Introduction

A plethora of studies highlight how the didactic actions of teachers are hampered when they also hold ethnic and homophobic prejudice (D’Urso & Petruccelli, 2022; Kennedy, 2010; Vervaet, et al., 2016), which also undermine an inclusive and diversity friendly school (Pace et al., 2021; D’Angelo & Dixey, 2001). How is it possible that some teachers can hold ethnic and homophobic prejudice when the basic values of democratic societies that are based on human dignity and human rights speak clearly against them? Bandura's (1999, 2002a, b) cognitive theory of moral agency offers a promising framework to study this topic. This theory states that individuals engage in specific cognitive processes to convince themselves that harmful actions which violate either their own moral standards, or the moral standards of society, are actually justified in some way. Although most teachers endorse human dignity and are highly committed to human rights (Osler, 2016), self-justification processes (i.e., moral disengagement) might allow them to hold negative attitudes towards immigrants and sexual minorities that are in contrast with moral standards (Maftei & Holman, 2022). Religiosity is another factor that might add to the complexity of these processes by amplifying the importance of moral disengagement for self-justifying prejudice and negative attitudes that are also clearly against the basic values of religions, even if many churches disapprove same-sex relationships.

The Role of Prejudice

Prejudice is defined as an antipathy or a negative judgment based on a false and inflexible generalization that can be felt or expressed (Allport, 1954). Prejudice can be considered attitudes, and as such, they are transmitted socially and maintained in societies. Furthermore, prejudice can be directed towards a group as a whole or towards an individual member of that group. