The Structure and Sources of Right-Wing Authoritarianism and Social Dominance Orientation

Abstract

Right-wing authoritarianism (RWA) and social dominance orientation (SDO) are widely used constructs in research on social and political attitudes. This study examined their hierarchical and correlative structure (across sexes, generations, and rater perspectives), as well as how genetic and environmental factors may contribute to individual differences in them (using different rater perspectives and nuclear twin family data). We found a substantive common aspect (beyond shared artificial variance arising from socially desirable responding) underlying both RWA and SDO: Aggression against subordinate groups. We discussed how this aspect could help to explain the commonly-reported correlation between the two concepts in Western countries. Estimates of genetic and environmental components in RWA and SDO based on self-reports were quite comparable to those based on peer reports. When controlling for error variance and taking assortative mating into account, individual differences in RWA were primarily due to genetic contributions including genotype-environment correlation, whereas variance in SDO was largely attributable to environmental sources shared and not shared by twins. The findings are discussed in terms of the utility of RWA and SDO as basic constructs to describe individual differences in social attitudes, and with respect to the different patterns of genetic and environmental influences that underlie them.

Full text

The Structure and Sources of Right-Wing Authoritarianism and Social Dominance
Orientation
Christian Kandler1, Edward Bell², and Rainer Riemann1
Pre-print version of a manuscript accepted for publication in
European Journal of Personality
July 1st, 2016
____________________
1Bielefeld University, Department of Psychology and Sport Sciences, Bielefeld, Germany
²Brescia University College at Western University, School of Social and Behavioral Sciences,
London, Canada
Correspondence concerning this article should be addressed to Christian Kandler,
Department of Psychology and Sport Sciences, Bielefeld University, Universitätsstr. 25, D-
33615 Bielefeld, Germany. Email: christian.kandler@uni-bielefeld.de

2
2
Structure and Sources of Social Attitudes
Abstract
Right-wing authoritarianism (RWA) and social dominance orientation (SDO) are
widely used constructs in research on social and political attitudes. This study examined
their hierarchical and correlative structure (across sexes, generations, and rater
perspectives), as well as how genetic and environmental factors may contribute to
individual differences in them (using different rater perspectives and nuclear twin family
data). We found a substantive common aspect (beyond shared artificial variance arising
from socially desirable responding) underlying both RWA and SDO: Aggression against
subordinate groups. We discussed how this aspect could help to explain the commonly-
reported correlation between the two concepts in Western countries. Estimates of genetic
and environmental components in RWA and SDO based on self-reports were quite
comparable to those based on peer reports. When controlling for error variance and taking
assortative mating into account, individual differences in RWA were primarily due to genetic
contributions including genotype-environment correlation, whereas variance in SDO was
largely attributable to environmental sources shared and not shared by twins. The findings
are discussed in terms of the utility of RWA and SDO as basic constructs to describe
individual differences in social attitudes, and with respect to the different patterns of
genetic and environmental influences that underlie them.
Keywords: right-wing authoritarianism; social dominance orientation; multi-method study;
nuclear twin family study; social attitudes

3
3
Structure and Sources of Social Attitudes
The Structure and Sources of Right-Wing Authoritarianism and Social Dominance
Orientation
Decades of research have demonstrated the importance of right-wing
authoritarianism (RWA) and social dominance orientation (SDO) as two basic dimensions of
individual differences in social and political attitudes (e.g., Altemeyer, 1981; Duckitt &
Sibley, 2010; Pratto, Sidanius, Stallworth, & Malle, 1994). In recent years, however,
researchers have begun to question the unidimensionality of RWA and SDO, providing
empirical support for the existence of two or more different dimensions in each one (e.g.,
Duckitt & Bizumic, 2013; Funke, 2005; Ho et al., 2012). Moreover, recent behavioral genetic
research points to different sources of individual differences in RWA and SDO: Whereas
RWA is moderately to strongly influenced by genetic factors, with about 50% of its variance
attributable to genetic influences (e.g., Funk et al., 2013; Lewis & Bates, 2014; Ludeke &
Krueger, 2013), the genetic component in operationalizations of SDO appears to be small or
even negligible, less than 30% (Kandler, Bleidorn, & Riemann, 2012; Orey & Park, 2012).
The current study examined the hierarchical and correlative structure and sources of
individual differences in RWA and SDO based on multi-generational, multi-method, and
genetically informative data. This work addressed several research questions. First, do RWA
and SDO each have conceptually different facets, as recent studies have suggested? Second,
is the correlative and hierarchical structure invariant across sexes, generations (parents and
offspring), and informant perspectives (self-reports and peer reports)? Third, is the
expected differential nature regarding the contributions of the genetic and environmental
factors to individual differences in RWA and SDO comparable across self- and peer reports?
And finally, what can we learn about the intergenerational transmission of individual
differences in RWA and SDO (beyond the important but rather limited information offered

4
4
Structure and Sources of Social Attitudes
by the classical twin approach) by using a nuclear twin family design that takes both the
assortative mating of the twins’ parents and genotype-environment interplay into account?
The Structure of Individual Differences in RWA and SDO
Based on the formative work by Adorno and colleagues (Adorno, Frenkel-Brunswick,
Levinson, & Sanford, 1950), Altemeyer (1981) developed the concept of RWA and produced
a scale to capture it. He defined RWA as a core preference for establishing or maintaining
collective security, societal stability, and cultural tradition. His RWA measure encompasses
an item pool taken in part from previous measures of authoritarianism (e.g., the F-scale of
Adorno et al., 1950; see also Fromm, 1941) and conservatism. It predicts a broad spectrum
of social and political attitudes such as cultural conservatism, positive opinions about
punitive policies, and diverse forms of outgroup prejudice and discrimination (e.g.,
Altemeyer, 1996; Duriez & Van Hiel, 2002; Kandler, Lewis, Feldhaus, & Riemann, 2015). Even
though Altemeyer noted that RWA includes three conceptually distinct aspects, which he
labelled as aggression against subordinates (preference for harsh, punitive, and strict social
control), submission to authority (a predilection for obedient, respectful, and uncritical
support for authorities), and conventionalism (support for traditional norms, values, and
rules), he did not attempt to measure them as separate facets of RWA.
Nonetheless, as suggested earlier, researchers have recently challenged the assumed
unidimensionality of RWA and have attempted to measure it multidimensionally (Duckitt,
Bizumic, Krauss, & Heled, 2010; Funke, 2005; Van Hiel, Cornelis, Roets, & De Clrecq, 2007).
Duckitt and Bizumic (2013), for instance, observed three factor-analytically distinct
dimensions and found measurement invariance across Serbian and New Zealand samples.
They also showed that the three facets predicted different prejudices.

5
5
Structure and Sources of Social Attitudes
Similar findings have been reported for the distinctiveness of specific facets of SDO,
a construct defined by its originators as a core presumption of ingroup superiority and a
preference for intergroup hierarchy (Pratto et al., 1994). Even though SDO as a unitary
construct predicts several intergroup attitudes and social conservatism (Sidanius & Pratto,
1999), Ho and colleagues (2012) found two factor-analytically distinct facets: SDO-
Dominance (belief in the superiority of one’s own group and support for its dominance over
outgroups) and SDO-Egalitarianism (preference for social and economic inequality between
groups). The factorial distinctiveness also held for SDO-D and SDO-E dimensions that had an
equal number of pro-trait and contra-trait item formulations (Ho et al., 2015). Also, studies
have demonstrated the differential predictive validity of the SDO facets: Whereas SDO-D
was more related to racism, outgroup discrimination, and intergroup aggression, SDO-E
better predicted hierarchy-enhancing beliefs and attitudes toward social policies (Ho et al.,
2012; Jost & Thompson, 2000; Kugler, Cooper, & Nosek, 2010).
Although their contributions to various specific prejudices and behaviors are
primarily additive (Sibley, Robertson, & Wilson, 2006), RWA and SDO are typically positively
correlated, at least in Western, individualistic countries, where authoritarian aggression and
conservatism have generally entailed a defense of social and economic hierarchy (Aspelund,
Lindeman, & Verkasalo, 2013; Duriez, Van Hiel, & Kossowska, 2005). A few studies,
however, have also provided evidence that group-based dominance has a stronger
relationship with RWA than does opposition to equality (Cohrs, Moschner, Maes, &
Kielmann, 2005; Kugler et al., 2010). This may be because aggression against and
domination over subordinate groups is an aspect of RWA as well as SDO and thus could
explain the correlation between RWA and SDO to some degree. The correlation between
RWA and SDO may also be due to rater biases to some extent, because intergroup

6
6
Structure and Sources of Social Attitudes
aggression and dominance are less socially acceptable than conservative attitudes or
preferences for social and economic inequality. As a consequence, the correlation between
RWA and SDO could be reduced and the structure of RWA and SDO could look different if
common substantive aspects and measurement artifacts were taken into account.
The Sources of Individual Differences in RWA and SDO
Despite their common aspects, behavioral genetic research points to differential
contributions of genetic and environmental factors to individual differences in RWA and
SDO. As noted above, previous studies have found a moderate-to-strong genetic
contribution to the variance in RWA (e.g., Bouchard & McGue, 2003; Ludeke, Johnson, &
Bouchard, 2013; McCourt et al., 1999), whereas the heritability of SDO or specific facets,
such as ethnocentrism and attitudes toward inequality, tend to be lower (e.g., Kandler et al.,
2012; Kandler, Bell, Shikishima, Yamagata, & Riemann, 2015; Olson, Vernon, Harris, & Jang,
2001). Therefore, individual differences in SDO appear to be more influenced by
environmental factors than is the case for RWA, and this difference may be even greater
when the common aspect of RWA and SDO is controlled.
Even though the classic twin study design allows us to disentangle heritability from
two kinds of environmental influences (i.e., environmental factors shared by twins and
those that are not shared), studies solely relying on twin data cannot take the similarity of
the twins’ parents regarding RWA and SDO into account. This is problematic because
parental similarity seems to be at least moderate for social attitudes (Bouchard, 2009;
Watson et al., 2004), and such similarity may not only be due to shared social backgrounds
(i.e., social homogamy) but also to genetically-driven assortative mating (Kandler et al.,
2012; Watson, Beer, & McDate-Montez, 2014). This source of nonrandom mother-father
similarity would act to increase the genetic similarity between parents and offspring and

