Awareness of the psychological bias of naïve realism as a subtle strategy for improving stereotypes towards Moroccan women in Spain

The present research explored whether the exposure to the psychological bias of naïve realism (i. e., the conviction that one ’ s own views are objective, and the other ’ s views are biased) could promote stereotype change along the dimensions of morality, sociability, and competence. Due to their double discrimination, we considered Moroccan women living in Spain as the target group. Participants (307 Spaniards; M age = 35.92, SD = 13.37; 61.6% women) were randomly assigned to a naïve realism (vs. control) condition and were further asked to evaluate Moroccan or Spanish women. Findings reveal that Moroccan women were evaluated as more moral, sociable and competent in the naïve realism than in the control condition. The ingroup favouritism towards Spanish women (compared to Moroccan women) disappeared in the moral dimension: once aware of the naïve realism bias, participants perceived Moroccan women as honest, sincere and trustworthy as Spanish women. Taken together, our findings suggest that making people aware of the bias of naïve realism might be beneficial for intercultural relations as it contributes to enhance outgroup evaluations and reduce the ingroup bias on morality, a core dimension in social perception. Moroccans are the largest immigrant group 16.1% of 2021). They are discriminated in the This discrimination might be closely linked to the existing stereotypes about this ethnic group, being the most devalued immigrant Spain L 2013; L 2014). The present study aimed at testing whether the exposure to psychological bias of naïve realism could improve the perception of Moroccan women living in Spain. In doing so, our work aims at exploring whether such a subtle cognitive strategy may be functional to promote stereotype change.


Introduction
Labor discrimination is a challenging reality for different ethnic minorities and women around the world (e.g., Cuadrado, García-Ael, & Molero, 2015). Consequently, immigrant women face double discrimination. In Spain, Moroccans are the largest immigrant group (the 16.1% of foreign residents in this country; INE, 2021). They are discriminated in different areas, especially in the labor market, where Moroccan women are even more vulnerable: more unemployment, lower wages, more temporary employment, less occupational mobility (Moreno-Colom & Loṕez-Roldań, 2018). This discrimination might be closely linked to the existing stereotypes about this ethnic group, being the most devalued immigrant community in Spain (e.g., López-Rodríguez, Cuadrado, & Navas, 2013;López-Rodríguez, Navas, Cuadrado, Coutant, & Worchel, 2014). The present study aimed at testing whether the exposure to psychological bias of naïve realism could improve the perception of Moroccan women living in Spain. In doing so, our work aims at exploring whether such a subtle cognitive strategy may be functional to promote stereotype change.
Ethnic stereotypes might be the substrate of multiple forms of discrimination. People are evaluated in two main dimensions of social perception, warmth and competence (e.g., Fiske, Cuddy, Glick, & Xu, 2002). The warmth dimension serves to evaluate others' intentions, a basic process from evolutionary perspective, and it is shaped by judgments of sociability and morality (Leach, Ellemers, & Barreto, 2007). Morality, indeed, constitutes a core dimension in social perception (Brambilla & Leach, 2014) and take a prominent role when evaluating both the ingroup (Leach et al., 2007) and the outgroup (for a review, see Brambilla, Sacchi, Rusconi, & Goodwin, 2021). When judging the morality of social targets, we are posing them in a continuum of benevolence and deciding if we can trust them. Moroccan immigrants in Spain are believed to be less sociable, competent, and specially less moral than other immigrant groups (e.g., López-Rodríguez et al., 2013). This perception might justify a worst treatment towards them compared to Spanish or other immigrant groups. Finding ways to increase the perception of morality, sociability and competence of Moroccans can open new pathways to improve the intercultural relations between Moroccans and Spaniards.
Modifying stereotypes is not an easy task. Different strategies can be used to alter ethnic stereotypes, such as receiving a multicultural education, recategorizing the ingroup and the outgroup in a superordinate group (Gaertner & Dovidio, 2000) or presenting counter stereotypical information about members from stigmatized ethnic groups (Ramasubramanian, 2011). Experiences of intergroup contact can also be a source for disconfirming previous stereotypes and learning about the others (Brambilla, Ravenna, & Hewstone, 2012;Pettigrew & Tropp, 2006). However, most of these strategies are not easy to implement and multiple obstacles can prevent administrations and policymakers from translating such strategies into real policies. The segregation of ordinary life prevents ethnic encounters (Dixon et al., 2020) that could facilitate having positive and pleasant interactions with Moroccans. Some strategies need a systematized implementation, and their positive effects would be visible in the long-term. Other interventions might be ineffective because they rise reactance among the target of influence, especially under the xenophobic discourse of several political parties in polarized societies. These difficulties open to new ways to understand prejudice and to modify ethnic attitudes that could be easily implemented and complement previous strategies by avoiding reactance in polarized environments.
Here, we propose that ethnic stereotypes can be altered by using a subtle and indirect cognitive strategy orientated to alter our biased processing. Our thoughts, attitudes and behaviors are influenced by unconscious psychological biases, which affect the way we perceive ourselves and the world around us. In addition, we are convinced that others are more susceptible to be biased than us (Pronin, Lin, & Ross, 2002). This (erroneous) assumption can be explained by a universal psychological bias, the naïve realism (Ross & Ward, 1996), which refers to people' s tendency to consider their own perceptions and judgments to be objective reflections of reality while considering others' views to be irrational and biased.
Some authors have shown that making people aware of the influence of unconscious psychological biases leads them to be less affected by such biases (e.g., Pronin & Kugler, 2007;Rand, 2003). Therefore, one method that can be used to counteract such biases is making people aware of their existence in order to control and correct them. This is precisely what Nasie, Bar-Tal, Pliskin, Nahhas, and Halperin (2014) tested when they analyzed the effect of awareness of naïve realism on openness to the narratives of the outgroup in the context of the Israeli-Palestinian conflict. Their findings supported that becoming aware of the naïve realism bias and its effects in conflicts made participants more open to the narrative of the adversary, especially for those participants with a hawkish political ideology.
Extending such evidence, López-Rodríguez et al. (2021) tested the effect of a simplified version of this intervention in the context of intercultural relations between Spaniards and Moroccan immigrants. To explicitly avoid any reactance to the intervention, only an explanation of the psychological bias was presented, without any relation to intercultural conflicts. This way, participants cannot easily link the cognitive intervention to intergroup attitudes. This research confirmed that the sole awareness of the psychological bias of naïve realism generated more acceptance of cultural differences of Moroccan immigrants among Spanish participants.
Our main goal was to explore whether the sole awareness of the naïve realism bias can alter how people perceive social groups on basic content dimensions. Indeed, prior work on naïve realism have not tested whether such a strategy may be helpful to change how we stereotype social groups. Specifically, we tested whether exposure to naïve realism may help to change the stereotypes of Spaniards towards Moroccan women along the fundamental dimensions of morality, sociability, and competence. We also aimed at exploring whether such a strategy could reduce the ingroup bias towards Spanish women compared to Moroccan women, especially in the case of morality, the primary dimension of group perception .

