Sex Roles (2007) 57:159–171 DOI 10.1007/s11199-007-9231-3

ORIGINAL ARTICLE

The Relationship between Social Dominance Orientation and Gender: The Mediating Role of Social Values Luca Caricati

Published online: 23 May 2007 # Springer Science + Business Media, LLC 2007

Abstract Invariance hypothesis posits that, across cultures and contexts, men are higher in SDO than women. Social Dominance Theory (SDT) suggests that this difference is biologically determined and resulting from the differences in reproductive strategies between sexes. In this study we tested the hypothesis that values can explain gender–SDO differences. SDT suggests that SDO mediates gender differences in values. The contrasting hypothesis suggests that gender–SDO differences are mediated by values. Using SEM with a sample of Italian University students (N = 162, 80 men), the results provided strong support for the hypothesis of a mediation effect of values. Indeed, when controlling for values, gender difference in SDO disappear. Very similar findings were obtained for both GBD and OEQ. Results are discussed.

differences in values are mediated by SDO. The second model, according to the Role Learning Theory (e.g. Eagly and Wood 1999), predicts that gender differences in SDO are mediated by the adhesion to values. The primary goal of this study was, therefore, to test these two alternative models. This was done by mediational analysis using Structural Equation Modelling. Since there is a growing debate about the nature of SDO (see e.g. Duckitt 2001; Rubin and Hewstone 2004; Turner and Reynolds 2003), to test the nature of the relation between gender, SDO and values seems to be relevant in the definition of the Social Dominance nature.

Keywords Social dominance orientation . Social values . Gender differences

Social Dominance Theory (SDT; Sidanius 1993; Sidanius and Pratto 1999) is one of the most famous theories in intergroup relations. SDT rests on the observation that social groups are always hierarchically organized and that people share some ideologies legitimizing the hierarchical organization (i.e. hierarchy-enhancing legitimizing myths) and others which are opposed to hierarchy (i.e. hierarchy-attenuating legitimizing myths). At a personal level, the adhesion to these two kinds of ideology is captured by a construct named Social Dominance Orientation (SDO), defined as “a general attitudinal orientation toward intergroup relations, reflecting whether one generally prefers such relations to be equal, versus hierarchical” (Pratto et al. 1994, p. 742). Thus, from the SDT’s point of view, people are more or less oriented to see social groups organized along a superiority-inferiority dimension and to support social inequality. People who score high on SDO should be more oriented to endorse hierarchyenhancing legitimizing myths, whereas people who score low should be more oriented to endorse hierarchy-attenuating

The aim of this study was to explore the nature of the relation among gender, Social Dominance Orientation (SDO, Pratto et al. 1994) and social values. Indeed, although there are some resemblances between the content of certain values (e.g. power, universalism and benevolence) and SDO as well as between gender differences in SDO and values, there is a lack of knowledge about their relationships. In this respect, two alternative models can be advanced. The first one, according to Social Dominance Theory (Sidanius and Pratto 1999; Sidanius et al. 1994), stresses that gender

L. Caricati (*) Department of Psychology, University of Parma, Borgo Carissimi, 10, 43100 Parma, Italy e-mail: [email protected]

Defining Social Dominance Orientation

160

legitimizing myths. Consistently, several empirical pieces of evidence showed that SDO positively correlated with general conservative beliefs such as political and economic conservatism, ethnic prejudice, and right-wing ideologies (Pratto et al. 1994; Sidanius and Pratto 1999; Sidanius et al. 1996; Pratto et al. 1997). Moreover, SDO seems to be positively correlated with ideologies such as nationalism, patriotism, sexism, anti-Black racism, cultural elitism, belief in meritocracy, pro-military attitudes, and favouring punitive legal policies (Pratto 1999; Sidanius and Liu 1992). In addition, SDO negatively correlated with favouring civil-rights policies, social welfare and environmentalism (Pratto et al. 1994). Recently, Jost and Thompson (2000) showed that SDO is a bi-dimensional construct composed of two distinct, but related, sub-scales: the first, named Group-Based Dominance (GBD), measures people’s tendency to accept and support group-based social dominance; the latter, named Opposition to Equality (OEQ), refers to people’s tendency to oppose to social equality. Accordingly, OEQ was positively correlated with political conservatism and negatively with the support of affirmative action. Moreover, OEQ was strongly associated with economic system justification. On the contrary, GBD was correlated neither with political conservatism nor with the support of affirmative action, and it was only weakly correlated with economic system justification (Jost and Thompson 2000). These results suggest that OEQ is an ideological component of SDO which serves a system justifying ideological function (Jost and Banaji 1994). Thus, according to Jost and Thompson (2000), although they correlate with each other, GBD and OEQ are two distinct, albeit related, ideological components of SDO.

Ideological Asymmetry and the Invariance Hypothesis SDT suggests that any social structure has a dominant group (or groups) at the top of the hierarchy and a subordinate group (or groups) at the lower end of the hierarchy. SDT indicates three distinct stratification systems (Sidanius and Pratto 1999): (a) age-set system in which the oldest are the powerful, (b) gender-set system in which men have more power than women and (c) arbitrary-set system in which the powerful and the powerless are defined on the basis of any salient social characteristic (e.g. ethnicity, nation or race). Even though subordinates may often endorse legitimizing myths, typically dominants are expected to endorse these myths to a stronger degree and therefore they are higher in SDO than are subordinates (Sidanius and Pratto 1999). This phenomenon is known as Ideological Asymmetry (IA; Sidanius and Pratto 1999; Sidanius et al. 1994).

Sex Roles (2007) 57:159–171

According to gender-set system, IA is also true for gender so that men are higher in SDO than women (Pratto et al. 1997; Sidanius and Pratto 1999; Sidanius et al. 1996). However, even though the levels of SDO are expected to interact with other structural variables such as culture or status in the arbitrary-set system, in the gender-set system this interaction is not expected. Indeed, gender-based difference is assumed to be universal and independent from social and cultural contexts (e.g. genetically determined); a principle known as the Invariance Hypothesis (IH). According to Sidanius and Pratto (1999), IH is partly driven by men’s desire to justify their dominant position in society but SDT places an emphasis also on the differences in reproductive strategies. The result is that men are “chronically” more oriented toward group-based social inequality (SDO) than women. Consistently with these expectations, Sidanius and Pratto (1999) found that across ten samples (Australia, Canada, Israel, Mexico, Palestine, Republic of China, New Zealand, the former USSR, Sweden and the United States) men scored higher in SDO than women. (see also Sidanius et al. 2000). These results, however, do not account directly for any stable gender difference in SDO, rather they suggest that gender differences in SDO are less sensitive to the cultural factors. With respect to both GBD and OEQ, Jost and Thompson (2000), have argued that, for dominant groups, GBD and OEQ are congruent ideologies that enhance and justify the advantaged ingroup, whereas, for subordinate groups, GBD and OEQ are not congruent with each other. This should occur since GBD tends to enhance the position of both groups while OEQ tends to justify the disadvantaged position of the ingroup (e.g. my ingroup deserves what it gets). Accordingly, Jost and Thompson (2000) found that dominants (White Americans) had a stronger relationship between GBD and OEQ than would subordinates (African Americans). Moreover, for both groups, OEQ was correlated with general conservative beliefs whereas GBD was correlated with ingroup favouritism.