7
7
Structure and Sources of Social Attitudes
between siblings, except for monozygotic (MZ) twins who are genetically identical. Because
genetic influences are indicated when MZ twins are more similar on a particular trait than
other first-degree relatives, and because estimates of shared environmental contributions
increase as the similarity between MZ twins compared to other first-degree relatives
decreases, environmental influences shared by family members would be overestimated
and heritability underestimated if assortative mating were not taken into account. To avoid
that problem, it is necessary to investigate the genetic and environmental sources of
individual differences in a trait using an extended twin family design. This was done in the
current study by using the nuclear twin family design (NTFD; Keller, Medland, & Duncan,
2010; see also Kandler, Gottschling, & Spinath, 2016) to analyze the sources of individual
differences in RWA and SDO.
By controlling for the influence of parental similarity, the NTFD also allows for a
disentanglement of diverse influences on twins’ similarity on a particular trait, namely
genetic and environmental parent-child transmission as well as non-parental environmental
factors shared by twins. An additional important phenomenon that the NTFD allows
researchers to test for is passive genotype-environment correlation (Scarr & McCartney,
1983), which can occur if the genotype twins inherit from their parents correlates with the
parental environment in which the twins are raised. For example, authoritarianism is partly
heritable, but children with a genetic predisposition toward authoritarianism may also be
exposed to an authoritarian parenting style which enhances their innate tendencies in that
regard. In this way there may be a correlation between children’s genetic makeup and the
sort of parental environment in which they are raised. Knowledge of such correlations adds
considerably to our understanding of the traits in question. The inclusion of assortative
mating, diverse shared environmental factors, and passive genotype-environment

8
8
Structure and Sources of Social Attitudes
correlation in the analyses offered here provides an important extension to the classical
twin study approach.
Aims of the Current Study
Earlier research, cited above, provided support for studying both RWA and SDO
multidimensionally. This article provides an extended test of these concepts’
multidimensionality by using Principal Axis Factoring (PAF) analysis to explore the necessary
number of dimensions and Confirmatory Factor Analyses (CFA) to examine the hierarchical
and correlative structure of RWA and SDO. This extended analysis includes an examination
of the association between the two core attitudinal dimensions when controlling for
common substantive aspects (e.g., aggression against subordinate groups) and common
measurement artifacts (e.g., social desirability). Our expectation was that once these
common aspects were brought into the analyses, the correlation between RWA and SDO
would be reduced.
Since older people appear to endorse conservative and authoritarian values more
readily than younger people, and men tend to prefer intergroup inequality and dominance
against outgroups more than women (Brandt & Henry, 2012; Duriez & Van Hiel, 2002;
Feather, 1977; Sidanius, Pratto, & Bobo, 1994), the most parsimonious and appropriate SEM
model was tested for structural invariance across males and females as well as across
different generations.
Previous research on RWA and SDO was primarily based on self-reports, even though
other measures, such as ratings from well-informed peers, are quite informative and
provide comparable psychometric quality and accuracy in capturing core attitudes and
prejudices (Cohrs, Kämpfe-Hargrave, & Riemann, 2012). Therefore, we also tested the best
fitting model across self-reports and peer reports.

9
9
Structure and Sources of Social Attitudes
To get basic information on the underlying influences on RWA and SDO, we used
classical twin modeling to examine the genetic and environmental contributions to
individual differences in these phenomena while controlling for their common aspects. The
previous research suggesting that RWA is significantly influenced by genetic factors whereas
differences in SDO are mainly due to environmental sources was based on self-reports, so
the current research extended that analysis by testing whether these patterns were
comparable across self- and peer reports. To the best of our knowledge, no study has done
this before. In addition, we used a latent phenotype modeling approach to control for
variance due to random error of measurement and nonrandom rater-specific perspectives.
To delve deeper into the genetic and environmental influences on RWA and SDO, the
NTFD was used to examine the extent to which individual differences in these traits were
attributable to genetic factors, environmental transmission from parents to offspring
(including separate estimates for mothers’ and fathers’ contributions), siblings’ shared
environmental influences beyond parental effects, passive genotype-environment
correlation, and individual-specific environmental influences, while taking assortative
mating of the twins’ parents into account. Since RWA can be expected to be substantially
heritable with only a small amount of its variance attributable to environmental influences
shared by twins (Bouchard, 2004; Ludeke & Krueger, 2013), we hypothesized that shared
environmental influences on the variance in RWA would be reduced when assortative
mating and passive genotype-environment correlation were brought into the analysis.
Method
Sample Characteristics
The complete sample was drawn from the Jena Twin Study of Social Attitudes
(JeTSSA; Stößel, Kämpfe, & Riemann, 2006). This sample encompasses 1437 target

10
10
Structure and Sources of Social Attitudes
individuals from 481 twin families (see Table 1). The sample included 394 complete twin
pairs: 48 male and 178 female MZ twin pairs and 20 male, 81 female, and 67 opposite sex
dizygotic (DZ) twin pairs. About 65% of the mothers and 52% of the fathers of twins
provided data. The sample was heterogeneous with regard to education and employment
status. Parts of the same data set were also analyzed in Kandler, Lewis et al. (2015) and in
Study 2 of Cohrs et al. (2012) for different purposes.
In addition to the self-raters, 1322 acquaintances (e.g., friends, spouses, or
colleagues) completed the peer-report versions of the RWA and SDO scales. The peers had
to be reasonably well-acquainted (at least one-year acquaintanceship: MYEARS = 11.48, 1-58
years), with preference given to those peer raters who knew one twin very well but not the
co-twin. For 748 twins (86%), at least one peer rater provided assessments, and for 574
twins (66%) two peer reports were available.
Measures
Twins, parents, and independent peer raters provided ratings on a 12-item Right-
Wing-Authoritarianism (RWA) scale (Altemeyer, 1996; Funke, 2005) and a 16-item Social
Dominance Orientation (SDO) measure (Pratto et al., 1994; Sidanius & Pratto, 1999). The
first RWA item (“People ought to pay attention to the Bible and other old traditional beliefs
instead of developing their own moral standards of ‘Good and Bad’.”) was excluded from
the analyses because of its very weak psychometric characteristics (e.g., low and even
negative correlations with other RWA items and with the RWA sum score). Since the RWA
measure was developed to consider RWA’s multidimensionality, four RWA items were
indicative of RWA-Aggression (RWA-A) and four pertained to RWA-Submission (RWA-S),
whereas the remaining three items were used to capture RWA-Conventionalism (RWA-C;
see Funke, 2002, 2005, for more details).

11
11
Structure and Sources of Social Attitudes
Similarly, the 16 SDO items can be assigned to the two SDO facets (see Ho et al.,
2012, 2015). The subscale SDO-Dominance (SDO-D) encompassed seven items, and
captured a preference for the suppression of other groups. Nine items were used to
measure SDO-Egalitarianism (SDO-E), which reflected a preference for nonegalitarian
intergroup relations. However, the differentiation between SDO-D and SDO-E is strongly
confounded with negative and positive item formulations in terms of social desirability, so
the possibility that the two SDO components were artificial was examined in the current
study. The internal consistency, peer consensus, and self-peer agreement for the RWA and
SDO sum scores and the three RWA subscale scores as well as the two SDO subscale scores
are shown in Table 2.
Analyses and Results
The Structure of Individual Differences in RWA and SDO
Principal axis factoring. We initially conducted a PAF analysis exploring the number
of potential latent dimensions that are necessary to describe the intercorrelation among the
27 items based on all available self-report data (N = 1437). We used promax rotation (k = 4)
to allow correlations between resulting dimensions. Both the Scree Plot (Cattell, 1966) and
Velicer’s Minimal Average Partial (MAP) test (O’Connor, 2000) yielded two dimensions
(accounting for 30.33% of individual differences in the items’ content) which were quite
orthogonal (r = .01). These two dimensions were interpretable as RWA and SDO (see Table
3). However, all seven SDO-D items showed positive loadings on the first dimension (i.e.
RWA), four of which were high, indicating that the SDO-D component reflected a common
aspect of RWA and SDO, which could account for the commonly-found positive correlation
between RWA and SDO sum scores in Western countries (r = .26 in the current study).

12
12
Structure and Sources of Social Attitudes
The first dimension was also indicated by high loadings on items confounded with
positive item formulations in terms of RWA, whereas the second dimension showed high
loadings on items confounded with negative item formulations regarding SDO. Thus, the
correlation between RWA and SDO scale scores might also be due to method artifacts, such
as individual differences in socially desirable responding. Therefore, we subsequently ran
CFAs taking this artificial aspect and the potential common aspect of RWA and SDO (i.e.,
domination over and aggression against subordinate groups) into account.
Confirmatory factor analyses. We ran a series of models allowing for two correlated
dimensions in terms of RWA and SDO with corresponding item loadings. The first model,
containing only the RWA and SDO dimensions, provided a poor fit to the data (χ² = 3800.81;
df = 323; RMSEA = .089; CFI = .783), as indicated by a root mean square error of
approximation (RMSEA) value > .08 and a comparative fit index (CFI) < .90 (Hu & Bentler,
1999; Steiger, 1990). In the next model we added a method factor accounting for artifact
variance due to positive item formulations pertaining to social desirability. It yielded a much
better fit (Δχ² = 2268.15; Δdf = 7; p = .00) and an acceptable global fit to the data (χ² =
1532.66; df = 310; RMSEA = .054; CFI = .917). Latent RWA and SDO scores were positively
interrelated (.42), indicating common substance beyond artificial shared variance due to
socially desirable responding. In a third model, illustrated in Figure 1a, we added blended
loadings from SDO-D items on the latent RWA factor thus incorporating the common
substantive aspect of RWA and SDO. It provided a significantly better model fit than the
second model (Δχ² = 109.82; Δdf = 7; p = .00; RMSEA = .052; CFI = .923). The standardized
factor loadings on RWA including the blended loadings were positive and ranged between
.05 and .73. The loadings on SDO ranged between .16 and .78, and the method factor
weights varied between .15 and .58. In this model, latent RWA and SDO scores were

13
13
Structure and Sources of Social Attitudes
uncorrelated (-.04), indicating that RWA and SDO scores reflected largely unrelated
individual characteristics when the common aspect pertaining to group-based dominance
(SDO-D) was controlled.
We then ran three alternative models which took the method factor and common
aspect into account, but which also tested the proposed multidimensionality of either RWA
or SDO or both. These multidimensional models had model fit statistics that were quite
close to those observed for the more parsimonious unidimensional model shown in Figure
1a. The first model (Figure 1b) included the three facets of RWA and allowed for blended
loadings from SDO-D items on the RWA dimension via its subordinate RWA-Aggression
facet. It provided an acceptable to good fit to the data (χ² = 1332.31; df = 300; RMSEA =
.050; CFI = .928). The substantive and artifact factor loadings as well as first and second
order factor loadings were positive. Latent RWA and SDO showed no significant association
(.02). Fixing the blended loadings from SDO-D items on RWA-Aggression to zero led to a
significant reduction in model fit (Δχ² = 123.34; Δdf = 7; p = .00; RMSEA = .052; CFI = .921)
and a positive correlation between RWA and SDO dimensions (r = .33). This again indicated
that RWA and SDO reflected largely unrelated individual characteristics when the common
aspect pertaining to group-based dominance was controlled.
A second alternative model, which allowed for the two facets of SDO (Figure 1c),
showed an acceptable model fit (χ² = 1516.83; df = 308; RMSEA = .054; CFI = .918). The
blended loading on RWA was .17 and the correlation between RWA and SDO was significant
(r = .25). Fixing the blended loading from the SDO-D facet on RWA to zero led to a significant
reduction in model fit (Δχ² = 21.56; Δdf = 1; p = .00; RMSEA = .054; CFI = .917) and increased
the correlation between RWA and SDO to .40.