Participants
After controlling for completeness and duplicates, 353 residents in Spain volunteered to participate in this study. Criteria for exclusion of participants were pre-established: giving a wrong answer to at least one of the four items of manipulation checks (n = 34) or being born outside of Spain (n = 12). The final sample comprised 307 participants (61.6% women) who ranged from 18 to 79 years (M age = 35.92, SD = 13.37). Most of the participants were actually working (76.5%) and had a university education (65.5%). A sensitivity analysis using G*Power 3 program (Faul, Erdfelder, Lang, & Buchner, 2007) revealed that with a sample size of 307 participants, an ANOVA (fixed effects, special, main effects and interactions) had 80% power to detect a small-to-medium minimum effect size of f = 0.160 (η 2 p = .025). In our study, all the significant effects reported in the results section (except one:.018) were above η 2 p = .025. These sizes are consistent and in line with those obtained in Social Science research.

Procedure
Participants were selected from the general population by convenience sampling and were asked to fill an online questionnaire. Specifically, second-year undergraduate Psychology students sent a link to their acquaintances in exchange for credits (0.25 points) following a snowball procedure. The study was approved by the authors' University Ethics Committee. Participation was voluntary and anonymous, and participants were adequately informed of their rights and provided their informed consent. Following an intergroup design of 2 (Experimental manipulation: naïve realism vs. control) × 2 (Women's ethnicity: Moroccan women vs. Spanish women), participants were randomly assigned to either a condition to raise awareness of the naïve realism bias or a control condition, and to evaluate either Moroccan women or Spanish women. Participants in the naïve realism condition (n = 137; 71 evaluated Moroccan women and 66 Spanish women) were induced to be aware of this psychological bias by reading a text describing such a bias (see Supplementary Materials). Participants in the control condition (n = 170; 85 evaluated Moroccan women and 85 Spanish women) were asked to read a text about a process of organizational psychology (see Supplementary Materials). Samples in conditions were imbalanced because of the selection criteria based on the manipulation checks in the naïve realism condition.

