Introduction

In the light of current events in America following the death of George Floyd, events in Slovenia regarding the illegitimate deportation of asylum seekers (the severity of the issue is on the rise due to governmental changes), and other discriminatory practices around the globe, it is clear that discrimination is a persistent concern which needs to be addressed on the individual and societal level (Pincus, 1996). As humanity has witnessed atrocities caused by such race and ethnicity-based discrimination several times throughout the history, it is of global importance to help uncover the psychological and societal factors that give rise to in-group and out-group biases and to find ways of fostering social tolerance and mutual acceptance between diverse cultural groups. This research examines the influence of several factors on two intergroup biases, namely (1) neo-racism, which is discrimination based on race, ethnicity and culture and (2) ethnocentrism which is a strong favouritism of one’s own ethnic group over others.

It is well documented that intercultural contact may help to reduce discrimination towards other cultural groups (Sparkman & Eidelman, 2018). In line with the ‘contact hypothesis’ (Allport, 1954), positive intercultural contact with an out-group not only reduces prejudice towards that specific group, but this also extends to other out-groups not involved in the contact (Tadmor et al., 2012). However, the influence of intercultural contact on ethnocentrism has remained fairly unexplored, and the evidence so far is mixed (Cakal & Petrović, 2017; Kosmitzki, 1996).

Furthermore, apart from personal life experiences of an individual such as having contact with members of a different culture, a broader sociocultural context needs to be taken into account when studying intergroup biases (Tajfel et al., 1971). Cultural dimensions influence how people are socialised and which specific views and behaviours are reinforced (Salter et al., 2017). And since racism and ethnocentrism do not happen in a social vacuum but are formatted within a specific culture, it is crucial to study these social phenomena together with the context in which they occur. However, the extent to which cultural dimensions may have an influence on in-group favouritism and out-group discrimination, and which dimension has the strongest influence on intergroup biases, has not been systematically investigated yet. Moreover, the greatest contribution of this research is to combine cultural dimensions and intercultural contact into a single study in order to investigate and compare their predictive value on intergroup biases, as, to the best of knowledge, this is not something that it has been reported before in the literature.

To fill this gap, this study seeks to examine cultural dimensions (namely individualism/collectivism, social dominance orientation, and uncertainty avoidance) together with intercultural contact and explores which factor can best predict neo-racism and ethnocentrism among a sample of Slovenian adults.

The findings may have important implications for understanding different factors associated with discrimination and ethnocentrism, and for developing effective culturally-informed practices for preventing and reducing discrimination.

Literature Review

According to social identity theory (SIT) (Tajfel & Turner, 1986), intergroup biases occur because an individual’s identity is partly based on belonging to a group and in order to identify with a certain social group, it is necessary to differentiate among them. Differentiation is based on some distinct characteristics (such as gender, age, or ethnicity) which leads to the formation of positive or negative attitudes towards the in-group and out-group members (Tajfel et al., 1971). Negative attitudes may take the form of prejudice and stereotypes, which are overgeneralised and inflexible beliefs about an individual based solely on their group membership and may result in negative, discriminative behaviours (Allport, 1954; Fiske, 1998). The most severe form of discrimination based on race and ethnicity can and has led to major atrocities such as slavery, segregation and genocide (for example genocide in Bosnia, Rwanda, Germany and Iraq).

It is important to understand that prejudice is not static and that new forms of discrimination are emerging across cultures (Salter et al., 2017). Racism has lately taken on a new, more subtle and insidious form comprised of less overt expressions of racial bias, compared to so-called traditional racism (Tougas et al., 2004). Many scholars (e.g. Mamdani, 2002; Munck et al., 1993) claimed that racism used to be postulated on the idea of race, whereas the new form of racism tends to focus on (irreducible) cultural differences. As McConahay (1986) described, neo-racism consists of blaming minorities for their unjust demands and undeserved gains (e.g. employment equity), and for overall social problems (e.g. unemployment rate). It is often justified by cultural insecurities, such as perceiving minorities as a threat to existing cultural values and national identity, due to incompatibility between cultural groups. For this reason, neo-racism may be disguised in the ideology of patriotism as a defence of cultural traditions (Council of Europe, 2015).

