples with high acquiescence
(p= 0.006). Acquiescence effects on individual questions are generally similar at the individ-
ual and country-levels.
4. Discussion
Table 2 shows that the negative association between education and acquiescence that had been
reported by some investigators (Heaven, 1983; Javeline, 1999; Mirowsky & Ross, 1991; Watson,
1992) is a nearly worldwide phenomenon. Also extreme responding is favored by low education
and low income in most countries (Table 3). Therefore, psychological factors related to education,
such as intelligence, rationality, self-control or self-confidence, appear to suppress both acquies-
cent and extreme responses.
The demographic variables used in Tables 2 and 3 explain only 1–5% of the individual differ-
ences in most country samples, but country-level characteristics explain up to 63.2% of the vari-
ance in acquiescence and 74.5% of the variance in extremity in comparisons between countries
(Tables 5 and 6). These values are comparable to those for substantive variables such as religiosity
and subjective well-being (Meisenberg, 2004, in press).
Given the great variety of questions that were averaged into the acquiescence and extremity
scores, systematic contamination of the bias measures by question content is unlikely. This is con-
firmed by the observation that for both response styles, measures formed from two different sub-
sets of the questions still correlated with external country characteristics in the same way as did
the original measures.
The observation that extreme responding is most closely related to low IQ at the country-level
(Tables 4 and 6) suggests that it expresses a crude ”either-or thinking”that is oblivious to fine
G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550 1547
Author's personal copy
distinctions. This is confirmed by the observation that within countries, extreme responding is re-
duced in persons with higher education and/or higher income.
The positive association of age and extreme responding might be related to the secular rise in
IQ known as the Flynn effect (Flynn, 1987), which implies greater intellectual sophistication in the
younger cohorts. The association of extreme responding with age is not seen in most African and
Asian countries, for which the extent or even existence of a Flynn effect is unknown. However,
also in East Asia extreme responding is barely age-dependent although the Flynn effect was strong
in these countries (Flynn, 1987; Lynn & Hampson, 1986). In this case we can postulate that the
bias-enhancing effect of low cognitive development is offset by more specific cultural changes that
reduce extreme responding in the older but not the younger generation.
Males are no less intelligent than females (Born, Bleichrodt, & van der Flier, 1987; Halpern,
2000). Therefore their more extreme responding is more likely related to other personality traits,
for example greater impulsivity and weaker inhibitory controls (Bjorklund & Kipp, 1996). More
extreme responding in males has not been observed in at least three other studies of extreme re-
sponse style (Bachman & O’Malley, 1984; Clarke, 2000; Greenleaf, 1992), and therefore its repro-
ducibility needs to be ascertained.
In regression models that contain education, lgGDP and/or corruption as covariates, acquies-
cence is not independently related to intelligence at the country-level (Tables 4 and 5) if the anal-
ysis is performed without a measure of education. This is unexpected because earlier studies have
shown a negative association between acquiescence and intelligence at the individual level (Elliott,
1961; Meisenberg et al., 2006; Messick & Frederiksen, 1958). The same is suggested by the neg-
ative correlations of acquiescence with education and income in Table 2. One possible explanation
is that in every society poor and uneducated individuals are at least somewhat aware of their lim-
itations. They cope with these limitations by developing a habit of agreeing to most statements
and proposals that are brought forward by their brighter compatriots. Since this is a response
to interpersonal comparisons, it produces an independent negative association between acquies-
cence and intelligence at the individual but not the country-level.
Thus acquiescence appears to be favored by a lack of self-confidence, self-esteem or assertive-
ness, or a habit of subordination and conformity to others, and only indirectly by low intelligence.
This can explain its association with corruption. People living in corrupt societies must be subser-
vient to powerful others, and they might carry this habit over into their survey responses – or else
the social conformity that produces acquiescent responding also favors corruption. Self-confi-
dence and assertiveness can also explain the negative association of acquiescence with education
at the level of individuals (Table 2) and country samples (Table 5), because educated people are
likely to answer questions with greater confidence.
Acquiescence appears to favor responses that conform to the traditional or prevailing values in
the society including family values and moral/religious values. Perhaps the greater acquiescence of
older individuals is related to a lesser willingness to question traditional values and social conven-
tions. Acquiescence as conformity can explain why the scales that are most likely to be con-
founded by acquiescence are those that measure conformity to traditional values, including the
infamous California F-scale (Messick & Frederiksen, 1958; Ray, 1979).
Several authors have attributed country-level variation in response styles to specific cultural
traits including individualism, collectivism, uncertainty avoidance, power distance and masculin-
ity (Harzing, 2006; Johnson, Kulesa, Cho, & Shavitt, 2005; Smith, 2004). These claims need to be
1548 G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550
Author's personal copy
evaluated against the competing hypothesis that response styles are more directly related to coun-
try-level intelligence, the educational level of the country’s population or that of the interviewed
sample, or other ‘‘development indicators.”There is also a need for the inclusion of cognitive
measures in cross-cultural surveys of personality, attitudes and values.