Measures
A detailed description of each measure can be found in Supplementary Materials. The response scale for the items ranged from 1 to 5.

Manipulation checks
We used the four-item true/false scale adapted by López-Rodríguez and colleagues (2021) from Nasie et al. (2014) to assess if participants in the naïve realism condition had understood the text.
In order to be sure about the equality of participants' initial levels of prejudice and contact with Moroccans in the different conditions, we also measured:

Quantity and quality of contact
Participants had to indicate, with a single item each, the quantity and quality of contact they had with Moroccan immigrant women or Spanish women.
Additionally, tolerance towards Moroccan immigrants (Cuadrado, Ordóñez-Carrasco, López-Rodríguez, Vázquez, & Brambilla, 2021), openness to Moroccan immigrants' narrative regarding integration , and support for affirmative action policies in the workplace towards Moroccan immigrants and Moroccan immigrant women (created ad hoc) were measured as exploratory dependent variables in order to test the effect of the experimental manipulation (naïve realism vs. control) on them. The description of these variables can be consulted in Supplementary Materials.
Regarding socio-demographic data, participants indicated their sex, age, level of education, main activity, birth country, and political orientation.

Preliminary Analyses
Before conducting the main analyses, we examined whether there were significant differences between participants as a function of the experimental manipulation (control vs. naïve realism) and women's ethnicity (Moroccan women vs. Spanish women) in their sociodemographic characteristics, initial level of modern racism and quantity and quality of contact with the women target. We did not find a multivariate experimental manipulation × women's ethnicity significant effect on modern racism, political orientation, participants' age nor quantity and quality of contact, Wilk's λ = 0.99, F(5, 299) = 0.525, p = .757, η 2 p = .009. The univariate effects of the interaction on each variable were also not significant (ps > 0.256). We also found no significant differences between participants depending on experimental manipulation × women's ethnicity across sex, education nor main activity (ps > 0.058).

Effects of naïve realism on stereotypes and ingroup bias
We conducted a MANOVA in order to test the effect of the experimental manipulation and women's ethnicity on the evaluations of morality, sociability and competence. The two-way interaction between experimental manipulation and women's ethnicity was significant, Wilk's λ = 0.97, F(3, 301) = 2.88, p = .036, η 2 p = .028. The univariate analyses and pairwise comparisons revealed that Moroccan women were evaluated as more moral, F(1303) = 5.45, p = .020, η 2 p = .018, sociable, F(1303) = 7.70, p = .006, η 2 p = .025, and competent, F(1303) = 4.40, p = .037, η 2 p = .026, in the naïve realism than in the control condition (see Fig. 1). The evaluations of Spanish women on these three dimensions did not significantly vary depending on the experimental manipulation (ps > 0.189).
We also found a multivariate main effect of women's ethnicity on stereotypes, Wilk's λ = 0.80, F(3, 301) = 24.67, p < .001, η 2 p = .197. The univariate analyses and pairwise comparisons showed that Spanish women were perceived as more moral, sociable and competent than Moroccan women (ps < 0.001), confirming an ingroup favouritism towards Spanish women. As shown in the steepness of the slopes of Fig. 1, this favouritism diminished for participants who were aware of the naïve realism bias, and even disappeared in the moral dimension, F(1303) = 2.17, p = .142, η 2 p = .007.

Effects of naïve realism on tolerance, openness to narratives, and support for affirmative action
The multivariate effect of the experimental manipulation (naïve realism vs. control) on tolerance, openness to narratives of Moroccan immigrants about integration, and affirmative action towards Moroccans/Moroccan women -including as covariates political orientation and level of modern racism of the participants-was no significant, Wilk's λ = 0.99, F(3, 301) = 1.45, p = .230, η 2 p = .014. The univariate effects of the experimental manipulation on each variable were also not significant (ps > 0.136). Fig. 1. Two-way interaction between experimental manipulation and women's ethnicity on perceived morality, sociability and competence.