Ethnicity or culture-based in-group favouritism, called ethnocentrism has been characterised as a “strong sense of ethnic group self-centredness and self-importance” (Bizumic & Duckitt, 2012, p. 4) and has been found to be a universal phenomenon (Van den Berghe, 1995). It has been associated with resource allocation in favour of the in-group, preferential collaboration with the in-group, and competition with out-group members. (Bizumic & Duckitt, 2012). This in-group bias does not necessarily include negative attitudes towards other groups however, it is still considered to be an important theoretical concept in explaining intergroup processes, especially because of its potentially adverse effects on positive relations (Hewstone et al., 2002). Although it is a universal phenomenon, its magnitude may be mediated by certain aspects of culture, like for example uncertainty avoidance and collectivism (Fischer & Derham, 2016).

As already stated, it is important to study social phenomena within the cultural context in which they occur. Because culture is seen as “a complex, multidimensional structure rather than a simple categorical variable” (Clark, 1987, p. 461), it is thus necessary to approach each cultural dimension separately. To identify various cultural dimensions, this research used two prominent cultural frameworks, namely Hofstede’s (1980, 2001) and Schwartz’s (1994), which propose several societal value dimensions by which cultures differ and can be compared. Hofstede’s (2001) model includes six dimensions: individualism vs. collectivism, uncertainty avoidance, power distance, masculinity vs. femininity, long-term orientation and lastly, indulgence vs. self-restraint. Schwartz (1994) developed three major dimensions: embeddedness vs. autonomy, hierarchy vs. egalitarianism, and mastery vs. harmony. Such classification of cultural values reflects which attitudes, beliefs and behaviours are more frequent in a certain culture compared to others, although there is some within-culture variation (Hanel et al., 2018). Cultural dimensions thus indicate the overall thinking and behavioural tendencies of groups of individuals, enabling us to understand and predict different psychological phenomena within a broader societal context.

Individualism/Collectivism

Hoftede’s (2001) individualism-collectivism dimension (similar to Schwartz’s (1994) embeddedness-autonomy), refers to the extent to which individuals are socialised to be independent or interdependent i.e. to what extent an individual is embedded in a social group, values, tradition, security, obedience, and prefers to avoid change, or on the other hand values autonomy in thinking and the pursuit of individual interests (Fischer & Derham, 2016). There exists disparate literature on how this cultural dimension could affect intergroup attitudes. Scholars agree that in-group favouritism and out-group discrimination are present in both collectivist and individualist societies; however, Triandis (1995) argues that those intergroup biases are more likely to occur among collectivists because they show greater concern for group identification and differentiation of the in-group from the out-group. Many studies support the notion that in-group bias (Schröder et al., 2013), specifically ethnocentrism, (Kongsompong et al., 2010; Neuliep et al., 2001) is stronger in collectivist societies. On the other hand, evidence for out-group discrimination among collectivists or individualists is mixed. While Oyserman’s (1993) findings are in line with Triandis’s predictions, Gouveia et al. (2011) found that individualists were more prejudiced towards Gypsies in Spain than collectivists. One of the proposed explanations for those results was that individualists, within a principally collectivist society, express positive affection only to close-knit people while maintaining their own space and independence, which results in lower intentions of contact and negative attitudes towards Gypsies.

Social Dominance Orientation

Hofstede’s (2001) power distance dimension (similar to Schwartz’s (1994) hierarchy-egalitarianism) is the extent to which individuals are socialised to support group-based hierarchies. On the individual level, this general attitudinal orientation is referred to as the social dominance orientation (SDO) (Pratto et al., 1994). As per social dominance theory (Pratto, 1999), racism and ethnocentrism are both positively related to higher SDO. This is because societies maintain group-based hierarchies’ trough cultural beliefs which reinforce stereotypes and prejudice. In line with the theory, higher SDO was found to be associated with ethnic prejudice (i.e. out-group discrimination, Pratto et al., 2000), but the evidence regarding its association to ethnocentrism is mixed. For example, Jost and Thompson (2000) found a positive correlation between attitudes opposing equality and ethnocentrism for European Americans, but negative correlation for African Americans. However, when they specifically measured support for group-based dominance, they found a positive correlation with ethnocentrism in both samples. Taken together, prior studies show that hierarchy enhancing versus attenuating societal ideology is an important factor influencing inter-group attitudes, but the manifestation of biases may also depend on other aspects of culture.