Problems Associated with SDO Empirical pieces of evidence have shown that SDO possesses its own specificity with respect to other constructs correlated with it (e.g. authoritarianism). Altemeyer (1998) defines SDO as “an extraordinary measure of prejudice” and “the best measure we have of missing link in the dominationsubmission authoritarian social system” (p.85). However, some doubts about the validity of SDO and its interpretation still remain (Rubin and Hewstone 2004; Turner and Reynolds 2003). Particularly, the invariance hypothesis has been greatly criticized by both social identity and role learning theorists. For example, Huang and Liu (2005)

Sex Roles (2007) 57:159–171

showed that, in accordance with SDT, Taiwanese men scored higher than women in SDO, but this was true only when gender was made salient. When gender was not salient, both men and women showed the same levels of SDO. Moreover, Wilson and Liu (2003) showed that both New Zealander males’ and females’ SDO scores were dependent on gender identification. Also Dambrun et al. (2004) confirmed with a sample of French students that SDO is dependent on identification with the ingroup: when gender identification was statistically controlled, men and women showed the same levels of SDO; however, SDO did not mediate the strength of gender identification. Recently, also Chatard et al. (2005) showed that gender differences in SDO are mediated by gender identification. Particularly, the authors showed that, across four countries (Belgium, France, Switzerland and Tunisia), gender differences in SDO disappeared when gender identification was inserted as mediator. This evidence is problematic for a conceptualization of SDO as an individual, relatively stable personality trait. Social identity theorists have argued that SDO may be better defined as a general attitude toward social inequality or, using the words of Turner and Reynolds (2003) “SDO is a product of social life rather than an underlying cause” (p. 200). These arguments are in line with those advanced by role learning scholars (see Eagly and Wood 1999). Particularly, Foels and Pappas (2004) advanced the idea that the gender differences in SDO (and its sub-scales) are due to the different socialization processes for men and women. The effects that socialization has on an individual’s masculine and feminine behaviour have been largely discussed in the last decades and documented through the use of gender role measures such as the Bem Sex Role Inventory (Bem 1974) and the Personal Attributes Questionnaire (Spence et al. 1974). In this vein, Foels and Pappas (2004) tested the hypothesis that, when controlled for socialization, gender difference in GBD and OEQ would be weakened or eliminated. Their results (Study 1) provide some evidence that “the difference between men and women is not invariant when controlling for social factors, which suggests that gender differences in social dominance may be learned through masculine socialization” (p. 748). In Study 2, Foels and Pappas (2004) confirmed and extended the results of Study 1 showing that GBD was related to masculinity but not to femininity, whereas OEQ was related to femininity and also partly related to masculinity. The authors concluded that “Socialization can and does influence an individual's level of both SDO-D [i.e. GBD] and SDO-E [i.e. OEQ], with the relationship between sex and SDO-D weakened, and sex and SDO-E reduced to virtually zero (p=.499) when masculinity and femininity were controlled” (p. 752).

161

The Present Study The present study was carried out in order to take into account the relationship between gender, SDO and social values. Social values can be considered as “desirable, transsituational goals, varying in importance, that serve as guiding principles in people’s lives” (Struch et al. 2002, p. 16). Schwartz (1992) identifies a set of ten values (conformity, tradition, security, self-direction, stimulation, hedonism, power, achievement, benevolence and universalism). Schwartz’s system has two high-order dimensions of values: Self-Transcendence (universalism and benevolence) vs. Self-Enhancement (power and achievement); and Conservation (security, conformity and tradition) vs. Openness to Change (stimulation and selfdirection). Opposite values have opposite motivational emphases while adjacent values share motivational emphasis. According to Schwartz (1992), hedonism shares features of both Self-Enhancement and Openness to Change. The above structure has been verified in cross-cultural analyses carried out in each of 65 cultures (comprised Italy) which largely replicated the expected values structure (Schwartz 1992, 1994; Schwartz and Sagiv 1995). Further confirmatory factor analysis of value relations across 23 countries also replicated the structure (Schwartz and Boehnke 2004). With respect to SDO, there are four values that should be more relevant: power, achievement, benevolence and universalism. Power is defined by Schwartz (1992; Struch et al. 2002) as the desire to obtain social prestige and to have control and domination over people and resources. Similarly, achievement is defined as the desire to reach social prestige through a demonstration of competence. Thus, SDO, power and achievement are expected to be strongly, and positively, correlated. At the same time, universalism and benevolence (which refer to social equality and fairness) should be negatively correlated to SDO. Accordingly, McFarland (1998; 1999; McFarland and Adelson 1996), found that SDO was negatively correlated with universalism and positively with power, whereas it was not correlated with tradition, conformity, security and self-direction. Values of hedonism, stimulation and benevolence were not administered by McFarland and colleagues. More recently, Duriez and Van Hiel (2002), with a Belgian sample, and Heaven and Connors (2001), with Australian undergraduate psychology students, highlighted that SDO was positively correlated with values of self-enhancement (e.g. power and achievement) and negatively correlated with values of self-transcendence (e.g. universalism and benevolence). Moreover, SDO does not seem correlated with values referring to both conservation and openness to change. Considering gender, previous studies have shown that men and women differ in respect to the importance they assign to each value. A recent article published in this Journal (Lyons et al. 2005) which surveyed results from

162

research on gender differences carried out with Rokeach Value Survey (RVS, Rokeach 1973), showed a clear gender difference in value orientation. Generally, men are more oriented to self-enhancement values (i.e. power and achievement) while women are more oriented to both conservation and self-transcendence values (Bond 1988; Feather 1984; Hitlin and Piliavin 2004; Rokeach 1973, Schwartz et al. 2001). Role learning theory explains these gender differences in values as the product of the current division of the labour (Eagly et al. 2000). Occupational and family roles provide men and women with different experiences that directly affect behaviours, attitudes, representations and basic values. Moreover, shared different roles for the two sexes create different gender role expectations which influence men’s and women’s feelings, thoughts and behaviours. Generally, men are predominant in occupations that enjoy more power and status. On the contrary, women are more often occupied in activities that promote care for young children, the elderly or the sick people. Thus, the effect of the differences in labour system is that men are more oriented to agentic-powerful values (e.g. power and achievement) whereas women are more oriented to communal-caring values (e.g. benevolence, conformity and tradition). Role learning theory can also explain gender differences in openness to change and security values. Indeed, since the men occupy more often higher status positions, they should be more direct to self-reliance, independence and self-gratification. Thus, men should be more oriented to values promoting stimulation and good life. Women, occupying lower status positions, are more dependent from others and more vulnerable. Thus, women should be more oriented to the security value (see Schwartz and Rubel 2005). In summary, gender-related differences in values overlap gender differences in SDO and, therefore, exploring the connection between values, gender and SDO seems a logical next step. However, there is a lack of research into the relationship between gender, values and SDO.