Discussion
This study explored whether the sole awareness of the psychological bias of naïve realism can promote stereotype change. In doing so, we considered Moroccan women since they are highly devaluated in Spain. We thus tested whether the awareness of the psychological bias of naïve realism enhanced perceptions along the dimensions of morality, sociability, and competence (Brambilla & Leach, 2014) for Moroccan women as well as reduced the ingroup bias towards Spanish women compared to Moroccan women.
These findings reveal a novel and relevant effect of this subtle intervention on stereotype change. They extend previous research about the contextualized effects of the awareness of the psychological bias of naïve realism on the openness to the narrative of the adversary in the Israeli-Palestinian conflict (Nasie et al., 2014), as well as on the acceptance of cultural differences of Moroccan immigrants by Spaniards . Concretely, we show that the mere awareness of the naïve realism bias leads participants to improve the stereotyped evaluations of morality, sociability and competence towards Moroccan immigrant women. This has obvious positive implications for improving the intercultural relations between Spaniards and Moroccan women which, in turn, could improve the general perception of Moroccans as a whole group. Hence, these findings provide us of an interesting, subtle, fast and easy intervention to improve the perception of members of devaluated outgroups. The replication of these results with Moroccan immigrant men, as well as with other groups and in different contexts, might entail a stronger evidence to support the effect of this manipulation on outgroup stereotypes.
Moreover, our findings show that this intervention also reduces the omnipresent ingroup favoritism bias (Tajfel & Turner, 1979) on the evaluations of competence and sociability, and even such bias disappears when morality is evaluated. That is, once aware of the naïve realism bias, participants perceived Moroccan women as honest, sincere and trustworthy as Spanish women. It is a promising and noteworthy finding since morality is especially relevant in social perception because it foreshadows the essential judgment of whether someone represents an opportunity or a threat (for reviews see Brambilla & Leach, 2014;Brambilla et al., 2021). Previous research has also evidenced that people consider morality as the most important characteristic for feeling good about one's ingroup, and that perceived ingroup morality was the strongest predictor (compared to sociability and competence) of pride in the group (Leach et al., 2007). Given the prominent role of morality in ingroup and outgroup evaluations, the fact that awareness of the naive realism bias equals participants' evaluations towards ingroup women with those towards women from a devalued outgroup, opens up a promising avenue of research about future applications of this simple intervention to improve intergroup relations.
Morality has a prominent impact not only on social perception, but also on behavioral responses. Indeed, a recent research conducted in Spain (Cuadrado, López-Rodríguez, & Constantin, 2020) has shown that only moral stereotypes consistently predict facilitation behavioural tendencies towards Moroccans (via positive emotions). Thus, evaluating Moroccan women as moral as Spanish women have also positive implications regarding helping intentions towards this outgroup.
We also found that our manipulation did not affect openness to Moroccan immigrants' narrative, tolerance and support for affirmative action policies towards Moroccan immigrants. By contrast, Nasie et al. (2014) confirmed that their intervention had benefits for considering the outgroup narrative. However, while Nasie and colleagues provided participants a more articulated description of the naive bias, detailing that it can intensify conflicts and misinterpretations between individuals and groups, here we opted for a simplified version of the manipulation and we only made participants aware of the bias without providing information that could activate group categorization or ethnic attitudes. This might explain the difference between our findings and those reported by Nasie and colleagues. Supporting this claim, López-Rodríguez et al. (2021), whose simplified manipulation we replicate here, also reported no effects on openness to Moroccan immigrants' narrative regarding integration. The influence of this intervention had not been tested so far on tolerance and support for affirmative action policies. Thus, it is possible that the sole exposure to a psychological bias may not be sufficient to modify participants' endorsement of ideological variables, and more complex manipulations are needed (as that reported by Nasie and colleagues).

Conclusions
This study suggests that a subtle intervention entailing the mere awareness of the naïve realism bias (a key principle in Social Psychology), and not including explicit reference to intercultural relations, can improve outgroup evaluations and reduce ingroup favoritism. Training in critical thinking can contribute to gain awareness of our tendency to perceive the world in a unilateral and simplistic manner, which may encourage people to a less categorical perception of others and, in the long-term, may benefit intercultural relations.

Funding
This work was supported by the Ministry of Science and Innovation of Spain [grant number PRX19/00263]. This publication is part of the project PID2019-105114-GB-I00, funded by MCIN/ AEI /10.13039/501100011033, and the project P18-RT-668 funded by CECEU/FEDER.

Declaration of Competing Interest
None.