14
14
Structure and Sources of Social Attitudes
The final alternative model, which allowed for five subordinate and two
superordinate factors (Figure 1d) provided an acceptable model fit to the data (χ² = 1436.06;
df = 305; RMSEA = .052; CFI = .922). The blended loading on RWA was .19 and the
correlation between RWA and SDO was significant (r = .17). Fixing the blended loading from
the SDO-D facet on RWA to zero led to a significant reduction in model fit (Δχ² = 24.12; Δdf =
1; p = .00; RMSEA = .053; CFI = .921) and increased the correlation between RWA and SDO
dimensions to .22, again indicating that the correlation between RWA and SDO scores was
lower when the common aspect was taken into account.
In sum, both the PAF analysis and CFAs suggested that modeling RWA and SDO as
multidimensional constructs did not provide an appreciably better model fit, indicating that
the two substantive dimensions comprising RWA and SDO were sufficient, at least for our
data and the measurement instruments used to capture RWA and SDO. In the next stage of
the analysis, the most parsimonious model (Figure 1a) was tested for measurement and
structural invariance across sexes, generations, and rater perspectives.
Sex differences. PAF analyses with promax rotation (k = 4) were conducted
separately for men and women. The Scree Plot and MAP test yielded two dimensions for
both males (accounting for 29.67% of the variance) and females (accounting for 30.82% of
the variance). These two dimensions were interpretable as RWA and SDO (see
supplementary Table S1). The two-dimensional structure was largely orthogonal and quite
comparable across men and women (r = -.01 for males and r = .01 for females). For both
men and women, about half of the SDO-D items showed higher loadings on the first
dimension (i.e. RWA) than the second (i.e. SDO).
A Multigroup Confirmatory Factor Analysis (MCFA) was conducted to test the most
parsimonious model presented in the previous section (see Figure 1a) for measurement and

15
15
Structure and Sources of Social Attitudes
structural invariance across men and women. The unconstrained model (in which none of
the parameters were set as equal) yielded an acceptable to good model fit (χ² = 1715.99; df
= 606; p = .00; RMSEA = .037; CFI = .907). Comparisons with more constrained models
yielded small but significant differences between males and females regarding factor
weights (Δχ² = 67.60; Δdf = 44; p = .01) but non-significant differences in the factor variance-
covariance structure (Δχ² = 2.31; Δdf = 4; p = .68). The measurement invariance across sexes
was primarily due to variation in method factor weights which were larger for women (.17
to .61, average: .37) than men (.14 to .54, average: .33). That is, social desirability tended to
be larger in women’s than men’s self-reports. Most importantly for the purposes of this
study, the link between latent RWA and SDO scores was not statistically significant for both
men and women.
Generational differences. We also compared the structure of social attitudes
between parents and their offspring based on PAF analyses with promax rotation (k = 4).
Because age 40 can mark an important turning point in life with respect to life tasks
(McAdams & Olson, 2010), we used this age as the cut-off to avoid an overlap in age
between generational groups. That is, all parents younger than 40 and all twins older than
39 were excluded from this analysis. The Scree Plot and MAP test yielded two dimensions
for both the offspring generation, aged between 17 and 39 (accounting for 30.32% of the
variance), and the parental generation, aged between 40 and 86 (accounting for 29.47% of
the variance). These two dimensions could be interpreted as RWA and SDO (see
supplementary Table S2). The dimensions were quite orthogonal (r = .05 for offspring and r
= -.01 for parents). Five of the SDO-D items showed higher loadings on the first dimension
(i.e. RWA) in the younger generation, whereas this was the case for only three of these
items in the older generation. However, the general structure of the two largely orthogonal

16
16
Structure and Sources of Social Attitudes
dimensions accounting for the variance and covariance in social attitudes was quite
comparable across generations.
To test measurement and structural invariance across generations, a MCFA was
performed and yielded an acceptable model fit for the unconstrained model (χ² = 1438.19;
df = 606; p = .00; RMSEA = .035; CFI = .903). There were significant differences in factor
weights between generations (Δχ² = 86.52; Δdf = 44; p = .00), but no significant differences
in the factor variance-covariance structure (Δχ² = 6.75; Δdf = 4; p = .15). The measurement
invariance was primarily due to the fact that the older generation tended to show larger
factor loadings on RWA (.06 to .79, average: .41) and smaller blended loadings from SDO-D
items on RWA (.02 to .35, average: .20) than the younger generation (RWA: .01 to .72,
average: .38; blended loadings: .09 to .46, average: .32). Again, allowing for blended
loadings, the dimensions RWA and SDO were neither significantly correlated in the younger
nor the older generation.
Rater perspective differences. Focusing on the invariance across rater perspectives,
we first analyzed the structure of social attitudes for both self- and averaged peer reports
based on PAF analyses with promax rotation (k = 4). Only the data from individuals with
available self-reports and at least one peer assessment were included in the analyses. If two
peer reports were provided, peer ratings were averaged across raters. The Scree Plot and
MAP test yielded two dimensions for both self-reports (accounting for 31.39% of the
variance) and peer reports (accounting for 29.47% of the variance). These two dimensions
could be construed as RWA and SDO (see supplementary Table S3). The dimensions were
largely orthogonal (r = .06 for self-reports and r = .16 for peer reports). More than half of the
SDO-D items showed higher loadings on the first dimension (i.e. RWA) for both self- and
mean peer reports.

17
17
Structure and Sources of Social Attitudes
To test for measurement and structural invariance across rater perspectives, we
conducted a MCFA and found that the unconstrained model yielded an acceptable model fit
(χ² = 1472.26; df = 606; p = .00; RMSEA = .035; CFI = .904). There were significant differences
between self- and averaged peer reports regarding factor weights (Δχ² = 80.86; Δdf = 44; p =
.00), but again no significant differences in the factor variance-covariance structure (Δχ² =
3.53; Δdf = 4; p = .47). Aggregated peer-rated items showed similar substantive factor
loadings (average: .45 for RWA, .40 for SDO, and .30 for the blended loadings) compared to
self-rated items (average: .43 for RWA, .42 for SDO, and .34 blended loadings), but tended
to show higher loadings on the method factor (.20 to .62, average: .41) than self-reports (.09
to .64, average: .34). The latter indicated that the higher internal consistency in averaged
peer reports was attributable to systematic measurement artifact. The dimensions RWA and
SDO were neither significantly correlated for self-ratings nor for averaged peer reports.
Summary. In the current study, models allowing for a hierarchical RWA and SDO
structure provided comparable model fits to the more parsimonious model allowing for two
basic dimensions RWA and SDO as well as blended loadings. In other words, only two
substantive dimensions were necessary to sufficiently describe individual differences in
RWA and SDO item contents of the used measures. The consistent zero correlation between
the two substantive factors indicated that RWA and SDO scores reflected largely unrelated
individual characteristics when the common aspect pertaining to group-based dominance
and method bias in terms of socially desirable responding were controlled.
The Sources of Individual Differences in RWA and SDO
Multi-rater twin modeling. Based on the results of the CFA and PAF analyses
reported above, for all behavior genetic aspects of the study we computed RWA and SDO
factor scores taking the blended loadings from SDO-D items on the RWA dimension into

18
18
Structure and Sources of Social Attitudes
account. This was done for all individuals’ self-reports, averaged peer reports, and
composites (i.e., aggregations of self- and peer reports).
Because age and sex differences can act to increase variance in RWA and SDO
affecting the covariance between two specific family members and thus biasing the
estimates of genetic and environmental variance components, age and sex effects were
taken into account. Older people showed higher levels of RWA (self-reports: β = .21, p = .00;
averaged peer reports: β = .10, p = .01) and SDO (self-reports: β = .06, p = .04; averaged peer
reports: β = .10, p = .01) than younger individuals. Females tended to show higher levels of
RWA (self-reports: β = .08, p = .01; averaged peer reports: β = -.03, p = .37) and lower levels
of SDO (self-reports: β = -.09, p = .00; averaged peer reports: β = -.12, p = .00) than males.
We used a regression technique to correct the factor scores for sex differences and age
effects (McGue & Bouchard, 1984). Subsequent analyses were based on standardized
residuals derived from these regressions.
Univariate twin model analyses were then conducted to decompose the variance in
RWA and SDO into three components attributable to: additive genetic influences (i.e.,
narrow-sense heritability, h²); environmental influences shared by twins (c²); and nonshared
environmental influences (e²) including error of measurement (see Neale & Maes, 2004, for
more details on twin modeling procedures). Additive genetic influences reflect the
combined effects of twins’ segregating genes, which are completely shared by MZ twins and
on average 50% shared by DZ twins. Larger MZ compared to DZ correlations indicate genetic
influences. Shared environmental influences (such as growing up in the same neighborhood)
reflect environmental circumstances that act to make twins similar, whereas nonshared
environmental effects (such as those arising from interactions with different peers or
spouses) reflect individual-specific influences (including random error of measurement) that

19
19
Structure and Sources of Social Attitudes
make twins dissimilar. This model assumes the absence of nonadditive genetic effects,
assortative mating, gene-environment correlation, and Gene × Environment interaction,
which cannot be estimated in the classical twin design.
The classical twin model was run for self-reports, averaged peer reports, and
composites. In addition, we ran multiple-rater twin model analyses with self- and averaged
peer reports as indicator variables of latent true scores. This structural equation model
allowed us to decompose variance in twins’ self-reports that was common with averaged
peer reports, thus correcting for unsystematic measurement error variance and systematic
variance due to specific self- or peer rater perspectives (see Nelling, Kandler, & Riemann,
2015, for more details). The results are shown in Table 4.
Peer reports provided lower twin correlations and lower estimates of heritability in
the case of RWA than self-reports. However, despite variation in the value of the variance
components, the different rater perspectives provided quite consistent results regarding the
contribution of genetic and environmental influences on RWA and SDO: For all rater
perspectives individual differences in RWA were partially heritable, whereas for all
perspectives the variance in SDO was not significantly influenced by genetic differences
(recall that the common aspect with RWA has been controlled). Thus, the patterns reported
in previous research in which heritability for SDO was substantially lower than those for
RWA held across the reviewer perspectives examined in this study.
Individual differences in both RWA and SDO were influenced by shared
environmental and nonshared environmental factors. Moreover, for both constructs
environmental influences shared by twins turned out to be more important than
environmental influences not shared by twins when random error and nonrandom method
variance were controlled (see results for true scores).