The Context of the Present Research Before proceeding, a further specification about the effect of cultural context is needed. Indeed, most of the previous researches we reviewed have been carried out in the US or into Anglo-Saxon contexts. Instead, the data presented in this study refer to an Italian sample. Nevertheless, there are several reasons to believe that data from Italy can be compared with other data. Firstly, the division of the labour in Italy is still gender-stereotyped and unequally distributed. Indeed, a national survey carried out in 2005 by the EURISPES (an Italian institution for political, economical

Sex Roles (2007) 57:159–171

and social studies) revealed that women come to have lower salary and a more irregular payment than men. Moreover, female occupational rate, even though growing, is well inferior to that of men. Further, men are employed in higher status jobs more frequently than women. According to Role learning theory, these differences in labour seem to affect general beliefs about gender differences. Indeed, crosscultural studies on hostile and benevolent sexisms showed that gender stereotypes in Italy are substantially similar to those in other occidental countries (e.g. Glick et al. 2000): men are seen as more competent and women as more warm. In the same way, some Italian researches on the beliefs about gender differences and gender roles found very similar results to those obtained in the American context \(e.g. Papafagou, unpublished work; Ragazzi, unpublished work), that is, Italians believe that men are (and should be) employed in more prestigious-agentic jobs and women are (and should be) employed in more caringoriented jobs (e.g. housework, teaching and nursing). These differences seem to be applicable also to gender differences in SDO. Consistently, and accordingly to previous crosscultural researches (e.g. Pratto et al. 2000; Sidanius et al. 2000), Caricati (2007a) found that, in the Italian context, men were higher in SDO than women. Also Aiello et al. (2005) found expected gender differences in SDO between Italian men and women. Regarding value structure and priorities, researches show that, even though value structure is relatively invariant across samples, the importance of each value changes according to the specific culture or nation. However, some regularity among value priority can be found. Generally, collectivistic cultures assign more relevance to conservative values, whereas individualistic cultures show higher priority in openness to change values (e.g. Schwartz 1992; Triandis 1996). From this point of view, Italy can be considered as an individualistic nation like US and some other European nations. Consistently, Italy and US are very similar in value priority if compared with pan-cultural normative baseline (Schwartz and Bardi 2001). A difference between similar nations could rest in the strength of value orientation (that is, Americans could be more individualistic than Italians) but this does not seem to affect the general organization of value system. Accordingly, Caricati (2007b) found that, for Italian adolescents, values of openness to change and selftranscendence were more important than values of conservation and self-enhancement. Particularly, like in the US (Schwartz and Bardi 2001), Italians ranked benevolence, hedonism and self-direction as the more important values and tradition and power as the less important values. A difference was found anyhow: Italians were less oriented to achievement than Americans. Further researches on gender differences in values priorities show that, across several countries, it is

Sex Roles (2007) 57:159–171

possible to find precise patterns for men and for women. Particularly, Schwartz and Rubel (2005, Study 1) showed that data from 19 European countries (included Italy) clearly indicated that women ranked higher on selftranscendence and conservation values. Men, instead, were higher in self-enhancement and openness to change values. In Study 4, Schwartz and Rubel (2005), analysing 64 samples (8 samples from Africa, 11 from Asia, 3 from Middle East, 16 from Central and Eastern Europe, 16 from Western Europe, 6 from Latin America, 2 from North America and 2 from Oceania) showed a similar pattern: women were consistently higher in benevolence and universalism (but not in conservative values) and men were higher in self-enhancement and openness to change values. Cross-cultural variances in gender differences in values priorities have been found in this research, that is, gender differences were more pronounced in poor and collectivistic cultures whereas they tended to be reduced in more individualistic-autonomous and rich countries. In other words, as stated above for general value priority, also gender differences in value priorities are affected by cultural variability but, inside culture groups e.g. collectivistic vs. individualistic), some regularities can be found. Accordingly, Caricati (2007b) and Capanna et al. (2005) found that Italian men were higher in power, achievement and hedonism whereas women scored high in benevolence and conformity. The whole set of these results seems to suggest that Italian and American contexts are very similar for what concerns variables which are relevant in this research and that cultural context does not affect seriously neither gender differences in both SDO and values, nor general beliefs about gender, social dominance and values. Thus, it seems that data from an Italian context can be compared with data from other contexts.

The Questions SDT proposes a precise causal model about the relationship between gender, values and SDO. Since gender differences in SDO are hypothesized to be less sensitive to cultural factors, value expressions may be a product of legitimizing ideologies, which is why women and men may use different value expressions. Thus, SDO is expected to mediate the gender differences in values. The predicted path is the following (see also Sidanius et al. 1994, p. 1001): Gender
! SDO
! Values In the logic of this pattern, gender differences in values are mediated by SDO so that values become legitimizing myths

163

that enhance (e.g. power and achievement) or attenuate (e.g. universalism and benevolence) inequality between groups. The question we raise is as follows: could the causal model proposed by SDT explain the relationship between values, SDO and gender? Indeed, one can reasonably argue that gender differences in SDO are due to gender differences in values. Since values are “social” in their essence, gender differences in values may be traced back to both cultural characteristics (such as ideologies, Maio et al. 2003) and socialization processes (Hitlin and Piliavin 2004). Thus, if gender differences in SDO are mediated by the adhesion to values, then the Invariance Hypothesis is challenged. This alternative path is the following: Gender
! Values
! SDO Our question arises from two observations. On one hand, a great body of research has shown consistently that SDO is both context sensitive and influenced by socially based human features. On the other hand, it seems questionable that SDO scale would really measure a personality trait. Indeed, items such as “We should strive to make income more equal” or “It would be good if all groups could be equal” ask people what they think, and not what they do, so those items measure attitude and not personality (Ray 2006). Grouped together, these observations suggest the possibility that SDO may be really an expression of an attitude toward group inequality which depends on the “motivations,” or “guiding principles,” driving human behaviour, that is, values. If this is the case, then gender differences in SDO mainly could be due to different adhesion to values. In this respect the literature suggests that gender differences should affect only three (namely power, achievement and benevolence) of the four values expected to be correlated with SDO (as stated above, SDO does not seem correlate with values of conservation and openness to change). Indeed, previous findings suggest no differences in predisposition or socialization relevant to universalism, nor are the goals of universalism differently attainable (PrinceGibson and Schwartz 1998). However, since universalism refers to welfare of all others, including those which are not part of the ingroup, we expect that universalism will be negatively and strongly related to SDO. Thus, if SDO is a product of both socialization and social life, then gender difference on SDO should be weakened or eliminated whether controlled for values (namely, benevolence, power and achievement). Findings in support of this prediction would suggest that the gender difference in SDO is not invariant as IH posits. The primary goal of this study, therefore, is to test these two alternative patterns of relations (i.e. SDT model versus “value as mediator” model).