Uncertainty Avoidance/Need for Closure

Uncertainty avoidance (Hofstede, 2001) relates to the extent to which individuals are socialised to avoid uncertain situations. In Schwartz’s model (1994) the elements of this aspect are found in embeddedness (formerly called conservatism) and include security values, avoiding social change, and maintaining stability. It has been proposed that higher uncertainty avoidance increases out-group discrimination, which is mediated by the cognitive mechanism, need for closure (NFC) (Brizi et al., 2015). Individuals with high NFC tend to seek firm answers and draw quick conclusions in order to avoid ambiguity. People with low NFC tend to gather more knowledge before making decisions and feel less uncomfortable in uncertain situations (Kruglanski & Fishman, 2009). Researchers found that higher NFC individuals rely more heavily on stereotypes and are more discriminatory (Todor, 2014). With regards to ethnocentrism, Hogg’s (2007) uncertainty-identity theory explains that greater uncertainty avoidance leads to greater in-group association and identification, resulting in stronger in-group favouritism and collectivistic behaviour.

Intercultural Contact

Aside from cultural dimensions, contact with members of a different culture was also found to influence intergroup attitudes. The contact hypothesis, which was first proposed by Allport (1954) and later became very influential due to Pettigrew and Tropp’s (2006) meta-analysis, suggests that intercultural contact may reduce racism. Furthermore, Tadmor et al. (2012) demonstrated that contact with a specific cultural group does not only reduce discrimination towards members of that group but reduces discriminatory attitudes in general and thus enhances intergroup tolerance. Despite this, some of the studies—for example research by Graf et al. (2014)—highlight the importance of meeting certain criteria in order to improve intergroup attitudes, such as whether the experience with an out-group member is positive or negative. They demonstrated that negative contact predicted racism more strongly than positive contact predicted its reduction.

Regarding intercultural contact and ethnocentrism, findings are diverse. Since racism and ethnocentrism usually correlate, many studies suggest that intercultural contact reduces levels of ethnocentrism as well (Cakal & Petrović, 2017; Perreault & Bourhis, 1999). On the other hand, Kosmitzki (1996) found that contact with a different cultural group (while living in a foreign culture) may actually enhance ethnocentric attitudes, which he calls the reaffirmation of cultural identity. However, there is a lack of further research on cultural reaffirmation.

Slovenian Context

This study is set in the Slovenian context where the ethnic composition apart from Slovenians (83,1%) includes Croatians (1,8%), Serbs (2%), Muslims (1,6%), and Hungarians, Italians, Roma and others (less than 1%; Statistical Office of the Republic of Slovenia). Since the ethnic composition is quite homogenic, Slovenians usually have intercultural encounters and interactions via travelling or sojourning in foreign countries or by interacting with immigrants, tourists, and sojourners in Slovenia. For people living in urban areas such encounters are fairly frequent on a day-to-day basis compared to people in rural areas. With regards to cultural dimensions, Slovenia scored high (71) in power distance (SDO) and uncertainty avoidance (88) in Hofstede’s study (1980), which means that people generally support hierarchical order and have a preference for avoiding uncertainty. With the low score (27) on individualism dimension, Slovenia is considered a collectivist society. It needs to be kept in mind however, that Hofstede’s data were collected between 1967 and 1973 and more recent findings suggest that in line with modernization theory, the individualism scores are rising and power distance scores are dropping (Beugelsdijk et al., 2015).

The Present Study

This study seeks to examine the extent to which intergroup biases can be predicted by the cultural dimensions and by intercultural contact. There were two principal research questions:

  1. (1)

    Does intercultural contact, social dominance orientation, need for closure, and individualism/collectivism significantly predict in-group bias (ethnocentrism)?

  2. (2)

    Does intercultural contact, social dominance orientation, need for closure and individualism/collectivism significantly predict outgroup discrimination (neo-racism)?

In line with the cultural context described above and the reviewed literature, we may predict that social dominance orientation and uncertainty avoidance will be positively related to intergroup biases, but the relationship with collectivism is not prespecified. We also predict that greater intercultural contact will result in lower intergroup biases. However, we make no assumptions concerning the strength of predictive value of each variable.