20
20
Structure and Sources of Social Attitudes
Nuclear twin family modeling. Estimates of shared environmental influences on
RWA and SDO might reflect true shared familial influences, such as common parental
influences that act to make twins more similar. However, estimates of shared
environmental influences may also reflect genetically driven assortative mating by the
twins’ parents on the phenotypes of interest (Kandler et al., 2012; Watson et al., 2014).
Thus, as mentioned earlier, environmental influences shared by family members could be
overestimated if assortative mating were not controlled.
The NTFD takes the spousal similarity of twins’ parents on the measured phenotypes
into account. This design also allows for a disentanglement of twins’ shared environmental
effects due to parental influences from shared environmental influences not provided by
parents, such as common peer influences. In addition, the NTFD design allows estimates of
the contribution of passive genotype-environment correlation. Stated more technically, the
model incorporating twins and their parents (see Figure 2 and Table 5) allows for estimates
of genetic contributions to the variance (h²) as well as different environmental
contributions. The latter components include influences from mothers (m²), fathers (f²), and
both (2mfμ), as well as environmental influences shared by twins beyond parental
influences (s²), and environmental influences that are not shared by twins (e²; including
error of measurement). The model also allows for an estimation of the contribution of the
covariance between parents’ genetic makeup and offspring’s shared parental environments,
i.e. passive genotype-environment correlation as discussed above: h²m(1 + μ) + h²f(1 + μ). As
can be seen in Table 5, the shared genetic influences accounting for parent-child covariance
and DZ twin covariance takes assortative mating (μ) into account (½h² [1 + μ]).
The correlations between twins’ parents were r = .45 for RWA and r = .42 for SDO.
The parent-child correlations ranged between r = .23 (father and male child) and r = .42

21
21
Structure and Sources of Social Attitudes
(mother and female child) for RWA and between r = .09 (father and male child) and r = .32
(father and female child) for SDO. Thus, the average parent-child correlations tended to be
lower than the DZ twin correlations (see Table 4), indicating shared environmental effects
that do not arise through interactions with parents.
The results of the NTFD modeling are depicted in Table 6 and Figure 3. The NTFD
models provided at least acceptable model fit, as indicated by the RMSEA and CFI. We
provided model parameter estimates for full and reduced (more parsimonious) models,
where non-significant parameters derived from the full model were fixed to zero. The model
fitting results yielded significant genetic effects (h) contributing to the parent-child and
siblings’ similarity for RWA but not for SDO. Beyond the genetic contributions to the parent-
offspring similarity, the environmentally mediated maternal effects (m) were important for
both RWA and SDO, whereas the results indicated paternal effects (f) only in case of SDO.
This was somewhat surprising, given that authoritarianism is often thought of as a
predominantly male trait, although it does make sense if one considers that mothers usually
spend more time interacting with their children than fathers. Environmental factors shared
by twins primarily manifested via sibling-specific shared effects (s), i.e. environmental
factors not associated with parenting. The remaining effects were attributable to nonshared
environmental effects including error of measurement.
Figure 3 also shows (in parentheses) estimates of genetic and environmental
influences corrected for attenuation due to error of measurement, estimated as the
difference between MZ twin true score correlation and MZ twin self-reports correlation (see
Table 4): ε² = .88 – .70 = .18 for RWA and ε² = .66 – .35 = .31 for SDO. Corrected for
attenuation, about 70% of individual differences in RWA were attributable to genetic
involvement in terms of genetic effects (54%) and passive genotype-environment

22
22
Structure and Sources of Social Attitudes
correlation (16%), whereas 30% was due to environmental factors, mainly non-parental
influences shared by twins. For SDO, about 50% of individual differences were due to shared
environmental influences (in particular beyond parental influences) and the other 50% were
due to environmental factors not shared by twins.
Discussion
The findings of the current study on the structure and sources of individual
differences in RWA and SDO have several implications. These implications pertain to the
utility of RWA and SDO as basic dimensions of individual differences in social attitudes, the
meaning of their association, and the interpretations of their underlying genetic and
environmental sources.
The Structure of Individual Differences in RWA and SDO
Our study indicates that only two substantive dimensions, interpretable as RWA and
SDO, are sufficient to describe individual differences in the content of RWA and SDO items.
This two-dimensional factor solution showed structural invariance across sexes,
generations, and rater perspectives. According to previous research, specific facets of RWA
(authoritarian aggression, submission, and conventionalism) and SDO (preference for
ingroup superiority and intergroup inequality) can be modelled in addition to the two core
dimensions, and previous studies have shown that these facets may be important with
respect to predicting specific social attitudes and prejudices (e.g., Duckitt & Bizumic, 2013;
Ho et al., 2012). However, our research indicates that at a descriptive and abstract level,
RWA and SDO conceived of non-hierarchically are attitudinal dimensions that adequately
describe individual differences in core preferences that organize a body of social attitudes,
at least when using the measures employed in the current study (see the limitation section
for a discussion on measurement issues).

23
23
Structure and Sources of Social Attitudes
Our data indicated that the correlation between RWA and SDO was reduced and
even became negligible when common substance (i.e., aggression against subordinate
groups) and measurement artifact (i.e., positive response in terms of socially desirable
responding) were taken into account. This correlative independence was invariant across
sexes, generations, and rater perspectives. Against the background of international variation
in the links between authoritarianism and social dominance orientation reported in earlier
studies – that is, positive associations in Western, individualistic countries and zero
correlations in Eastern, collectivistic countries (Duriez et al., 2005; Kandler, Bell et al., 2015)
– this result is remarkable. Our study was based on data gathered in Germany, an
unambiguously Western, educated, industrialized, rich, and democratic (WEIRD) society
(Henrich, Heine, & Norenzayan, 2010).
It is conceivable that authoritarian people in economically developed and
individualistic countries tend to prefer the superiority of their own group and dominance
against other groups. According to the dual process motivational (DPM) model of ideology
(Duckitt & Sibley, 2010), RWA reflects a “dangerous world” orientation to social and political
affairs, whereas SDO reflects a “competitive jungle” worldview. In economically rich
countries, authoritarian people may perceive other groups (e.g., immigrants from
economically underdeveloped countries or religious and ethnic minorities) as a threat to
economic stability and cultural tradition, and thus as a threat to societal preservation and
security. This may give rise to support for initiatives that seek to maintain cultural stability
and defend existing economic arrangements. The former may be reflected in the social
conservatism that characterizes RWA, while the latter would be something that would
appeal to authoritarians for security reasons but would also be supported by people high in
SDO who want to exert their dominance over subordinate economic groups. It is

24
24
Structure and Sources of Social Attitudes
conceivable that individual differences in SDO are not as extensive in collectivistic and less
competitive countries because of a perception of greater social and economic equality. This
restriction in the amount of variance would reduce the covariation and thus the correlation
between RWA and SDO in those countries. Further research is needed to test that
hypothesis.
Another, related issue that should be considered when trying to make sense of the
relationship between RWA and SDO in different countries pertains to the levels of religious
and ethnic homogeneity found in the countries or regions from which the samples were
drawn. Often when people think about dominance over “other groups” or how “inferior
groups” ought to be treated (see SDO-D items, Table 3), religious and ethnic differences
come to mind. For example, Poland and Japan, which have low or non-existent RWA-SDO
correlations, are quite homogeneous on those characteristics. Perhaps in such countries
issues pertaining to religious and ethnic interaction and conflict are less salient and not tied
to traditionalism and cultural conservatism, among both the populace in general and
political elites, unlike in more heterogeneous societies. That would reduce, for
authoritarians, the need to suppress religious and the ethnic minorities, and thus eliminate
a form of dominance over other groups that authoritarians in other contexts share with
people high in SDO. It would also reduce the RWA-SDO correlation. More cross-cultural
research, acquired across substantial time periods, is required to shed further light on the
influence of religious and ethnic divisions on the links between RWA and SDO as well as on
the extent of individual differences in SDO.
The Sources of Individual Differences in RWA and SDO
Our data provide strong support for the idea that, at least among our participants,
individual differences in SDO, when controlling for the common variance with RWA, were

25
25
Structure and Sources of Social Attitudes
driven almost exclusively by environmental factors. The results are in line with previous
genetically informative studies (e.g., Orey & Park, 2012) and experimental studies (e.g.,
Lehmiller & Schmidt, 2007) which have shown that SDO appears to be more context-
sensitive than RWA and is primarily influenced by individual-specific environmental stimuli.
However, all previous studies on that topic relied on self-reports, and single-rater studies
cannot disentangle individual environmental influences from error of measurement. Our
multiple-rater study controlled for error of measurement and found that about half the
variance in SDO was attributable to environmental influences shared by twins, with the
other half coming from individual-specific environmental factors.
Although in the absence of relevant data it is plausible to assume that shared
environmental influences on SDO (which by definition increase twins’ similarity) primarily
reflect shared parental effects, our nuclear twin family study indicates that only about 25%
of the variance explained by shared environmental factors are attributable to parental
influences. The more powerful, non-parental effects may reflect intra-generational shared
social contexts such as those stemming from increasing identification with peers, cliques,
and generation-specific idols. Those influences appear to be important not only in the adult
years but even when children begin to attend primary school, which marks the beginning of
a trend toward independence from parents and the development of the agentic self, which
allows one to articulate one’s own goals, values, and attitudes (Kandler et al., 2016;
McAdams, 2015).
However, that there was only a small shared environmental component arising from
common parental influences does not mean that parental effects on their offspring’s social
dominance orientation were marginal. On the contrary, trait-relevant behaviors exhibited by
parents that are observed by both twins may not have the same effect on each twin.