164

Resuming, we advance the following major hypotheses about values and SDO: (1a) According to both Role learning theory and previous results, men should score higher in achievement, power, hedonism, self-direction and stimulation. Women, instead, should be higher in conservation, tradition, security and benevolence. (1b) Men should score higher than women in SDO. (2) SDO should be positively linked to achievement and power and negatively with benevolence and universalism. (3) Values should mediate gender differences in SDO. Accordingly,” Value as mediator” model should fit the data better than SDT model. The second goal of this study is to test whether sub-scales of SDO, namely GBD and OEQ, are influenced by both people’s gender and values. To our knowledge, since Jost and Thompson’s (2000) and Foels and Pappas’s (2004) studies, other findings about both GBD and OEQ are not available. Generally, we predict that both GBD and OEQ should be negatively linked to benevolence and universalism and positively linked to power and achievement. However, since OEQ refers to social system justification (Jost and Thompson 2000), it should be also linked to values of conservation (i.e. conformity, tradition and security). With regard to gender differences, since men are considered as a high status group and women as a low status group, we predict that men should be more oriented to OEQ than women, meanwhile men and women should be more similar with respect to GBD (even if, theoretically, a difference may still occur). Finally, the same two alternative patterns of mediation addressed for SDO could be tested also for OEQ and GBD. Thus, our hypotheses about SDO subscales can be resumed as following; (4) Men should be higher in OEQ. (5) GBD and OEQ should show the same correlations of SDO. In addition, OEQ should be correlated with conservative values. (6) As for SDO, values should mediate gender differences in both GBD and OEQ. Accordingly, “value as mediator” model should fit the data better than “GBD as mediator” and “OEQ as mediator” models.

Method Sample One hundred and sixty two students at the University of Parma (Italy) were enrolled in this study. Eighty two

Sex Roles (2007) 57:159–171

(50.2%) of these were women and 80 (49.4 %) were men. The mean age of the sample was 23 years (SD=4.32; range: 20–51 years). Participation was voluntary and subjects were not paid. Materials and Procedure Participants anonymously completed paper-and-pencil questionnaires in the classroom before the lesson. Firstly, participants completed the Italian version of the Portrait Value Questionnaire (PVQ; Caricati 2007b; Schwartz 2004; Schwartz et al. 2001) already validated with back translation in a large representative Italian sample by Schwartz et al. (2001) and Capanna et al. (2005). PVQ is a questionnaire that surveys ten typologies of values according to Schwartz’s (1992) theory. PVQ includes short verbal portraits of 40 people. Each one depicts a person’s aspirations, goals or wishes that point implicitly to the importance of a value (e.g. “It’s very important to him to show his abilities. He wants people to admire what he does”). For every item, participants have to indicate how much the portrayed persons are similar to themselves on a 6 point scale (“very much like me,” “like me.” “some-what like me,” “a little like me,” “not like me,” “not like me at all”). According to previous studies (see e.g. Schwartz 1992), the score for each value type was the mean rating of the single items postulated a priori to measure that value (for example, conformity was calculated as the mean of items 7, 16, 28, 36). Internal reliability (Cronbach’s α) of values in this study was .61 for conformity, .50 for tradition; .67 for benevolence; .75 for universalism; .63 for self-direction; .70 for stimulation; .81 for hedonism; .84 for achievement; .68 for power and .74 for security. The relatively low reliabilities of some values reflect the fact that each value score depends on a smaller set of items which are intended to capture a rather wide construct. Observed reliabilities, however, were in the range of variation commonly observed for values (see Devos et al. 2002). Anyway, the magnitude of reliability coefficients was not correlated neither with the priority assigned to each value nor with the strength of the correlation between values, SDO and its subscales. In other words, differences in internal reliabilities do not account for the results reported below. After completing the PVQ, participants were asked to fill in the Italian version (validated with a back translated scale administered to a large sample of Italian students by Aiello et al. 2005) of the SDO composed of 16 items. Eight items indicate orientation to group-based social dominance (GBD), whereas another eight indicate group-based social inequality (OEQ; reverse coded). Internal reliability of SDO was very good (Cronbach’s α=.88) as well as for GBD (α=.76) and OEQ (α=.89).

Sex Roles (2007) 57:159–171

165

Sidanius and Pratto 1999) men scored higher in SDO than women.

Results First, data concerning analyses of SDO will be presented. Second, we will present results for both GBD and OEQ. In the last section, further analyses will be presented. Gender Differences in Values and SDO A multivariate oneway ANOVA was performed on values as dependent variables and gender as independent. Overall effect was significant (Wilks’ λ=.85, F(10,151)=2.75, p<.01; η2 =.15). Observing univariate results in Table 1 it is possible to see that, in accord with hypothesis 1a, the women scored significantly higher on conformity, tradition and benevolence but only marginally on security. The men, alternatively, scored higher on power and hedonism. Furthermore, men and women did not differ with respect to universalism, as foreseen. Unexpectedly, men and women showed the same levels of achievement. However, this latter result is consistent with that reported by Ryckman and Houston (2003). In addition, men and women did not differ on self-direction and stimulation values. This result can be explained by the age of the participants. Indeed, some researches showed that gender differences in openness to change values are attenuated or reduced in student samples (Schwartz and Rubel 2005). Moreover, univariate analysis of variance revealed that, as expected from hypothesis 1b, gender affected SDO (F(1,160)= 4.40, p<.01; η2 =.026). Consistent with previous findings (e.g.

The Relations Among Sex, Values and SDO: Testing the Models As we stated previously, and according to both SDT and IH, if SDO mediates adhesion to social values, then gender effect on SDO should not disappear if controlled for values. Rather, SDO should mediate the relations between gender and values. On the contrary, if IH is a consequence of the adhesion to values, then gender difference on SDO should disappear when controlled for values. Table 2 shows inter-correlations among considered variables. The upper half of the table shows correlations for female sample (correlations for men were very similar. The only relevant differences were that GBD was significantly correlated with achievement (r=.32, p<.01) and power (r=.30, p<.01)). The lower half of the Table 2 shows overall correlations. In order to test the two alternative models we employed path analysis using AMOS 5 software (Arbuckle 2003). In order to test mediation analysis it is required that the mediating variable correlates with both the dependent and independent variables (Baron and Kenny 1986). As showed in the lower half of Table 2, and partially in accord with hypothesis 2, three values satisfied these requirements, namely: benevolence, power and hedonism. However, we did not add hedonism to the model for theo-

Table 1 Total means (standard deviation), and means separated for gender, values, SDO, GBD and OEQ. Men (N=80)

Values Tradition Benevolence Conformity Hedonism Power Security Stimulation Universalism Self-direction Achievement Social dominance SDO GBD OEQ

Women (N=82)

F(1,160)

η2

M

SD

M

SD

2.94 4.60 3.55 4.40 3.01 3.51 4.08 4.70 4.83 4.01

.70 .71 .76 1.00 1.00 .89 .86 .70 .71 .98

3.31 4.91 3.89 3.95 2.64 3.75 3.88 4.78 4.76 4.00

.88 .65 .77 .98 .83 .94 .94 .71 .68 1.04

8.68**** 8.50**** 8.10**** 8.11**** 6.52**** 2.83** 2.01 0.51 0.33 0.02

.051 .050 .048 .048 .039 .017 .012 .003 .002 .000

3.11 3.36 2.87

.91 1.03 1.13

2.81 3.10 2.52

.92 .92 1.15

4.40*** 2.72* 3.64***

.026 .017 .022

A higher score indicates a higher grade of measured variables. Value scores range from 1=not like me at all to 6=very much like me. SDO, GBD and OEQ scores range from 1=Strongly disagree to 7=Strongly agree. * p =.10 ** p<.10 *** p<.05 **** p<.01