Method

Design

This study had a cross-sectional, correlational design using a survey to explore the influence of four predictor variables (intercultural contact, social dominance orientation, individualism/collectivism, and uncertainty avoidance) on two criterion variables (in-group favouritism – ethnocentrism, and outgroup discrimination – neo-racism).

Sample and Procedure

Participants (N = 164, 68.3% female) were recruited through convenience and self-selective sampling. All were Slovenian nationals and aged over 18 years. The majority of participants were between 18 and 29 years old (40.2%). With regards to hometown population, both urban and rural areas were well represented, with 34,8% of the participants living in big cities and 31.7% living in villages with a population below 3.000 (category 1 and 2 combined). As far as educational qualification is concerned, 42.7% had a Bachelor's degree and 36% had a Master's degree (see Table 1 for demographic information).

Table 1 Summary statistics (raw scores) for neo-racism, ethnocentrism, intercultural contact, social dominance orientation (SDO), need for closure (NFC), horizontal (HI) and vertical individualism (VI), horizontal (HC) and vertical collectivism (VC), gender, age, hometown population, and education
Full size table

Data were collected through an advertised survey using the Slovenian online survey tool https://www.1ka.si/. A link to the online survey was posted in several Slovenian Facebook groups and was available on the websites of different Slovenian organizations. Participants took part on a voluntary basis. First, they were briefed about the aims, process and confidentiality of the study, the right to withdraw at any time during or after the study, and about the storage of the data. Participants were then asked to provide informed consent to participate. Next, they were instructed to indicate their level of agreement or disagreement with each presented statement on a Likert-type scales, and to answer multiple choice questions. Upon completion, participants were debriefed and thanked for their participation.

Measures

All measures were administered in Slovene. The back-translation method was used for translation from English language. The online survey included six scales (see below) and questions regarding demographic information, namely age, gender, education and hometown population.

Uncertainty Avoidance

To measure uncertainty avoidance on the individual level, the 15-item Need for Closure Scale (Roets & Van Hiel, 2011) was used. Items such as “I don’t like situations that are uncertain” were scored on a 6-point Likert-type scale ranging from strongly disagree to strongly agree. Subsequent analysis by Crowson (2013) confirmed the scale’s factorial validity and internal consistency (α = 0.87), and it was found to be a reasonable alternative to the 41-item NFC scale. The Cronbach’s alpha in this study was α = 0.83.

Social Dominance Orientation

SDO was measured using the 8-item Social Dominance Orientation Scale (Ho et al., 2015). It comprised of statements such as “Some groups of people are simply inferior to other groups”. Items were scored on a 7-point Likert-type scale ranging from strongly oppose to strongly favour. Negatively worded items were reverse-scored. It was developed as a short-version of the 16-item SDO scale, and the authors confirmed its validity and reliability across several different samples (Cronbach’s alpha between α = 0.78 and 0.90.). The internal reliability in this study was α = 0.77.

Individualism/Collectivism

As a measure of individualism/collectivism, the 16-item Individualism/Collectivism Scale (Triandis & Gelfland, 1998) was used. The scale measured four subdimensions: horizontal (HI) and vertical individualism (VI), and horizontal (HC) and vertical collectivism (VC). Each of the subdimensions included 4 items (e.g. “I'd rather depend on myself than others”). Items were scored on a 9-point Likert-type scale ranging from strongly disagree to strongly agree. Cozma’s (2011) review of individualism/collectivism measures revealed this is a generally reliable and stable measure (Cronbach’s alpha for subdimensions varied between α = 0.73 and 0.82), but highlights the possible cross-cultural issues. The Cronbach’s alpha for each subdimension in this study was α = 0.46 for HI, α = 0.65 for VI, α = 0.39 for HC, and α = 0.63 for VC.

Intercultural Contact

The intercultural contact of participants was measured using the Bidimensional Multicultural Experience Scale (Sparkman & Eidelman, 2018). This scale included 10 items which offer five categorical answers, for example “How many times have you travelled outside of Slovenia?” (1 = never, 2 = 1–2 times, 3 = 3–4 times, 4 = 5 times or more, 5 = On a regular basis). Authors reported alpha reliability between α = 0.70 and 0.83 across their three studies. The internal reliability of this scale in our study was α = 0.78.