References
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.
Bjorklund, D. F., & Kipp, K. (1996). Parental investment theory and gender differences in the evolution of inhibition
mechanisms.Psychological Bulletin, 120, 163–188.
Born, M. P., Bleichrodt, N., & van der Flier, H. (1987). Cross-cultural comparison of sex-related differences on
intelligence tests. A meta-analysis.Journal of Cross-Cultural Psychology,18, 283–314.
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.
Clarke, I. (2000). Extreme response style in cross-cultural research: An empirical investigation.Journal of Social
Behavior and Personality,15, 137–152.
Couch, A., & Keniston, K. (1960). Yea sayers and nay sayers: Agreeing response set as a personality variable.Journal of
Abnormal and Social Psychology,60, 151–174.
Elliott, L. L. (1961). Effects of item construction and respondent aptitude on response acquiescence.Educational and
Psychological Measurement,21, 405–415.
Flynn, J. R. (1987). Massive IQ gains in 14 nations: What IQ tests really measure.Psychological Bulletin, 101, 171–191.
Greenleaf, E. A. (1992). Measuring extreme response style.Public Opinion Quarterly,56, 328–351.
Halpern, D. (2000). Sex differences in cognitive abilities (3rd ed.). Mahwah (NJ): Erlbaum.
Harzing, A.-W. (2006). Response styles in cross-national survey research. A 26-country study.International Journal of
Cross Cultural Management,6, 243–266.
Heaven, P. C. L. (1983). Authoritarianism or acquiescence? South African findings.Journal of Social Psychology, 119,
11–15.
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.
Inglehart, R., Basa
´n
˜ez, M., Dı
´ez-Medrano, J., Halman, L. & Luijkx, R. (2004). Human beliefs and values. A cross-
cultural sourcebook based on the 1999–2002 values surveys.Me
´xico: Siglo XXI Editores.
Javeline, D. (1999). Response effects in polite cultures. A test of acquiescence in Kazakhstan.Public Opinion Quarterly,
63, 1–28.
Johnson, T., Kulesa, P., Cho, Y. I., & Shavitt, S. (2005). The relation between culture and response styles. Evidence
from 19 countries.Journal of Cross-Cultural Psychology,36, 264–277.
Lynn, R. (2006). Race differences in intelligence: An evolutionary analysis. Athens (GA): Washington Summit.
Lynn, R., & Hampson, S. (1986). The rise of national intelligence: Evidence from Britain, Japan and the USA.
Personality and Individual Differences,7, 23–32.
Lynn, R., Meisenberg, G., Mikk, J., & Williams, A. (2007). National IQs predict differences in scholastic achievement in
67 countries.Journal of Biosocial Science,39, 861–874.
Lynn, R., & Vanhanen, T. (2006). IQ and global inequality. Athens (GA): Washington Summit.
Marin, G., Gamba, R. J., & Marin, B. V. (1992). Extreme response style and acquiescence among Hispanics.Journal of
Cross-Cultural Psychology,23, 498–509.
Meisenberg, G. (2004). Talent, character, and the dimensions of national culture.Mankind Quarterly,45, 123–168.
Meisenberg, G. (in press). Intelligenz- und Wertentwicklung in Gesellschaften an der Schwelle zur Moderne
[Development of intelligence and values in societies at the threshold to modernity]. In G. W. Oesterdiekhoff, & H.
Rindermann (Eds.), Kultur und Kognition: Die Beitra
¨ge von Psychometrie und Piaget-Psychologie zum Versta
¨ndnis
kultureller Unterschiede. Hamburg: Lit Verlag.
G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550 1549
Author's personal copy
Meisenberg, G., Lawless, E., Lambert, E., & Newton, A. (2006). The social ecology of intelligence on a Caribbean
island.Mankind Quarterly,46, 395–433.
Messick, S., & Frederiksen, N. (1958). Ability, acquiescence, and ‘‘authoritarianism”.Psychological Reports,4,
687–697.
Mirowsky, J., & Ross, C. E. (1991). Eliminating defense and agreement bias from measures of the sense of control: A
22 index.Social Psychology Quarterly,54, 127–145.
Ray, J. J. (1979). Is the acquiescent response style problem not so mythical after all? Some results from a successful
balanced F scale.Journal of Personality Assessment,43, 638–643.
Smith, P. B. (2004). Acquiescent response bias as an aspect of cultural communication style.Journal of Cross-Cultural
Psychology,35, 50–61.
Watson, D. (1992). Correcting for acquiescent response bias in the absence of a balanced scale.Sociological Methods &
Research,21, 52–88.
1550 G. Meisenberg, A. Williams / Personality and Individual Differences 44 (2008) 1539–1550