.24*** .12 .28*** .07 −.20* .09 −.09 .37*** .18* .19* – .61**** .26*** .00

.29**** .15 .36*** .01 −.31*** −.11 −.18 .23** .13 .19* .59**** – .09 .20**

retical motives. Indeed, hedonism refers to intra-personal dimension of gratification for oneself (e.g. pleasure, enjoying life) whereas SDO, power and benevolence refer to a social-relational dimension. Accordingly, we inserted only power and benevolence as mediating variables in the models (Anyway, further multiple regression analysis on hedonism revealed that its effect on SDO decreased when controlling for gender (β=.13, p=.10) and further reduced when benevolence and power were added (β=.05, p=.56)). Models were tested with standardized coefficients obtained from the maximum likelihood method estimation. For each model we tested, we report several indices of fit (CFI and RMSEA). Model 1: SDO as Mediator

.11 .08 .11 −.31*** −.32*** .05 −.08 .34*** – .40**** .29**** .29**** −.24*** .11

.22** .25*** .15 .01 −.12 −.17 −.27*** −.01 .31*** – .23*** .32**** .08 .22*** −.63**** −.55**** −.56**** .16 .15 .53**** – .19** −.02 −.06 −.03 −.13 .16** .05 −.25*** −.27*** −.20* .14 .19* – .48**** .16** .02 −.11 .06 −.10 .07 −.22*** −.02 .02 −.05 – .45**** .27**** .17** −.20*** −.30**** −.12 .11 −.03 .52**** −.22***

Model 2: Values as Mediators This alternative model was tested using the same statistical procedure. As shown in Fig. 2, and consistently with expectations, the effect of gender (−1=women; 1=men) on SDO disappeared (β=.05, ns) when benevolence and power were statistically controlled. The effects of benevolence (β=−.23, p<.01) and power (β=.27, p<.01) on SDO still

Benevolence

.86**** – .50**** −.00 −.08 −.24*** −.45**** −.01 .10 .21** .22*** .25*** .19** .12

.91**** .58**** – −.05 −.05 −.22*** −.58**** .05 .12 .08 .21*** .27*** .00 .15**

.08 −.07 −.08 .39*** – .29**** .17** −.25*** −.34**** −.16** −.14 −.27*** .32**** −.23***

−.00 −.07 .07 −.20* −.27*** .24** .21* – .35**** .16** .36**** .34**** .01 .04

As shown in Fig. 1 and consistently with expectations, controlling for SDO, the effects of gender (−1=women; 1= men) on benevolence (β=−.18, p<.05) and power (β=.16, p<.05) still remained significant. This was inconsistent with the view that SDO mediates the effect of gender on social values. However, when controlling for gender, the effects of SDO on benevolence (β=−.24, p<.01) and on power (β=.28, p<.01) still remained significant. Various indices indicated that the proposed model presented in Fig. 1 (without the dotted lines between sex and both values) does not fit the data (χ2(3)=9.93, p<.05; CFI=.78; RMSEA=.12). Clearly, this model does not support the SDT hypothesis.

- .18* - .24**

– .84**** .89**** −.03 −.07 −.26*** −.60**** .02 .12 .16** .24*** .30**** .10 .16**

Gender

.16*

SDO .28**

p<.10 p<.05 *** p<.01 **** p<.0001

Power

**

1. SDO 2. GBD 3. OEQ 4. Conformity 5. Tradition 6. Benevolence 7. Universalism 8. Self-Direction 9. Stimulation 10. Hedonism 11. Achievement 12. Power 13. Security Gender

.16*

*

12 11 10 9 8 7 6 5 4 3 2 1

Table 2 Correlations between SDO, values and gender. The lower half refers to total sample (N=162), the upper half refers to female sample (N=82).

.11 .23** .00 .58**** .17 −.01 .14 .11 −.18 .11 .24** .22** – −.13

Sex Roles (2007) 57:159–171

13

166

Fig. 1 Path analysis. SDO as a mediator of the effect of gender on benevolence and power. Single asterisk indicates p<.05, double asterisks indicate p<.01. Without dotted lines: χ2(3)=9.93, p<.05, CFI=.78, RMSEA=.12. Gender: women=−1, men=1. N=162.

Sex Roles (2007) 57:159–171

167

Benevolence - .22**

- .23** .05 ns

Gender

SDO

.20**

.27** Power

Fig. 2 Path analysis. Benevolence and power as mediators of the effect of gender on SDO. Double asterisks indicate p<.01. Without the dotted line: χ2(2)=1.04, p=.59, CFI=1.00, RMSEA=.00. Gender: women=−1, men=1. N=162.

remained significant even when gender was statistically controlled. These results reveal that, as hypothesized, both the values provide independent mediation of the effect of gender on SDO. Indeed, using z tests (Kenny 2005), it appears that benevolence (z=2.00, p<.05) and power (z= 2.07, p<.05) were two independent mediators both of which reduced significantly the variance accounted for by gender on SDO. The model presented in Fig. 2, without the dotted line between gender and SDO, had a good fit. The chisquare was not significant (χ2(2)=1.04, p=.59), and the various indices were very high (CFI=1.00; RMSEA=.00). This model accounted for 14% of the variance in SDO. Comparing the chi-squared of the two models, it became evident that, according to hypothesis 3, the “value as mediator” model best fits the data (χ2(1)=8.89, p<.01). These results challenge the invariance hypothesis. It seems that when controlled for power and benevolence, men and women did not differ in levels of SDO. In other words, the difference in SDO may be accounted for by men and women’s adherence to power and benevolence. On the contrary, data does not support the SDT’s hypothesis that SDO mediates the men and women’s adhesion to values. Analysing GBD and OEQ In order to assess gender effect on GBD and OEQ, a multivariate analysis of variance was performed on GBD and OEQ as dependent variables and gender as independent. The overall effect was marginally significant (Wilks’ λ=.97, F(2,159)=2.14, p=.11; η2 =.03). However, observing Table 1 it is possible to see that gender only marginally affected GBD but, as expected from hypothesis 4, significantly affected OEQ: men were more oriented to oppose social equity (M=2.87) than women (M=2.52). For what concerns their relations with values and as showed in Table 2 (lower half), GBD and OEQ shared some correlations. As expected from hypothesis 5, both scales were negatively correlated with benevolence and universalism and positively with achievement and power.