Ethnocentrism

Ethnocentrism was measured using The Ethnocentrism Scale (Neuliep & McCroskey, 2013). This scale was composed of 22 items of which 15 were scored (e.g. “Most other cultures are backward compared to my culture”) and the other 7 functioned to balance the number of positively and negatively worded items. Items were scored on a 5-point Likert-type scale ranging from strongly disagree to strongly agree. Negatively worded items were reverse-scored. Authors of the original scale estimated the Cronbach’s alpha between α = 0.80 and 0.90 (Neuliep & McCroskey, 2013). The Cronbach’s alpha in this study was α = 0.88.

Neo-Racism

Neo-racism was measured using the 13-item Neo-Racism Scale (Tougas et al., 2004). Items were scored on a 5-point Likert-type scale ranging from strongly disagree to strongly agree. Negatively worded items were reverse-scored. Authors of the original scale estimated the Cronbach’s alpha between α =  0.80 and 0.90 (Neuliep & McCroskey, 2013). The Cronbach’s alpha in this study was α = 0.88.

Results

Data from the six scales and demographic questions were coded, entered, and analysed using the IBM SPSS Statistics 20 program. Individual mean scores were computed for each scale, and the overall mean scores were calculated. Cronbach’s alpha scores confirmed the internal reliability of all substantive scales except for the Individualism/Collectivism Scale which was, for this reason and non-normal data distribution, not included in any further analyses (see Table 1).

It is shown in Table 1, that the ethnocentrism score for the sample was 29.21 (SD = 7.88), which is at the mid-point of 30. A score above 55 indicates a high level of ethnocentrism (Neuliep, 2009), however none of the participants reached that score, implying that participants were not very ethnocentric. The neo-racism mean score was 44.49 (SD = 11.96), which is above the mid-point of 36 and suggests that participants were moderately prejudiced towards other ethnic groups. However, scores varied greatly as the range of individual mean scores was between 16 and 76. Overall, participants scored relatively high on the scale measuring intercultural contact (M = 35.18, SD = 6.26) as well as the NFC Scale (M = 54.05, SD = 10.61), but scored below the mid-point of 28 on the SDO Scale (M = 22.78, SD = 8.18).

To examine whether there were any significant bivariate correlations between these variables, a series of Pearson Correlation analyses were conducted. Results are seen in Table 2.

Table 2 Correlation matrix between neo-racism, ethnocentrism, intercultural contact, social dominance orientation (SDO), need for closure (NFC), and hometown population
Full size table

Gender, age, and education did not significantly correlate with neo-racism or ethnocentrism.

As per other variables, correlation analyses revealed that gender was significantly correlated with SDO, r (162) = -0.21, p < 0.01. However, an Independent Samples t Test showed no significant differences between males (M = 25.3, SD = 7.8) and females (M = 21.6, SD = 8.1) in SDO scores, t(103.24) = 2.83, p = 0.63.

As per our research questions, two multiple regressions were carried out to investigate whether the predictor variables (intercultural contact, social dominance orientation, and uncertainty avoidance) could significantly predict participants’ neo-racist attitudes and ethnocentrism, while holding all other variables constant. Among demographic variables, only hometown population was included in the predictive model for neo-racism, since they were found to be significantly related.

Regression 1 investigated whether intercultural contact, SDO, NFC, and hometown population significantly predict neo-racism (see Table 3). An analysis of standard residuals indicated that the data contained no outliers (Std. Residual Min = -2.4, Std. Residual Max = 2.8). The histogram of standardised residuals showed that the data included approximately normally distributed errors, as did the normal P-P plot of standardised residuals, which showed points that were not entirely on the line, but very close. The data met the assumption of independent errors (Durbin-Watson value = 1.83), the assumption of collinearity (NFC, Tolerance = 0.87, VIF = 1.2; SDO, Tolerance = 0.93, VIF = 1.1; Intercultural contact, Tolerance = 0.78, VIF = 1.3, Hometown population, Tolerance = 0.92, VIF = 1.1), and the assumption of nonzero variances (Neo-racism, Variance = 143.17; NFC, Variance = 112.64; SDO, Variance = 67.06; Intercultural contact = 39.24; Hometown population = 1.99). The scatterplot of standardised residuals indicated that the data also met the assumptions of homogeneity of variance and linearity.