26
26
Structure and Sources of Social Attitudes
Siblings may experience, perceive, or interpret the same parental influence differently
(Plomin, Asbury, & Dunn, 2001), leading to differences in social attitudes. Those
environmental influences would, in effect, not be shared by individuals raised in the same
family. That is, parental influences may arise as unique environmental influences that act to
make siblings different from each other.
Individual environmental influences on the variance in SDO may also reflect extra-
familial effects, such as having different friends. Peers represent important non-familial
influences on social attitudes that may or may not be shared by twins – twins can have
different friends and typically have different spouses. Nonshared environmental influences
may also reflect individual events. Experimental studies, for instance, have shown that
perceived social and economic threats can increase social intolerance and negative attitudes
toward competing groups (Riek, Mania, & Gaertner, 2006) as well as authoritarianism
(McCann, 2008). That is, individuals may adapt their attitudes to individual life experiences.
Genetic influences do not appear to contribute significantly to individual differences
in SDO beyond the aspects common with RWA, but genetic factors account for a substantial
component in the latter. Genetic influences primarily explained twin pairs’ resemblance in
RWA. Environmental influences that act to increase twin pairs’ resemblance were reduced,
when – as hypothesized – assortative mating and passive genotype-environment correlation
were taken into account. But why are there genetic effects on RWA but not SDO? Beyond
genetically driven assortative mating that acts to maintain the genetic variance in a
population, it is conceivable that core personality traits, which show strong genetic
influences, can account for proportions of the genetic variance in RWA (Kandler,
Zimmermann, & McAdams, 2014). Openness and Conscientiousness, for example, turned
out to be robust predictors of RWA (Duckitt & Sibley, 2010; Perry & Sibley, 2012; Sibley &

27
27
Structure and Sources of Social Attitudes
Duckitt, 2010). About a half of the genetic variance in RWA appears to be shared with
personality traits (Lewis & Bates, 2014). One can think about other linked and heritable
attributes which contribute to individual differences in RWA to account for its strong genetic
component, such as verbal intelligence (Heaven, Ciarrochi, & Leeson, 2011) and differences
in core human motives, such as the needs for order and safety (Jost, Kruglanski, Glaser, &
Sulloway, 2003; Jugert & Duckitt, 2009).
It might also be that authoritarianism reflects a distinct characteristic of a broad
personality spectrum that is systematically related to other personality characteristics, such
as traits, abilities, and motives, but still has its own and unique genetic basis. Consistent
with this idea, Alford and Hibbing (2007) have differentiated between personal (or
individual) attributes such as closed-mindedness, which pertains to the rigidity toward
individual adaptations to particular environments, and political (or social) characteristics,
such as authoritarian conservatism, which pertains to the resistance to social and economic
change in a group or nation (see also Hibbing, Smith, & Alford, 2014).
Both RWA and SDO address the question of how the society should be organized
and work (e.g., social hierarchy, loyalty to authorities, and domination of other groups).
However, the specificity of the SDO dimension primarily reflects individual differences in the
preference for social and economic inequality versus equality. Despite individual
differences, children may intuitively sense what is good or bad for the society through
naturally evolved social instincts (Haidt, 2012; McAdams, 2015), but they first have to learn
how the society works (e.g., individualistic vs. collectivistic) and which position they have in
it (e.g., minority vs. majority ethnicity or higher vs. lower SES), which may affect the
development of their attitudes toward intergroup inequality. This might explain why SDO is
largely context-sensitive and may vary across time and cultures.

28
28
Structure and Sources of Social Attitudes
Controlling for assortative mating and error variance, the environmental influences
on individual differences in RWA turned out to be comparatively small. In particular, the
environmentally mediated parent-child transmission, which primarily ran from the mother
to the offspring, was marginal. However, these findings again do not mean that parental
influences on children’s authoritarianism are negligible. The current study found strong
evidence for the passive type of genotype-environment correlation. That is, genetic effects
on RWA is partly linked to environmental parent-offspring transmission. Politically
authoritarian parents may have an authoritarian parenting style – high levels of control and
disciplinary efforts, but low levels of warmth and support (Baumrind, 1991) – which
matches their children’s genetically predisposed authoritarianism. This is in line with
previous research which has shown that children raised by parents who prefer an
authoritarian-directive parenting style tend to conform easily and show more aggressive
behavior outside the home (Baumrind, 1991). Thus, parental authoritarianism, which is
substantially similar between fathers and mothers, has been partially genetically
transmitted from parents to their offspring and their preferred authoritarian parenting style
may further enhance their offspring’s authoritarianism. Future genetically informative
studies with measured family environments, such as parenting styles, may shed further light
on the nature of those mechanisms.
Limitations and Future Directions
The current genetically informative multi-method study has several strengths that
extend previous research on RWA and SDO. However, there are some limitations and
possibilities for future research that should be mentioned. First, there was an
overrepresentation of women relative to men and MZ relative to DZ twins in the sample.

29
29
Structure and Sources of Social Attitudes
Accordingly, the results based on our study have to be replicated in future investigations
with more balanced and more representative samples.
Second, our study used a rather short measure of RWA. Even though this measure
was previously used to establish the three facets of RWA – aggression, submission, and
conventionalism (Funke, 2002, 2005) – other studies suggest that longer measures of RWA
are needed to adequately portray these facets. For example, Duckitt and Bizumic (2013)
used a 36-item measure that employed twelve items to capture each facet with good
reliability. The short measure may be adequate to capture RWA as core dimension, but an
accurate investigation on the level of its specific facets should include broader measures.
Third, we used the classical SDO measure with unbalanced pro- and contra-trait item
formulations. The pro-trait formulations are largely confounded with the SDO-Dominance
facet, whereas the contra-trait formulations are largely confounded with the SDO-
Egalitarianism facet. Even though Ho et al. (2012) provided support for the predictive
specificity and discriminant validity of both SDO facets, we recommend a measure with a
balanced set of pro- and contra-trait item formulations for future investigations.
Fourth, we found that the substantial heritability of RWA reflects genetic factors
which partly co-vary with environmentally mediated parental influences via passive
genotype-environment correlation. However, other types of genotype-environment
interplay are also plausible. Genetic factors may influence the kinds of environments that
individuals avoid or choose to be in, leading to variation in self-selected environments (e.g.,
specific friends, religious communities, and political parties), which in turn may affect
individual differences in authoritarian attitudes (i.e., there may be active genotype-
environment correlation). Individuals may also differently evoke reactions from their social
environment that match their genetically predisposed authoritarian orientations. For

30
30
Structure and Sources of Social Attitudes
example, liberal and non-submissive people may evoke authoritarian responses from
conservatives that make them even more liberal (i.e., evocative genotype-environment
correlation may exist). Such effects may act to increase individual differences in RWA and,
thus, the difference between MZ and DZ twin similarity as a function of their genetic
relatedness. Consequently, estimates of heritability of RWA can also reflect effects due to
active and evocative genotype-environment interplay to some degree.
Conclusions
The current study provides empirical support for two broad dimensions which can be
explicitly interpreted as RWA and SDO. This two-dimensional structure is sufficient to
capture the differences in RWA and SDO item content. Taking the common substantive
aspect of RWA and SDO (i.e., domination over or aggression against subordinate groups)
and artifact (method variance due to pro- or contra-trait item formulations) into account,
RWA and SDO are largely orthogonal dimensions. The orthogonality of the two core
dimensions is structurally invariant across men and women, two generations, and self- and
peer reports.
Controlling for random and nonrandom error of measurement, genetic and
environmental sources contribute differently to RWA and SDO. Whereas individual
differences in RWA are primarily genetically influenced, individual differences in the unique
aspects of SDO are almost exclusively attributable to environmental influences. Assortative
mating may be one driving force that acts to maintain the genetic variance in RWA.
Moreover, genetic factors are associated to some degree with shared familial environments.
That is, genetic and environmental transmissions from parents to their offspring are
correlated to some degree. This passive form of genotype-environment correlation acts to
increase individual differences and the similarity of twins in RWA. With regard to the strong

31
31
Structure and Sources of Social Attitudes
environmental influences on the variance in SDO, one half of it turned out to be
environmental effects shared by twins. These influences primarily reflect factors beyond
parental influences. The other half is attributable to individual-specific environmental
influences.
In sum, the current extended twin family and multi-method study provides new
insights into the correlative and hierarchical structure of RWA and SDO as fundamental
dimensions of individual differences in social attitudes. It further reveals the utility of adding
peer reports on social attitudes to self-reports, and the need for a nuclear twin family design
to analyze the genetic and environmental sources of individual differences in social
attitudes.

32
32
Structure and Sources of Social Attitudes
References
Adorno, T., Frenkel-Brunswick, E., Levinson, D., & Sanford, N. (1950). The authoritarian
personality. New York: Harper.
Alford, J. R., Funk, C. L., & Hibbing, J. R. (2005). Are political orientations genetically
transmitted? American Political Science Review, 99, 153-167.
Alford, J. R., & Hibbing, J. R. (2007). Personal, interpersonal, and political temperaments. The
ANNALS of the American Academy of Political and Social Science, 614, 196-212.
Altemeyer, B. (1981). Right-wing authoritarianism. Winnipeg University of Manitoba Press,
Canada: Manitoba.
Altemeyer, B. (1996). The authoritarian specter. Harvard University Press: Cambridge.
Aspelund, A., Lindeman, M., & Verkasalo, M. (2013). Political conservatism and left-right
political orientation in 28 Eastern and Western European countries. Political
Psychology, 34, 409-417.
Baumrind, D. (1991). The influence of parenting style on adolescent competence and
substance use. Journal of Early Adolescence, 11, 56-95.
Bouchard, T. J. Jr. (2004). Genetic influence on human psychological traits: A survey. Current
Directions in Psychological Science, 13, 148-151.
Bouchard, T. J. Jr. (2009). Authoritarianism, religiousness, and conservatism: Is ‘‘obedience
to authority’’ the explanation for their clustering, universality and evolution? In: E.
Voland, W. Schiefenhövel (Eds.), The biological evolution of religious mind and
behavior (pp. 165-180). Berlin: Springer.
Bouchard, T. J. Jr., & McGue, M. (2003). Genetic and environmental differences on human
psychological differences. Journal of Neurobiology, 54, 4-45.