However, contrary to our expectation, OEQ was not correlated with conservative values whereas GBD was positively correlated with security. However, these results do not account for the causal model among gender, OEQ, GBD and values. Since the gender effect on GBD was weak, we tested our SDO model using only OEQ. (Given the marginally significant gender difference in GBD, we also tested both Model 1 and 2 with GBD. Results were very similar to the results of OEQ. In Model 1, the effects of the gender on benevolence (β=−.20, p<.01) and power (β=.17, p<.05) still remained significant. Path model (without dotted lines): χ2(3)=11.51, p<.01; CFI=.69; RMSEA=.13. In Model 2, effect of gender on GBD disappeared (β=−.04, ns) when benevolence and power were statistically controlled. The effect of values remained significant (β=−.22, p<.01 and β=.23, p<.01, respectively). Path model without dotted line: χ2(2)=.82, p=.66, CFI=1.00; RMSEA=.00. Difference in Chi-squared clearly showed that the Model 2 was better than the Model 1: χ2(1)=10.69, p<.01). Model 1: OEQ as Mediator As for SDO, path analysis showed that controlling for OEQ, the effects of gender (−1=women; 1=men) on benevolence (β=−.19, p<.05) and power (β=.15, p<.05) still remained significant. Also this result was inconsistent with the view that OEQ mediates the effect of gender on social values. However, controlling for gender, the effects of OEQ on benevolence (β=−.19, p<.01) and on power (β=.25, p<.01) still remained significant. Various indices indicated that the proposed model (without the lines between gender and both values) does not fit the data (χ2(3)=10.12, p<.05; CFI=.65; RMSEA=.13). Model 2: Values as Mediator Similar to the results for SDO, analysis of the Model 2 reveals that the effect of gender (−1=women; 1=men) on OEQ disappeared (β=.05, ns) when benevolence and power were statistically controlled. The effects of benevolence (β=−.19, p<.01) and power (β=.24, p<.01) on OEQ still remained significant even when gender was statistically controlled. As for SDO, and according to hypothesis 6, these results reveal that both the values provide independent mediation of the effect of gender on SDO. The model (without the line between gender and OEQ), had good fit. The chi-square was not significant (χ2(2)=1.06, p=.59), and the various indices were very high (CFI = 1.00; RMSEA=.00). This model accounted for 11% of the variance in OEQ and it was very supportive for “value as mediator” hypothesis. Comparing the chi-squared of the two models it became evident that, according to hypothesis

168

6, the “value as mediator” model is best fitted to the data (χ2(1)=9.06, p<.01).

Sex Roles (2007) 57:159–171

Therefore, results revealed a strong effect of universalism and power on both GBD and OEQ when gender was not taken into account.

Further Analyses Discussion Relations Between Values and SDO In order to verify which values were linked to SDO without taking into account the gender, a stepwise multiple regression analysis, in which all ten values (centered around 0) were inserted as predictors of SDO, was performed. The final full model was significant (F(3,158)=41.78; p<.0001; R2 =.44). Given the high correlations of value scores, we checked for multicollinearity. As Cohen et al. (2003) pointed out, values of Variance Inflation Factor (VIF) higher than 10 and Tolerance values lower than .10 indicate serious problems of multicollinearity. In this analysis, diagnostics showed that both VIF and tolerance values were quite acceptable. Moreover, Durbin-Watson statistic was 1.82 indicating few or not collinearity. Univariate results indicated that universalism (β=−.60; t (160)=−9.88; p<.0001, VIF=1.05; Tolerance=.95), power (β=.21; t (160)=3.45; p<.01, VIF=1.03, Tolerance=.97) and security (β=.18; t (160)=3.02; p<.01,VIF=1.03, Tolerance=.96) were the only statistically significant predictors. Thus, universalism explained about 40% of the variance of SDO and was the best predictor of SDO when all values were controlled and gender was not taken into account. Relations Between Values, GBD and OEQ Even in this case, in order to verify what values were predictors of both GBD and OEQ two separate stepwise multiple regression analyses were performed. We inserted ten values (centered on 0) as predictors of both GBD and OEQ, respectively. For GBD (Multicollinearity diagnostics yielded acceptable values, Durbin-Watson statistic=1.93) the final full model was significant (F(3,158)=15.54; p<.0001; R2 =.30). Univariate results showed three significant effects due to universalism (β=−.47; t (160)=7.18; p<.0001, VIF=1.05; Tolerance=.95), security (β=.25; t (160)=3.72; p<.001, VIF=1.04, Tolerance=.96) and power (β=.17; t (160)=2.51; p<.05, VIF=1.03, Tolerance=.97). Also for OEQ the final full model was significant (F(2,159)=47.53; p<.0001; R2 =.37), but only two values, namely universalism, (β=−.55; t (160)=8.73; p<.0001, VIF=1.02, Tolerance=.98) and power (β=.20; t (160)= 3.16; p<.01, VIF=1.02, Tolerance=.98) were predictors of OEQ (Durbin-Watson statistic=1.81).

As conceptualized by SDT, SDO reflects an orientation toward both social dominance and social inequality (Sidanius and Pratto 1999). SDT places a great emphasis on the fact that men should be “chronically” higher in SDO than women. Accordingly, the invariance hypothesis posits that, across cultures and contexts, men should be more SDO oriented than women. However, in this study we have found only limited support for these conclusions. Replicating previous findings, men scored higher in SDO than women, but this was not true when we controlled for values. Indeed, when controlling for values, men showed the same level of SDO as women, whereas gender differences on power and benevolence did not disappear when controlling for SDO. Using structural equation modelling, the results of the present study are clear. The SDT model does not fit the data, but the results provide strong support for the role learning model. These results are very unsupportive also for the view of a genetic causation of gender differences in SDO. Indeed, SDT places an emphasis on evolutionary mechanisms by suggesting that higher levels of SDO among men are also part of their sexual/reproductive strategies. Women can optimize reproductive success by being more selective in their choice of mates. This selectivity would explain the female preference for powerful, high-status mates (Pratto and Hegarty 2000). On the other hand, male reproductive success is increased when they have sexual access to several women (Pratto and Hegarty 2000), reaching control over other men and women would optimize male reproductive success. The result is that, everything else being equal, men should be more social dominance oriented than women. Our results, instead, imply that the gender difference in SDO could be explained without necessarily (or mainly) making appeal to genes (Sidanius et al. 1994) or sexual reproductive strategies (Sidanius and Pratto 1999). On the contrary, if one assumes that the gender difference in values is socially based, then our results suggest that SDO difference between men and women is neither mainly genetically determined (invariance hypothesis) nor context insensitive. Rather, and consistently with Foels and Pappas’s (2004) findings, the different socialization processes of the two sexes seem to affect gender differences in SDO.

Sex Roles (2007) 57:159–171

It seems that, considering the results of this study, the only way for social dominance scholars to maintain the genetic foundation of SDO would be to postulate that values are fixed in the genes. However, even though theoretically conceivable, this arrangement would be contrary to another assumption of SDT. Indeed, as Sidanius et al. (1994) have pointed out, values are kinds of legitimizing myths which are socially shared, hence not biologically based. Moreover, Sidanius et al. (1994) overtly propose a causal linkage for which legitimizing myths are driven by SDO (p. 1001). Thus, positing that values are biologically fixed and, at the same time, that they are driven by another biologically determined trait (i.e. SDO), would be quite problematic. There is another finding of this study that challenges the Invariance hypothesis, particularly, the lack of consistent gender differences in SDO sub-scales. Indeed, men and women show only marginal differences in the levels of GBD. The weakness of gender-related difference in GBD is very interesting and congruent with the results from Foels and Pappas (2004) which showed a greater influence of gender on OEQ than on GBD. Furthermore, and congruently with the Foels’ and Pappas’ finding of a strong correlation between GBD and masculinity and no correlation with GBD and femininity, we found that GBD was strongly correlated with both achievement and power for men but not for women. This is in line with Jost and Thompson’s (2000) statement that GBD and OEQ are two distinct components of SDO that are differently affected by the status group. Indeed, since GBD refers to the improvement of the ingroup, both men and women may endorse GBD without problems for their self-worth. On the contrary, for the subordinated group (e.g. women), OEQ endorsement threatens their self-worth, while for the dominant group (e.g. men) this is not true. Our findings are supportive for System Justification Theory (Jost and Banaji 1994) which posits that high status groups (e.g. men) should be more oriented to justify the social system and inequalities between groups than low status groups (e.g. women). However, the evidence of a strong gender difference in OEQ along with a marginally significant difference in GBD, challenges Social Dominance Theory. Indeed, according to Sidanius and Pratto (1999) and the asymmetry hypothesis, dominant groups (e.g. men) should be more oriented to group-based dominance than dominated groups (e.g. women). Our data does not confirm completely this asymmetric hypothesis. This is a serious problem for the evolutionary foundation of SDO. Indeed, why would genes influence mainly the gender difference in opposition to equality rather than differences in group based dominance? According to Dambrun et al. (2004), this question is “compatible with the analysis of Ward (1995) who suggested