Table 3 Multiple regression model analysis predicting neo-racism from intercultural contact, social dominance orientation (SDO), need for closure (NFC), and hometown population
Full size table

Using the enter method it was found that SDO, NFC, intercultural contact, and hometown population explained 34.5% of the variance in neo-racism scores (F (4, 159) = 20.97, p < 0.001) (see Table 3). While SDO (B = 0.686, p < 0.001) and hometown population (B = -1.322, p < 0.05) contributed significantly to the model, NFC (B = 0.146, p < 0.062) and intercultural contact (B = -0.182, p < 0.193) did not.

Secondly, regression 2 examined whether intercultural contact, SDO and NFC could significantly predict ethnocentrism (see Table 4). An analysis of standard residuals indicated that the data contained no outliers (Std. Residual Min = -2.3, Std. Residual Max = 2.3). Analysis also revealed that multicollinearity was not a concern (NFC, Tolerance = 0.87, VIF = 1.1; SDO, Tolerance = 0.93, VIF = 1.1; Intercultural contact, Tolerance = 0.82, VIF = 1.2). The data met the assumption of independent errors (Durbin-Watson value = 2.09) and the assumption of nonzero variances (Ethnocentrism, Variance = 62.19; NFC, Variance = 112.64; SDO, Variance = 67.06; Intercultural contact = 39.24). The histogram of standardised residuals and the normal P-P plot of standardised residuals showed that the data had approximately normally distributed errors. The data also met the assumptions of homogeneity of variance and linearity as indicated by the scatterplot of standardised predicted values.

Table 4 Multiple regression model analysis predicting ethnocentrism from intercultural contact, social dominance orientation (SDO), and need for closure (NFC)
Full size table

As presented in Table 4, the results of the regression 2 showed that the model was a significant predictor of ethnocentrism (F (3, 160) = 18.47, p < 0.001), accounting for 25.7% of the variance. SDO (B = 0.376, p < 0.001) and intercultural contact (B = -0.304, p < 0.01) significantly contributed to the model, while NFC did not.

Taken together, present results demonstrated that neo-racism scores were best predicted by SDO scores and the population of the hometown, while ethnocentrism scores were best predicted by SDO scores and intercultural contact.

Discussion

This study set out to examine whether neo-racism and ethnocentrism can be predicted by social dominance orientation (power distance), uncertainty avoidance (need for closure), individualism/collectivism, and by intercultural contact. Each of the predictor variables in relation to neo-racism and ethnocentrism will be discussed in turn.

SDO was found to have the strongest relationship to neo-racism and ethnocentrism among all predictors. This is consistent with the predictions of social dominance theory (Pratto, 1999), which posits that one of the main forces maintaining group inequalities is cultural ideology including shared beliefs, social role prescriptions, and institutional discrimination. Beliefs and group stereotypes which are collectively shared among members of a certain culture may seem true due to social-psychological phenomena such as false consensus bias and self-fulfilling prophecies. Furthermore, institutional discrimination supports the status quo of the dominant group through legal and extra-legal means. This is reflected in the social policies that increase social hierarchy as well as in the work of the police, the military, and judicial authorities—which should not be assumed to be objective and impartial. In Slovenia, systemic discrimination is present especially in a form of ethnic profiling by the police (Ambasada Amnesty International, 2020; Rog, 2021). These practices, together with shared cultural stereotypes, maintain group inequalities and endorse the occurrence of racist attitudes in Slovenia.

As for the predictive value of SDO on ethnocentric attitudes, the present results are consistent with social identity theory (Tajfel & Turner, 1986) which explains that the greater the in-group identification, the greater the positive in-group affect, together with the allocation of social value to one’s own group. The motivation of an individual to maintain positive social identity thus reflects in hierarchy-enhancing ideology (seeing one’s own group as superior). Previous research shows that this effect is especially salient when the relative status of the group is high (Levin & Sidanius, 2003), when boundaries between the groups are seen as legitimate (Levin et al., 2002), and when there is a competitive context (Pratto & Glasford, 2008). This implies that perhaps SDO is a good predictor of ethnocentric attitudes among Slovenians due to meeting some of the above conditions. The relative status of Slovenians compared to e.g. immigrants is high and the status differences are perceived as fair and legitimate by many (the distinctive characteristic between the groups may be holding a Slovenian passport). The competitive context might be understood in light of the rising number of immigrants coming to Slovenia (Statistical Office, 2019), which may be seen as competition for resources (e.g. job opportunities).