33
33
Structure and Sources of Social Attitudes
Brandt, M. J., & Henry, P. J. (2012). Gender inequality and gender differences in
authoritarianism. Personality and Social Psychology Bulletin, 38, 1301-1315.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen
& J. S. Long (Eds.), Testing structural equation models (pp. 136−162). Newbury Park,
CA: Sage.
Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral
Research, 1, 245-276.
Cohrs, J. C., Kämpfe-Hargrave, N., & Riemann, R. (2012). Individual differences in ideological
attitudes and prejudice: Evidence from peer-report data. Journal of Personality and
Social Psychology, 103, 343-361.
Cohrs, J. C., Moschner, B., Maes, J., & Kielmann, S. (2005). The motivational bases of right-
wing authoritarianism and social dominance orientation: Relations to values and
attitudes in the aftermath of September 11, 2001. Personality and Social Psychology
Bulletin, 31, 1425-1434.
Duckitt, J., & Bizumic, B. (2013). Multidimensionality of right-wing authoritarian attitudes:
authoritarianism-conservatism-traditionalism. Political Psychology, 34, 841-862.
Duckitt, J., Bizumic, B., Krauss, S., & Heled, E. (2010). A tripartite approach to Right-Wing
Authoritarianism: The Authoritarianism-Conservatism-Traditionalism model. Political
Psychology, 31, 685-715.
Duckitt, J., & Sibley, C. G. (2010). Personality, ideology, prejudice and politics: A dual process
motivational model. Journal of Personality, 78, 1861-1893.
Duriez, B., & Van Hiel, A. (2002). The march of modern fascism: A comparison of social
dominance orientation and authoritarianism. Personality and Individual Differences,
32, 1199-1213.

34
34
Structure and Sources of Social Attitudes
Duriez, B., Van Hiel, A., & Kossowska, M. (2005). Authoritarianism and social dominance in
Western and Eastern Europe: The importance of the sociopolitical context and of
political interest and involvement. Political Psychology, 26, 299-320.
Feather, N. T. (1977). Generational and sex differences in conservatism. Australian
Psychologist, 12, 76-82.
Fromm, E. (1941). Escape from freedom. New York: Farrar & Rinehart.
Funk, C. L., Smith, K. B., Alford, J. R., Hibbing, M. V., Eaton, N. R., Krueger, R. F., Eaves, L. J., &
Hibbing, J. R. (2013). Genetic and environmental transmission of political orientations.
Political Psychology, 34, 805-819.
Funke, F. (2002). »Three pairs of eyes are better than one« or »Too many cooks spoil the
broth«? - Validating a three-dimensional model of right-wing authoritarianism. Paper
presented at the 25th Annual Meeting of the International Society of Political
Psychology. Berlin 16th July, 2002.
Funke, F. (2005). The dimensionality of right-wing authoritarianism: Lessons from the
dilemma between theory and measurement. Political Psychology, 26, 195-218.
Haidt, J. (2012). The righteous mind: Why good people are divided by politics and religion.
New York: Pantheon.
Heaven, P. C. L., Ciarrochi, J., & Leeson, P. (2011). Cognitive ability, right-wing
authoritarianism, and social dominance orientation: a five-year longitudinal study
amongst adolescents. Intelligence, 39, 15-21.
Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world?
Behavioral and Brain Sciences, 33, 61-83.
Hibbing, J. R., Smith, K. B., & Alford, J. R. (2014). Predisposed: Liberals, conservatives, and
the biology of political differences. New York and London: Routledge.

35
35
Structure and Sources of Social Attitudes
Ho, A. K., Sidanius, J., Kteily, N., Sheehy-Skeffington, J., Pratto, F., Henkel, K. E., Foels, R., &
Stewart, A. L. (2015). The nature of social dominance orientation: Theorizing and
measuring preferences for intergroup inequality using the new SDO7 scale. Journal of
Personality and Social Psychology, 109, 1003-1028.
Ho, A. K., Sidanius, J., Pratto, F., Levin, S., Thomson, L., Kteily, N., & Sheehy-Skeffington, J.
(2012). Social dominance orientation: Revisiting the structure and function of a
variable predicting social and political attitudes. Personality and Social Psychology
Bulletin, 38, 583-606.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.
Jost, J. T., Kruglanski, A. W., Glaser, J., & Sulloway, F. J. (2003). Political conservatism as
motivated social cognition. Psychological Bulletin, 129, 339-375.
Jost, J. T., & Thompson, E. P. (2000). Group-based dominance and opposition to equality as
independent predictors of self-esteem, ethnocentrism, and social policy attitudes
among African Americans and European Americans. Journal of Experimental Social
Psychology, 36, 209-232.
Jugert, P., & Duckitt, J. (2009). A motivational model of authoritarianism: Integrating
personal and situational determinants. Political Psychology, 30, 693-719.
Kandler, C., Bell, E., Shikishima, C., Yamagata, S., & Riemann, R. (2015). Genetic foundations
of attitude formation. In: R. Scott & S. Kosslyn (eds.), Emerging Trends in the Social
and Behavioral Sciences. John Wiley & Sons, Inc.
Kandler, C., Bleidorn, W., & Riemann, R. (2012). Left or right? Sources of political
orientation: the roles of genetic factors, cultural transmission, assortative mating, and
personality. Journal of Personality and Social Psychology, 102, 633-645.

36
36
Structure and Sources of Social Attitudes
Kandler, C., Gottschling, J., & Spinath, F. M. (2016). Genetic and environmental parent-child
transmission of value orientations: An extended twin family study. Child Development,
78, 270-284.
Kandler, C., Lewis, G. J., Feldhaus, L. H., & Riemann, R. (2015). The genetic and
environmental roots of variance in negativity toward foreign nationals. Behavior
Genetics, 45, 181-199.
Kandler, C., Zimmermann, J., & McAdams, D. P. (2014). Core and surface characteristics for
the description and theory of personality differences and development. European
Journal of Personality, 28, 231-243.
Keller, M. C., Medland, S. E., & Duncan, L. E. (2010). Are extended twin family designs worth
the trouble? A comparison of the bias, precision, and accuracy of parameters
estimated in four twin family models. Behavior Genetics, 4, 377-393.
Kugler, M. B., Cooper, J., & Nosek, B. A. (2010). Group-based dominance and opposition to
equality correspond to different psychological motives. Social Justice Research, 23,
117-155.
Lehmiller, J. J., & Schmitt, M. T. (2007). Group domination and inequality in context:
evidence for the unstable meanings of social dominance and authoritarianism.
European Journal of Social Psychology, 37, 704-724.
Lewis, G. J., & Bates, T. C. (2014). Common heritable effects underpin concerns over norm
maintenance and in-group favoritism: Evidence from genetic analyses of right-wing
authoritarianism and traditionalism. Journal of Personality, 82, 297-309.
Ludeke, S. G., Johnson, W., & Bouchard, T. J. Jr. (2013). ‘‘Obedience to traditional
authority:’’ A heritable factor underlying authoritarianism, conservatism and
religiousness. Personality and Individual Differences, 55, 375-380.

37
37
Structure and Sources of Social Attitudes
Ludeke, S. G., & Krueger, R. F. (2013). Authoritarianism as a personality trait: Evidence from
a longitudinal behavior genetic study. Personality and Individual Differences, 55, 480-
484.
McAdams, D. P. (2015). The art and science of personality development. New York: Guilford.
McAdams, D. P., & Olson, B. D. (2010). Personality development: Continuity and change
over the life course. Annual Review of Psychology, 61, 5.1-5.26.
McCann, S. J. H. (2008). Societal threat, authoritarianism, conservatism, and U.S. state death
penalty sentencing (1977–2004). Journal of Personality and Social Psychology, 94, 913-
923.
McCourt, K., Bouchard, T. J. Jr., Lykken, D. T., Tellegen, A., & Keyes, M. (1999).
Authoritarianism revisited: Genetic and environmental influence examined in twins
reared apart and together. Personality and Individual Differences, 27, 985-1014.
McGue, M., & Bouchard, T. J., Jr. (1984). Adjustment of twin data for the effects of age and
sex. Behavior Genetics, 14, 325–343.
Neale, M. C., & Maes, H. H. M. (2004). Methodology for genetic studies of twins and
families. Dordrecht, the Netherlands: Kluwer Academic Publishers.
Nelling, A., Kandler, C., & Riemann, R. (2015). Substance and artifact in interest self-reports:
A multiple-rater twin study. European Journal of Psychological Assessment, 31, 166-
173.
O’Connor, B. P. (2000). SPSS and SAS programs for determining the number of components
using parallel analysis and Velicer’s MAP test. Behavior Research Methods,
Instruments and Computers, 32, 396-402.
Olson, J. M., Vernon, P. A., Harris, J. A., & Jang, K. L. (2001). The heritability of attitudes: a
study of twins. Journal of Personality and Social Psychology, 80, 845-860.

38
38
Structure and Sources of Social Attitudes
Orey, B. D. A., & Park, H. (2012). Nature, nurture, and ethnocentrism in the Minnesota Twin
Study. Twin Research and Human Genetics, 15, 71-73.
Perry, R., & Sibley, C. G. (2012). Big-five personality prospectively predicts social dominance
orientation and right-wing authoritarianism. Personality and Individual Differences, 52,
3-8.
Plomin, R., Asbury, K., & Dunn, J. (2001). Why are children in the same family so different?
Nonshared environment a decade later. Canadian Journal of Psychiatry, 46, 225-233.
Pratto, F., Sidanius, J., Stallworth, L., & Malle, B. (1994). Social dominance orientation: A
personality variable predicting social and political attitudes. Journal of Personality and
Social Psychology, 67, 741-763.
Riek, B. M., Mania, E. W., & Gaertner, S. L. (2006). Intergroup threat and outgroup attitudes:
A meta-analytic review. Personality and Social Psychology Review, 10, 336-353.
Scarr, S., & McCartney, K. (1983). How people make their own environments: A theory of
genotype–environment effects. Child Development, 54, 424-435.
Sibley, C. G., & Duckitt, J. (2010). The personality bases of ideology: a one-year longitudinal
study. Journal of Social Psychology, 150, 540-559.
Sibley, C. G., Robertson, A., & Wilson, M. S. (2006). Social dominance orientation and right-
wing authoritarianism: Additive and interactive effects. Political Psychology, 27, 755-
768.
Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social hierarchy
and oppression. New York, NY: Cambridge University Press.
Sidanius, J., Pratto, F., & Bobo, L. (1994). Social dominance orientation and the political
psychology of gender: A case of invariance? Journal of Personality and Social
Psychology, 67, 998-1011.

