As in the previous studies, conspiracy mentality (α = .89) only showed small or negligible correlations with RWA (α = .76), r = .15, p = .01, and SDO (α = .87), r = -.05, p = .41, whereas RWA and SDO were strongly associated, r = .57, p < .001.
We first aggregated the power ratings across all participants, ICC(2,1) = .53, p < .001. Table 4 lists these ratings in descending order. The inclusion of the five target groups of Study 2 allowed us to test whether the assumptions about the ordering of these groups with respect to their perceived power were indeed borne out. As expected, both Americans and capitalists were perceived as highly powerful. Further, the hierarchy of perceived power from capitalists being most powerful to Roma and Sinti being least powerful was both significant as a linear trend, and all group differences were significant in Bonferroni-corrected paired t-tests (α = .005). At the level of the 32 social groups, mean ratings of power, ICC(2,k) = .997, were unrelated to the ratings of likeability, ICC(2,k) = .992, r = −.04, p = .83, and positively related to ratings of threat, ICC(2,k) = .995, r = .56, p = .001. Thus, more powerful groups were seen as more threatening but neither as more nor as less likeable.
Bonferroni-corrected zero-order correlations between the three generalised political attitudes and ratings of likeability and threat (Table 4) showed similar patterns for RWA and SDO but not for conspiracy mentality. To formally test this, we conducted vector correlations between the Fisher r-to-z-transformed correlations. Results show highly similar correlations for RWA and SDO with likeability, r = .94, and threat, r = .96, ps < .001, whereas conspiracy mentality did neither show the same correlation pattern with likeability ratings as RWA, r = −.02, p = .87, or SDO, r = −.27, p = .13, nor for threat ratings, r = −.14, p = .45, with RWA, r = −.28, p = .12, and SDO correlations.
Specifically, the results revealed that conspiracy mentality was associated with lower ratings of likeability for politicians and capitalist as well as greater perceptions of threat coming from politicians, power companies, managers, capitalists, physicists and Turks. In contrast, RWA predicted greater ratings of likeability for power companies and managers and lower liking for out-groups (Jews, Muslims, Foreigners, Roma and Sinti, and asylum seekers), poor people (unemployed, drug addicts and homeless), feminists, gay men and artists. With a few exceptions (Jews, artists and homeless), the same groups were also seen as threatening, as were Turks and welfare recipients. Results for SDO were highly similar with SDO predicting perceptions of threat for largely the same groups (except Roma and Sinti and drug addicts), less liking for largely the same groups (except drug addicts but additionally less liking for feminists, Muslims, gay men, Turks and welfare recipients) and like RWA greater liking of power companies (but not managers).
To formally underline the fact that high-power groups were seen as less likeable but more threatening by people high in conspiracy mentality, we treated the 32 social groups as cases and computed correlations between their power ratings and the Fisher r-to-z-transformed correlations of the likeability and threat rating with conspiracy mentality. Results showed that indeed greater power was associated with more negative relations between conspiracy mentality and likeability, r = −.35, p = .05, whereas the relation between conspiracy mentality and threat ratings became more positive with increasing power, r = .47, p = .006. An opposite pattern emerged for RWA and SDO for which increasing power led to more positive relations with likeability ratings, r = .70 for RWA and r = .75 for SDO, and more negative relations with threat ratings, r = −.72 for RWA and r = −.76 for SDO, all ps < .001.
However, our hypotheses concerned the role of generalised political attitudes in the intraindividual relationship between power and likeability/threat. We were particularly interested in whether individuals' degree of conspiracy mentality would be associated with their more negative evaluation of powerful groups. To address whether this was the case, we calculated for each participant whether his or her ratings of power were associated with his or her ratings of likeability and threat across the 32 groups. Positive scores on the power–likeability coefficient indicated that participants rated more powerful groups as more likeable, whereas positive scores on the power–threat coefficient indicated that they perceived groups as increasingly threatening with increasing power. We hypothesised that power and threat should be more highly positively correlated, and power and likeability should be more highly negatively correlated, with increasing conspiracy mentality. We expected the reverse pattern for SDO and RWA. To test these hypotheses, we conducted two multiple regression analyses with the Fisher r-to-z-transformed intraindividual correlations as criteria, conspiracy mentality, RWA and SDO as predictors and an alpha level of α = .025 to account for conducting two analyses.
Whereas individual associations between power and likeability revolved around zero, average r = −.02, SD = .31, (ranging from r = −.77 to r = .74), the relation between power and threat was positive on average (r = .38, SD = .31; ranging from r = −.60 to r = .80). As expected, higher levels of conspiracy mentality were associated with more negative individual-level correlations between power and likeability, ß = −.17, p = .002, whereas the opposite was true for RWA, ß = .35, p < .001, and SDO, ß = .21, p = .001 (Figure 2(a)). In contrast, this pattern reversed for the association of power and threat with conspiracy mentality as a positive predictor, ß = .14, p = .02, unlike RWA, ß = −.27, p < .001, and SDO, ß = −.17, p = .01 (Figure 2(b)). Thus, people high in conspiracy mentality rated higher-power groups as less likeable and as more threatening, whereas people high in RWA and SDO had a more positive view of higher-power groups and perceived lower-power groups as more threatening. Importantly, all three effects were incremental to each other. A stepwise procedure with RWA entered in a first step, SDO in a second and conspiracy mentality in a third step yielded significant increases in the explained variance for each step, ps < .02. Exploratory analyses including the cross-products revealed no significant interaction effects.
Figure 2. Unique relations between conspiracy mentality, right-wing authoritarianism (RWA), and social dominance orientation (SDO) when simultaneously predicting intraindividual correlations of (a) power and likeability (R2corr = .26), and (b) power and threat (R2corr = .15). Point estimates are plotted for each predictor at −1 standard deviation (SD) and +1SD at the mean value of the other predictors. Regressions are based on Fisher r-to-z-transformed coefficients that were inverse-transformed for the plots.
Download figure to PowerPoint