169

that the gender difference in SDO should be understood as an attitudinal difference between the powerful (i.e. men) and the powerless (i.e. women)” (p. 294). Taken together these findings suggest that SDO could be the expression of a value that is opposed to both benevolence and universalism and that is more similar to power and achievement. Particularly, SDO may be a “value” that refers to domination/fairness toward groups as a whole and not toward the “generalised other” (as is power). This hypothesis is consistent with the general findings that SDO and Right-Wing Authoritarianism (RWA) are weakly correlated (Altemeyer 1998). The lack of a strong linkage between SDO and RWA has been interpreted on the basis of the different targets underlying the two constructs: RWA has an intragroup (or interpersonal) target whereas SDO has an intergroup target. In the same way, since both power and achievement refer to interpersonal relations, they should be weakly correlated to SDO. The above hypothesis is also consistent with the finding that universalism, but not benevolence, is the best predictor of SDO, accounting about the 40% of the variance. Benevolence, even though correlated with universalism, is not a predictor of SDO when all values are taken into account. Once more, these results become comprehensible if one assumes that universalism and benevolence refer to two different forms of “kindness”. Indeed, benevolence refers to a more private and interpersonal dimension (e.g. the wellbeing of the people with which we are frequently in personal contact; Schwartz 1992) whereas universalism refers to a more “humanitarian” dimension (e.g. the wellbeing for, and between, all people) hence the latter is mainly linked to social groups’ welfare. Since SDO refers to groups, it is quite normal that SDO is associated with universalism rather than benevolence. These results challenge Social Dominance Theory because, conceptualizing SDO as a personality trait, SDT posits SDO as invariant (invariance hypothesis) as well as an antecedent of social life (also values). The results obtained in this study, however, strongly support the idea that SDO is better conceptualised as an expression of a value referring to domination among groups as a whole. This is in line with Duckitt’s Dual Process Model of Ideology and Prejudice (Duckitt 2001; Duckitt et al. 2002). Duckitt suggests that SDO is more appropriately a measure of an ideological belief (or attitude) which refers to “power distance/humanitarian” dimension. This conception of SDO directly overlaps Schwartz’s (1992) dimension of “SelfEnhancement/Self Trascendence” values. Congruently, as stated above, SDO could be conceptualized as a groupbased component of “power distance/humanitarian” ideological dimension.

170

Moreover, our results clearly support the hypothesis that gender differences in SDO should be due to the different socially based features that characterize men and women. In other words, SDO seems to be a consequence of the people’s socially learned general attitudes and beliefs regarding social equality rather than the cause. Once more, we can agree with Turner and Reynolds (2003) when they state that “SDO is a product of social life rather than an underlying cause” (p. 200).

References Aiello, A., Chirumbolo, A., Leone, L., & Pratto, F. (2005). Uno studio di Adattamento e Validazione della scala di Orientamento/ Tendenza alla Dominanza Sociale (Pratto et al. 1994) [A study for the validation of the social dominance orientation scale]. Rassegna di Psicologia, 22, 65–75. Altemeyer, B. (1998). The other “authoritarian personality”. In M. Zanna (Ed.), Advances in experimental social psychology, vol. 30 (pp. 47–92). New York: Academic. Arbuckle, J. L. (2003). Amos 5.0 [Computer Software]. Chicago, IL: SPSS. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Bem, S. L. (1974). Measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42, 155–162. Bond, M. (1988). Finding universal dimensions of individual variation in multicultural studies of values: The Rokeach and Chinese value surveys. Journal of Personality and Social Psychology, 55, 1009–1015. Capanna, C., Vecchione, M., & Schwartz, S. H. (2005). La misura dei valori: un contributo alla validazione del Portrait Values Questionnaire su un campione italiano. [The measurement of values. A contribution to the validation of the Portrait Value Questionnaire on an Italian sample]. Bollettino di Psicologia Applicata, 246, 29–41. Caricati, L. (2007a). Orientamento alla dominanza sociale e ipotesi dell’asimmetria ideologica fra maschi e femmine: un’analisi con il modello di Rasch [Social Dominance Orientation (SDO) and asymmetry hypothesis between men and women: An analysis via the Rasch Model]. (in press). Caricati, L. (2007b). Organizzazione, gerarchia e struttura valoriale nei giovani adolescenti: il modello di Schwartz e l’utilizzo del Portrait Value Questionnaire (PVQ).[Organization, hierarchy and structure of values among young adolescents: The Schwartz’s Model and use of the Portrait Value Questionnaire (PVQ)]. Rassegna di Psicologia, 24, 147–158. Chatard, A., Guimond, S., Lorenzi-Cioldi, F., & Désert, M. (2005). Domination masculine et identité de genre. Les Cahiers Internationaux de Psychologie Sociale, 67–68, 113–123. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences, (3rd edn.). Hillsdale: Erlbaum. Dambrun, M., Duarte, S., & Guimond, S. (2004). Why are men more likely to support group-based dominance than women? The mediating role of gender identification. British Journal of Social Psychology, 43, 287–297. Devos, T., Spini, D., & Schwartz, S. H. (2002). Conflicts among human values and trust in institutions. British Journal of Social Psychology, 41, 481–494.