Intercultural contact was negatively correlated with both neo-racism and ethnocentrism. This is in line with the contact hypothesis (Pettigrew & Tropp, 2006) which claims that individuals who have more contact with members of other cultural groups exhibit less discriminatory beliefs compared to individuals with fewer intercultural encounters. Since most of the research has used correlational design, one could argue that the negative association between intercultural contact and out-group discrimination might be because less biased individuals are more likely to seek intercultural experiences. However, Tadmor et al. (2012) experimentally manipulated intercultural exposure and demonstrated its causal effect on reduced discrimination. Despite this, in the present study, intercultural contact was not an important predictor of neo-racism after controlling for the effect of other predictor variables, but it did significantly predict ethnocentrism.

Interestingly, in previous research the effect of intercultural contact on ethnocentrism seems to vary across cultures (and might be dependent on cultural history and ideologies) since this effect was found in certain cultures, but not in others. For example, Neuliep et al. (2001) found the influence of intercultural contact on ethnocentrism in the American sample (where, similar to the present study, participants who experienced intercultural contact had lower ethnocentrism scores than those without contact) but no such differences were observed among Japanese participants. This was attributed to the Japanese history of political and geographical isolation and policies restricting contact with foreigners, which resulted in the nurturing of in-group favouritism and makes Japanese culture less prone to the effects of intercultural contact. On the other hand, ethnically diverse cultures such as America are thought to be more open and willing to assimilate differences. This might as well be the case for cultures with historically richer and more diverse multicultural contact, such as Slovenia, whose territory was part of several ethnically different states including the Roman Empire, Austro-Hungarian Empire (with incursions by the Ottoman Army) and Yugoslavia. The influence of intercultural contact on ethnocentric attitudes might thus be greater among cultures with historically diverse cultural contact, who are prone to acknowledging other culture’s values and beliefs or which even acculturate.

Uncertainty avoidance, measured with NFC, correlated positively with neo-racism. This is consistent with previous findings, suggesting that individuals with high NFC, who tend to avoid ambiguity by drawing quick conclusions, are more likely to make overgeneralised conclusions and thus are more inclined to stereotype and discriminate (Brizi, et al., 2015; Todor, 2014). Although there was a positive correlation, uncertainty avoidance was not found to be an important predictor of neo-racism after controlling for the effect of other predictor variables. This suggests that power distance does not play a direct and important role in forming neo-racist attitudes. Moreover, we did not find significant correlation between NFC and ethnocentrism, although Hogg’s (2007) uncertainty-identity theory and some previous studies (Fischer & Derham, 2016) propose the positive link between uncertainty avoidant context and in-group favouritism. These results imply that when forming ethnocentric attitudes, socio-psychological processes other than uncertainty reduction might be at work.

With regards to significant bivariate correlations, this research found that intercultural contact was negatively correlated with SDO, which implies that participants who experienced more contact with other cultural members showed less support for group-based hierarchies. This is in line with the findings of Shook et al. (2016) who demonstrated the causal effect of intercultural contact (students living in interracial rooms versus same-race rooms) on the reduction in SDO levels, which they measured at the beginning and at the end of the semester. Furthermore, the present study found a negative correlation between intercultural contact and NFC. This implies, as suggested by Tadmor et al. (2012) on the basis of their experimental findings, that this motivated cognitive process of drawing quick conclusions in order to avoid ambiguity may play a mediating role in forming intergroup attitudes.

Among demographic variables, only hometown population significantly predicted neo-racism, where participants from rural areas had higher neo-racism scores than participants from urban areas. Although Slovenia is a very small country, the differences in beliefs and values between people from rural and urban areas persist, as previous research shows (Černigoj, 2016). This might be due to the lack of encounters between autochthonous people and minorities in smaller villages (compared to more encounters in the cities), considering that hometown population was found to correlate with intercultural contact. The differences in the expression of racist attitudes might also contribute to higher neo-racism scores in rural areas, since rural populations might more openly express xenophobic views, whereas in an urban context of more diverse community, such beliefs are perhaps less socially acceptable and therefore concealed. Lastly, realistic conflict theory (Sherif, 1954) proposes that discrimination occurs when there is a perceived threat. According to this theory, rural people, compared to urban people, may perceive minorities as more threatening to group cohesion, stability, and the values of their small and thus vulnerable communities.