39
39
Structure and Sources of Social Attitudes
Steiger, J. H. (1990). Structural model evaluation and modification: An interval estimation
approach. Multivariate Behavioral Research, 25, 173-180.
Stößel, K., Kämpfe, N., & Riemann, R. (2006). The Jena Twin Registry and the Jena Twin
Study of Social Attitudes (JeTSSA). Twin Research and Human Genetics, 9, 783-786.
Van Hiel, A., Cornelis, I., Roets, A., & De Clrecq (2007). A comparison of various
authoritarianism scales in Belgium Flanders. European Journal of Personality, 21, 149-
168.
Watson, D., Klohnen, E. C., Castillas, A., Simms, E. N., Haig, J., & Berry, D. S. (2004). Match
makers and deal breakers: Analyses of assortative mating in newlywed couples.
Journal of Personality, 72, 1029-1068.
Watson, D., Beer, A., & McDade-Montez, E. (2014). The role of active assortment in spousal
similarity. Journal of Personality, 82, 116-129.

40
40
Structure and Sources of Social Attitudes
Table 1. JeTSSA Sample Characteristics
Number
Age
Total
Male
Female
Pairs
M
SD
Range
MZ twins
452
96
356
226
33.98
14.41
17-82
DZ twins
336
107
229
168
34.10
13.06
18-73
Single twins
87
24
63
36.78
11.25
19-68
Mothers
313
313
56.50
10.64
39-84
Fathers
249
249
58.37
10.35
39-86
All
1437
476
961
394
43.37
16.83
17-86
Note. MZ: Monozygotic; DZ: Dizygotic.

41
41
Structure and Sources of Social Attitudes
Table 2. Psychometric Quality of Original Scale Scores
Variables
RWA
RWA-A
RWA-S
RWA-C
SDO
SDO-D
SDO-E
N items
11
4
4
3
16
7
9
α self-reports
.73
.58
.54
.44
.79
.69
.77
α peer reportsa
.75
.62
.55
.44
.82
.72
.80
α Øpeer reports
.79
.68
.60
.52
.83
.72
.82
r peer consensus
.54
.45
.37
.40
.28
.22
.27
r self-peer agreementa
.52
.43
.36
.31
.26
.19
.30
r self-Øpeer agreement
.57
.53
.38
.33
.30
.21
.35
Note. Øpeer: average of two peer reports; aone peer report was randomly selected; rater
agreement was estimated with product moment correlation; all correlations were
significant (p < .001).

42
42
Structure and Sources of Social Attitudes
Table 3. Factor Loadings of RWA and SDO Items on the First Two Factors Based on Principal
Axis Factoring (PAF) and on Modeled RWA and SDO Factors Based on Confirmatory Factor
Analyses (CFA)
Social attitudes
PAF
CFA
I.
II.
I.
II.
RWA-Aggression (A), RWA-Submission (S), and RWA-Conventionalism (C):
(1) A: Instead of more civil rights, we need an upholding of
law and order.
.63
-.06
.62
.00
(2) S: The days when women are submissive to their
husbands and follow similar social conventions belong
strictly in the past. A “woman’s place” should be wherever
she wants to be.(-)
.12
.19
.05
.00
(3) C: Turning away from tradition will emerge as fatal error
one day.
.30
.05
.29
.00
(4) A: There is no crime that legitimates the death penalty.(-)
.26
.07
.25
.00
(5) S: Obedience and respect for authority are the most
important virtues that children should learn.
.69
-.07
.67
.00
(6) C: Homosexual cohabitations should be put on a par with
marriage.(-)
.29
.17
.35
.00
(7) A: What our country really needs is a strong, determined
leader who will crush evil, and take us back to our true path.
.64
-.06
.60
.00
(8) S: It is wonderful that young people today have greater
freedom to protest against things they don’t like and to do
their own thing.(-)
.17
.15
.18
.00
(9) C: In the long run, virtuousness and law-abiding behavior
will bring us forward, not the permanent challenge of our
society’s foundations.
.55
-.03
.59
.00
(10) A: It is important to protect fully the rights of radicals
and deviants.(-)
.19
.03
.33
.00

43
43
Structure and Sources of Social Attitudes
(11) S: The real keys to a good life are obedience, discipline,
and sticking to the old and narrow.
.73
-.06
.73
.00
SDO-Dominance (D) and SDO-Egalitarianism (E):
(12) E: We should strive to make incomes as equal as
possible. (-)
-.33
.51
.00
.41
(13) E: Group equality should be our ideal.(-)
-.32
.44
.00
.33
(14) E: It’s OK if some groups have more of a chance in life
than others.
.15
.51
.00
.71
(15) D: To get ahead in life, it is sometimes necessary to step
on other groups.
.17
.35
.14
.54
(16) E: We should do what we can to equalize conditions for
different groups.(-)
.07
.52
.00
.32
(17) D: It’s probably a good thing that certain groups are at
the top and other groups are at the bottom.
.18
.57
.12
.78
(18) D: Inferior groups should stay in their place.
.38
.35
.27
.41
(19) E: We would have fewer problems if we treated people
more equally.(-)
-.08
.62
.00
.42
(20) E: It would be good if groups could be equal.(-)
.04
.66
.00
.49
(21) D: In getting what you want, it is sometimes necessary
to use force against other groups.
.23
.31
.26
.45
(22) E: All groups should be given an equal chance in life.(-)
.03
.50
.00
.32
(23) D: If certain groups stayed in their place, we would have
fewer problems.
.48
.23
.39
.30
(24) E: Social equality should increase. (-)
-.09
.61
.00
.33
(25) D: Sometimes other groups must be kept in their place.
.44
.08
.44
.16
(26) D: Some people are simply inferior to others.
.34
.17
.37
.30
(27) E: No group should dominate in society.(-)
.06
.28
.00
.22
Note. Dimensions I and II can be interpreted as RWA and SDO; factor loadings > .30 in
boldface; (-): Reverse codings; the last two columns include the RWA and SDO factor
loadings including blended loadings from SDO-D items on RWA from the best fitting model
based on CFA (see Figure 1b).

44
44
Structure and Sources of Social Attitudes
Table 4. Univariate Twin Modeling Results
Twin
correlations
Model fit statistics and
indices
Standardized
components
Variables
MZ
DZ
χ²
df
p
RMSEA
CFI
h²
c²
e²
RWA
Self-reports
.70**
.42**
10.49
3
.02
.081
.951
.54**
.16*
.30**
ØPeer reports
.48**
.38**
4.69
3
.20
.043
.972
.20*
.31**
.53**
Composites
.71**
.47**
8.37
3
.04
.077
.958
.47**
.24**
.29**
True scores
.88**
.60**
19.17
13
.12
.035
.985
.57**
.31**
.12**
SDO
Self-reports
.35**
.32**
1.23
3
.75
.000
1.00
.05
.30**
.65**
ØPeer reports
.21**
.20*
4.54
3
.21
.036
.975
.01
.20*
.79**
Composites
.34**
.33**
0.69
3
.88
.000
1.00
.02
.32**
.66**
True scores
.66**
.62**
7.83
13
.86
.000
1.00
.08
.58**
.34**
Note. Ø: averaged; MZ: monozygotic twin pairs (self-reports: n = 226; other reports,
composites, and latent true scores: n = 176); DZ: dizygotic twin pairs (self-reports: n = 168;
other reports, composites, and latent true scores: n = 134); h²: additive genetic variance
component; c²: shared environmental variance component; e²: nonshared environmental
incl. error variance component (except for true score).

45
45
Structure and Sources of Social Attitudes
Table 5. Model of Twins Reared Together and Twins’ Parents
Phenotypic statistics
Variance decomposition
Variance
h² + m² + f² + s² + 2mfμ + h²m(1 + μ) + h²f(1 + μ) + e²
MZ twin covariance
h² + m² + f² + s² + 2mfμ + h²m(1 + μ) + h²f(1 + μ)
DZ twin covariance
½h²(1 + μ) + m² + f² + s² + 2mfμ + h²m(1 + μ) + h²f(1 + μ)
Parents covariance
μ
Mother-twin covariance
½h² (1 + μ) + m + μf
Father-twin covariance
½h² (1 + μ) + f + μm
Note. h: additive genetic effects; s: sibling-specific shared environmental effects; e:
environmental effects not shared by twins (incl. measurement error); μ: phenotypic
assortative mating; m: maternal transmission; f = paternal transmission.

46
46
Structure and Sources of Social Attitudes
Table 6. Nuclear Twin Family Model Fitting Results: Fit Statistics and Standardized Path
Coefficient Estimates
Fit statistics
Standardized effects
Variables
Model
χ²(df)
p
RMSEA
CFI
h
μ
m
f
s
e
RWA
Full
10.18(13)
.68
.000
1.00
.73*
.47*
.22*
-.08
.31*
.52*
Reduced
11.55(14)
.64
.000
1.00
.68*
.47*
.20*
.37*
.53*
SDO
Full
18.06(13)
.16
.032
.947
.00
.43*
.15*
.20*
.50*
.81*
Reduced
18.06(14)
.20
.027
.958
.43*
.15*
.20*
.50*
.81*
Note. The reduced model row for each variable shows parameters of the model with the
best parsimony-fit relation; h: additive genetic effects; μ: phenotypic spouse similarity; m:
environmental transmission from mother to offspring; f: environmental transmission from
father to offspring; s: sibling-specific shared environmental effects; e: environmental effects
not shared by twins (incl. measurement error); *significant (p < .05).

47
47
Structure and Sources of Social Attitudes
Figure Caption
Figure 1. Confirmatory factor models: A model (a) with two substantive factors (RWA and
SDO), one method factor (positive item formulations in terms of social desirability), and
blended loadings from SDO-D items on the latent RWA factor (dashed paths); (b) with five
substantive factors (RWA + facets and SDO), one method factor, and blended loadings from
SDO-D items on the latent RWA Aggression factor (dashed paths); (c) with four substantive
factors (RWA and SDO + facets), one method factor, and a blended loading from the latent
SDO Dominance factor on latent RWA (dashed path); and (d) with 5 + 2 hierarchical
substantive factors, one method factor, and a blended loading from the latent SDO
Dominance factor on latent RWA (dashed path). For simplicity random error and residual
variance components are not shown.

48
48
Structure and Sources of Social Attitudes

49
49
Structure and Sources of Social Attitudes

50
50
Structure and Sources of Social Attitudes
Figure 2. Nuclear twin family model for monozygotic (MZ) and dizygotic (DZ) twins: G:
genetic factors; E: environmental factors; h: additive genetic effects; m: environmental
transmission from mother to offspring; f: environmental transmission from father to
offspring; s: sibling-specific shared environmental effects; e: environmental effects not
shared by twins (incl. measurement error); μ: phenotypic spouse similarity; latent factor
variances were fixed to one.