Sex Roles (2007) 57:159–171 Duckitt, J. (2001). A dual process cognitive-motivational theory of ideology and prejudice. In M. P. Zanna (Ed.), Advances in experimental social psychology, vol. 33 (pp. 41–113). San Diego, CA: Academic. Duckitt, J., Wagner, C., du Plessis, I., & Birum, I. (2002). The psychological bases of ideology and prejudice: Testing a dual process model. Journal of Personality and Social Psychology, 83, 75–93. 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. Eagly, A. H., & Wood, W. (1999). The origins of sex differences in human behavior: Evolved dispositions versus social roles. American Psychologist, 54, 408–423. Eagly, A. H., Wood, W., & Diekman, A. B. (2000). Social role theory of sex differences and similarities: A current appraisal. In T. Eckes & H. M. Taunter (Eds.), The developmental social psychology of gender (pp. 123–174). Mahwah, NJ: Erlbaum. EURISPES (2005). Rapporto Italia 2005. Consulted online at: http:// www.eurispes.it/. Feather, N. T. (1984). Masculinity, femininity, psychological androgyny, and the structure of values. Journal of Personality and Social Psychology, 47, 604–620. Foels, R., & Pappas, C. J. (2004). Learning and unlearning the myths we are taught: Gender and social dominance orientation. Sex roles, 50, 743–757. Glick, P., Fiske, S. T., Mladinic, A., Saiz, J. L., Abrams, D., Masser, B., et al. (2000). Beyond prejudice as simple antipathy: Hostile and benevolent sexism across cultures. Journal of Personality and Social Psychology, 79, 763–775. Heaven, P. C. L., & Connors, J. R. (2001). A note on the value correlates of social dominance orientation and right-wing authoritarianism. Personality and Individual Differences, 31, 925–930. Hitlin, S., & Piliavin, J. A. (2004). Values: Reviving a dormant concept. Annual Review of Sociology, 30, 359–393. Huang, L. L., & Liu, J. H. (2005). Personality and social structural implications of the situational priming of social dominance orientation. Personality and Individual Differences, 38, 249–499. Jost, J. T., & Banaji, M. R. (1994). The role of stereotyping in system justification and the production of false consciousness. British Journal of Social Psychology, 33, 1–27. 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. Kenny, D. A. (2005). Mediation analysis. Retrieved October 9, 2006, from: http://users.rcn.com/dakenny/mediate.htm. Lyons, S., Duxbury, L., & Higgins, C. (2005). Are gender differences in basic human values a generational phenomenon? Sex roles, 53, 763–778. Maio, G. R., Olson, J. M., Bernard, M. M., & Luke, M. A. (2003). Ideologies, values, attitudes, and behaviour. In J. Delamater (Ed.), Handbook of social psychology (pp. 283–308). New York: Plenum. McFarland, S. G. (1998). Toward a typology of prejudiced persons. Paper presented at the annual convention of the International Society for Political Psychology, Montreal, Canada, July 1998. McFarland, S. G. (1999). Personality, values, and latent prejudice: A test of a causal model. Paper presented at the annual convention of the International Society for Political Psychology, Amsterdam, The Netherlands, July 1999. McFarland, S. G., & Adelson, S. (1996). An omnibus study of personality, values, and prejudice. Paper presented at the annual

Sex Roles (2007) 57:159–171 convention of the International Society for Political Psychology, Vancouver, British Columbia, July 1996. Pratto, F. (1999). The puzzle of continuing group inequality: Piecing together psychological, social, and cultural forces in social dominance theory. In M. P. Zanna (Ed.), Advances in experimental social psychology, vol. 31 (pp. 191–263). San Diego: Academic. Pratto, F., & Hegarty, P. (2000). The political psychology of reproductive strategies. Psychological Science, 11, 57–61. Pratto, F., Liu, J. H., Levin, S., Sidanius, J., Shih, M., Bachrach, H., et al. (2000). Social dominance orientation and the legitimization of inequality across cultures. Journal of Cross-cultural Psychology, 31, 369–409. Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance orientation: A personality variable predicting social and political attitudes. Journal of Personality and Social Psychology, 67, 741–763. Pratto, F., Stallworth, L. M., & Sidanius. J. (1997). The gender gap: Differences in political attitudes and social dominance orientation. British Journal of Social Psychology, 36, 49–68. Prince-Gibson, E., & Schwartz, S. H. (1998). Value priorities and gender. Social Psychology Quarterly, 61, 49–67. Ray, J. J. (2006). Social dominance orientation: Theory or artefact?. Retrieved October 9, 2006, from: http://jonjayray.netfirms.com/ sdo.html. Rokeach, M. (1973). The nature of human values. New York: Free Press. Rubin, M., & Hewstone, M. (2004). Social identity, system justification, and social dominance: Commentary on Reicher, Jost et al., and Sidanius et al. Political Psychology, 25, 823–844. Ryckman, R. M., & Houston, D. M. (2003). Value priorities in American and British female and male university students. Journal of Social Psychology, 143, 127–138. Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries. In M. Zanna (Ed.), Advances in experimental social psychology, vol. 25 (pp. 1–65). New York: Academic. Schwartz, S. H. (1994). Are there universal aspects in the content and structure of values? Journal of Social Issues, 50, 19–45. Schwartz, S. H. (2004). Robustness and fruitfulness of a theory of universals in individual human values. In A. Tamayo & J. Porto (Eds.), Valores e trabalho [Values and work]. Brasilia: Editora Universidade de Brasilia. Schwartz, S. H., & Bardi, A. (2001). Value hierarchies across cultures: Taking a similarities perspective. Journal of Cross-cultural Psychology, 32, 268–290. Schwartz, S. H., & Boehnke, K. (2004). Evaluating the structure of human values with confirmatory factor analysis. Journal of Research in Personality, 38, 230–255. Schwartz, S. H., Melech, G., Lehmann, A., Burges, S., Harris, M., & Owens, V. (2001). Extending the cross-cultural validity of the

171 theory of basic human values with a different method of measurement. Journal of Cross-cultural Psychology, 32,519–542. Schwartz, S. H., & Rubel, T. (2005). Sex differences in value priorities: Cross-cultural and multimethod studies. Journal of Personality and Social Psychology, 89, 1010–1028. Schwartz, S. H., & Sagiv, L. (1995). Identifying culture specifics in the content and structure of values. Journal of Cross-cultural Psychology, 26, 92–116. Sidanius, J. (1993). The psychology of group conflict and the dynamics of oppression: A social dominance perspective. In S. Iyengar & W. McGuire (Eds.), Explorations in Political Psychology (pp. 183–219). Durham, NC: Duke University Press. Sidanius, J., Levin, S., Liu, J. H., & Pratto, F. (2000). Social dominance orientation, antiegalitarianism and the political psychology of gender: An extension and cross-cultural replication. European Journal of Social Psychology, 30, 41–67. Sidanius, J., & Liu, J. (1992). The gulf war and the Rodney King beating: Implications of the general conservatism and social dominance perspectives. Journal of Social Psychology, 132, 685–700. Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social hierarchy and oppression. New York: 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. Sidanius, J., Pratto, F., & Bobo, L. (1996). Racism, conservatism, affirmative action and intellectual sophistication: A matter of principled conservatism or group dominance?. Journal of Personality and Social Psychology, 70, 476–490. Spence, J. T., Helmreich, R., & Stapp, J. (1974). The personal attributes questionnaire: A measure of sex role stereotypes and masculinity-femininity. JSAS Catalogue of Selected Documents in Psychology, 4, 43. Struch, N., Schwartz, S. H., & van der Kloot, W. A. (2002). Meanings of basic values for women and men: A cross-cultural analysis. Personality and Social Psychology Bulletin, 28, 16–28. Triandis, H. C. (1996). The psychological measurement of cultural syndromes. The American Psychologist, 51, 407–415. Turner, J. C., & Reynolds, K. J. (2003). Why social dominance theory has been falsified. British Journal of Social Psychology, 42, 199– 206. Ward, D. (1995). Social dominance theory: Are the genes too tight? Paper presented at the Eighteenth Annual Meeting of the International Society of Political Psychology, Washington, DC. Wilson, M. S., & Liu, J. H. (2003). Social dominance orientation and gender: The moderating role of gender identity. British Journal of Social Psychology, 42, 187–198.