Conclusion

To summarise, this study has demonstrated that social dominance orientation was the best predictor of both neo-racism and ethnocentrism. This implies that cultures that score high on this dimension are especially prone to discrimination which may lead to intergroup conflicts. Intercultural contact was also a significant predictor of ethnocentrism, suggesting that apart from cultural dimensions, individual experiences also play an important role in shaping intergroup attitudes. Our findings, together with previous research, suggest that certain factors influencing the formation of negative attitudes must be identified and targeted in order to improve social tolerance, but that these factors may vary across different cultural contexts.

Limitations and Future Directions

We acknowledge the limitations of the present study. In order to make cross-cultural comparisons of cultural dimensions and general attitudinal scores, different sampling techniques should have been used to gain a representative sample of a certain population. For example, the results indicate low overall SDO (which is not consistent with Hofstede’s findings for Slovenia), but perhaps this is due to the sample including a disproportionally lower percentage of male participants, who (as per our results as well as the previous studies from different cultures; Pratto et al., 2000) score higher than woman on the SDO measure. Thus, if the sample was balanced for gender, the overall score of SDO might have been different. In these terms, present findings on the overall scores could at best be treated as indicative.

Furthermore, present research only investigated three cultural dimensions (of which one – individualism/collectivism was excluded from the analysis due to statistical violations). In order to examine which cultural dimension can best predict in-group and out-group biases, other cultural dimensions, as per Hofstede’s (2001) or Schwartz’s (1994) classification, should be included in future research.

Concerning the results of Individualism/Collectivism Scale (Triandis & Gelfland, 1998) which yielded low internal reliability and non-normal data distribution, a few methodological shortcomings may apply. Firstly, the Slovenian translation of the scale was not validated prior to the study. Secondly, the psychometric properties of the original scale itself have certain issues, such as an issue with form equivalence and invariance, and a problematic distinction between horizontal and vertical dimensionality (Li & Aksoy, 2007). Investigators of the subsequent analysis also suggested that the back-translation method is not sufficient to ensure similar meaning across cultures in certain items and that additional methods should be employed (Li & Aksoy, 2007).

Since cultural values together with intergroup attitudes are a subject of change over time (Beugelsdijk et al., 2015; Tashakkori &Thompson, 1988), there are limitations with regards to the cross-sectional design of the present research. Only longitudinal data could provide an insight into which cultural dimensions have changed over time and how these changes impacted in-group and out-group biases, hence establishing cause and effect relationships (especially since it is difficult to manipulate cultural dimensions and as such is not feasible to assess the causal relationship by applying the experimental design). Moreover, another shortcoming of the chosen method is self-report data, which can be subject to several biases such as social desirability bias (which might have occurred in questionnaires assessing socially undesirable traits, e.g. racism), diverse introspective abilities and diverse rating styles (prone to extreme or neutral answers). Complementary qualitative assessment would be beneficial in order to minimise those biases.

On a practical level, the results of the present study reinforce support for the idea of culturally-informed practices for preventing and reducing discrimination, which take into account specific cultural aspects such as cultural history, ideology and values. Additionally, they highlight the importance of targeting SDO reduction as a means of reducing racism and ethnocentrism. As our findings suggest, promoting positive intercultural contact may help to reduce ethnocentrism in some cultural contexts, but may not be effective in others. Similar findings were obtained by Kende et al. (2018), who found that intergroup contact was an effective prejudice-reduction strategy in egalitarian societies but was less effective in societies high in SDO. For this reason it is important to be aware of the cultural context in which intergroup biases occur. The persistent issue with research on prejudice-reduction interventions (see Aboud et al., 2012 for a review) and, more specifically, with research on strategies targeting the reduction in SDO (Dambrun, et al., 2009; Danso et al., 2007), is that the majority of studies were conducted in WEIRD (Western, educated, industrialised, rich and democratic) societies and as such, do not offer information as to whether those strategies are also effective in non-Western societies. More research is therefore needed to find appropriate strategies for reducing SDO and prejudice in different cultural contexts.