51
51
Structure and Sources of Social Attitudes
Figure 3. Standardized variance components based on the nuclear twin family model fitting
results: Estimates are based on the models with the best parsimony-fit relation (see Table 6);
h²: heritability estimates (i.e., variance due to additive and nonadditive genetic effects);
h²m(1 + μ) + h²f(1 + μ): variance due to passive genotype-environment correlation; m² + f² +
2mfμ: variance due to environmental transmission from parents to offspring; s²: variance
due to sibling-specific shared environmental effects; e²: variance due to environmental
effects not shared by twins; ε²: error of measurement based on the difference between MZ
twin true-score correlation and MZ twin self-report correlation (see Table 4); genetic and
environmental variance components corrected for attenuation due to error of measurement
are shown in parentheses.

52
52
Structure and Sources of Social Attitudes
Online Supplements
Table S1. Factor Loadings of RWA and SDO Items on the First Two Promax Rotated Factors Based on
Principal Axis Factoring Analyses for Males and Females
Social attitudes
Males
(N = 476)
Females
(N = 961)
I.
II.
I.
II.
RWA:
(1) A: Instead of more civil rights, we need an upholding of law and order.
.68
-.08
.68
-.05
(2) S: The days when women are submissive to their husbands and social
conventions belong strictly in the past. A “woman’s place” should be
wherever she wants to be.(-)
.15
.29
.11
.16
(3) C: Turning away from tradition will emerge as fatal error one day.
.39
.15
.33
.02
(4) A: There is no crime that legitimates the death penalty.(-)
.35
.18
.27
.03
(5) S: Obedience and respect for authority are the most important virtues
that children should learn.
.67
-.08
.74
-.06
(6) C: Homosexual cohabitations should be put on a par with marriage.(-)
.33
.18
.32
.19
(7) A: What our country really needs is a strong, determined leader who
will crush evil, and take us back to our true path.
.69
.02
.69
-.10
(8) S: It is wonderful that young people today have greater freedom to
protest against things they don’t like and to do their own thing.(-)
.21
.18
.19
.19
(9) C: In the long run, virtuousness and law-abiding behavior will bring us
forward, not the permanent challenge of our society’s foundations.
.53
-.02
.64
-.06
(10) A: It is important to protect fully the rights of radicals and deviants.(-)
.27
.09
.19
-.00
(11) S: The real keys to a good life are obedience, discipline, and sticking
to the old and narrow.
.73
-.04
.77
-.05
SDO-Dominance (D) and SDO-Egalitarianism (E):
(12) E: We should strive to make incomes as equal as possible. (-)
-.40
.59
-.38
.53
(13) E: Group equality should be our ideal.(-)
-.43
.42
-.34
.51
(14) E: It’s OK if some groups have more of a chance in life than others.
.14
.54
.15
.58
(15) D: To get ahead in life, it is sometimes necessary to step on other
groups.
.16
.40
.19
.40
(16) E: We should do what we can to equalize conditions for different
groups.(-)
.20
.52
-.02
.61
(17) D: It’s probably a good thing that certain groups are at the top and
other groups are at the bottom.
.14
.64
.18
.60
(18) D: Inferior groups should stay in their place.
.43
.34
.40
.43
(19) E: We would have fewer problems if we treated people more
equally.(-)
-.13
.69
-.10
.66
(20) E: It would be good if groups could be equal.(-)
.03
.69
.02
.71
(21) D: In getting what you want, it is sometimes necessary to use force
against other groups.
.19
.27
.27
.39
(22) E: All groups should be given an equal chance in life.(-)
.08
.53
-.02
.59
(23) D: If certain groups stayed in their place, we would have fewer
problems.
.61
.22
.49
.28
(24) E: Social equality should increase. (-)
-.12
.64
-.12
.66
(25) D: Sometimes other groups must be kept in their place.
.50
-.01
.49
.14
(26) D: Some people are simply inferior to others.
.38
.15
.39
.23
(27) E: No group should dominate in society.(-)
.08
.33
.05
.34
Note. The first and second dimensions can be interpreted RWA and SDO; factor loadings > .30 in
boldface; (-): Reverse codings.

53
53
Structure and Sources of Social Attitudes
Table S2. Factor Loadings of RWA and SDO Items on the First Two Promax Rotated Factors Based on
Principal Axis Factoring Analyses for Parental (Age ≥ 40) and Offspring (Age < 40) Generations
Social attitudes
Offspring
(N = 630)
Parents
(N = 557)
I.
II.
I.
II.
RWA:
(1) A: Instead of more civil rights, we need an upholding of law and order.
.63
-.05
.59
-.07
(2) S: The days when women are submissive to their husbands and social
conventions belong strictly in the past. A “woman’s place” should be
wherever she wants to be.(-)
.14
.16
.15
.20
(3) C: Turning away from tradition will emerge as fatal error one day.
.23
.09
.33
.06
(4) A: There is no crime that legitimates the death penalty.(-)
.38
-.02
.19
.10
(5) S: Obedience and respect for authority are the most important virtues
that children should learn.
.70
-.12
.70
-.01
(6) C: Homosexual cohabitations should be put on a par with marriage.(-)
.23
.22
.18
-.20
(7) A: What our country really needs is a strong, determined leader who
will crush evil, and take us back to our true path.
.68
-.11
.63
.02
(8) S: It is wonderful that young people today have greater freedom to
protest against things they don’t like and to do their own thing.(-)
.22
.09
.17
.19
(9) C: In the long run, virtuousness and law-abiding behavior will bring us
forward, not the permanent challenge of our society’s foundations.
.52
-.04
.51
.03
(10) A: It is important to protect fully the rights of radicals and deviants.(-)
.20
.09
.16
.02
(11) S: The real keys to a good life are obedience, discipline, and sticking
to the old and narrow.
.66
-.10
.77
.06
SDO-Dominance (D) and SDO-Egalitarianism (E):
(12) E: We should strive to make incomes as equal as possible. (-)
-.30
.53
-.36
.43
(13) E: Group equality should be our ideal.(-)
-.27
.52
-.35
.30
(14) E: It’s OK if some groups have more of a chance in life than others.
.17
.52
.14
.46
(15) D: To get ahead in life, it is sometimes necessary to step on other
groups.
.29
.21
.16
.44
(16) E: We should do what we can to equalize conditions for different
groups.(-)
.11
.56
.00
.47
(17) D: It’s probably a good thing that certain groups are at the top and
other groups are at the bottom.
.17
.54
.17
.56
(18) D: Inferior groups should stay in their place.
.34
.33
.35
.41
(19) E: We would have fewer problems if we treated people more
equally.(-)
-.05
.63
-.11
.58
(20) E: It would be good if groups could be equal.(-)
.09
.66
-.06
.65
(21) D: In getting what you want, it is sometimes necessary to use force
against other groups.
.31
.26
.19
.31
(22) E: All groups should be given an equal chance in life.(-)
.09
.45
-.04
.52
(23) D: If certain groups stayed in their place, we would have fewer
problems.
.46
.19
.45
.33
(24) E: Social equality should increase. (-)
-.04
.61
-.17
.59
(25) D: Sometimes other groups must be kept in their place.
.45
.10
.43
.12
(26) D: Some people are simply inferior to others.
.36
.20
.31
.15
(27) E: No group should dominate in society.(-)
.06
.25
.02
.27
Note. The first and second dimensions can be interpreted as RWA and SDO; factor loadings > .30 in
boldface; (-): Reverse codings.

54
54
Structure and Sources of Social Attitudes
Table S3. Factor Loadings of RWA and SDO Items on the First Two Promax Rotated Factors Based on
Principal Axis Factoring Analyses for Self-reports and Averaged Peer Ratings
Social attitudes
Self-reports
(N = 746)
Øpeer
reports
(N = 746)
I.
II.
I.
II.
RWA:
(1) A: Instead of more civil rights, we need an upholding of law and order.
.66
-.04
.77
-.05
(2) S: The days when women are submissive to their husbands and social
conventions belong strictly in the past. A “woman’s place” should be
wherever she wants to be.(-)
.16
.17
.12
.26
(3) C: Turning away from tradition will emerge as fatal error one day.
.25
.09
.36
.00
(4) A: There is no crime that legitimates the death penalty.(-)
.38
.02
.45
.05
(5) S: Obedience and respect for authority are the most important virtues
that children should learn.
.69
-.13
.75
-.12
(6) C: Homosexual cohabitations should be put on a par with marriage.(-)
.29
.20
.28
.27
(7) A: What our country really needs is a strong, determined leader who
will crush evil, and take us back to our true path.
.66
-.13
.72
-.14
(8) S: It is wonderful that young people today have greater freedom to
protest against things they don’t like and to do their own thing.(-)
.23
.12
.28
.24
(9) C: In the long run, virtuousness and law-abiding behavior will bring us
forward, not the permanent challenge of our society’s foundations.
.56
-.06
.58
-.02
(10) A: It is important to protect fully the rights of radicals and deviants.(-)
.22
.05
.26
.15
(11) S: The real keys to a good life are obedience, discipline, and sticking
to the old and narrow.
.69
-.12
.70
-.11
SDO-Dominance (D) and SDO-Egalitarianism (E):
(12) E: We should strive to make incomes as equal as possible. (-)
-.26
.56
-.29
.55
(13) E: Group equality should be our ideal.(-)
-.27
.54
-.30
.55
(14) E: It’s OK if some groups have more of a chance in life than others.
.16
.53
.08
.52
(15) D: To get ahead in life, it is sometimes necessary to step on other
groups.
.26
.26
.16
.28
(16) E: We should do what we can to equalize conditions for different
groups.(-)
.09
.57
.16
.62
(17) D: It’s probably a good thing that certain groups are at the top and
other groups are at the bottom.
.17
.58
.15
.55
(18) D: Inferior groups should stay in their place.
.38
.31
.39
.33
(19) E: We would have fewer problems if we treated people more
equally.(-)
-.04
.65
-.11
.73
(20) E: It would be good if groups could be equal.(-)
.08
.66
.08
.64
(21) D: In getting what you want, it is sometimes necessary to use force
against other groups.
.31
.27
.24
.21
(22) E: All groups should be given an equal chance in life.(-)
.08
.48
.05
.60
(23) D: If certain groups stayed in their place, we would have fewer
problems.
.47
.18
.50
.24
(24) E: Social equality should increase. (-)
-.03
.61
-.02
.70
(25) D: Sometimes other groups must be kept in their place.
.43
.06
.48
.15
(26) D: Some people are simply inferior to others.
.35
.19
.42
.10
(27) E: No group should dominate in society.(-)
.08
.28
.04
.44
Note. The first and second dimensions can be interpreted as RWA and SDO; factor loadings > .30 in
boldface; (-): Reverse codings.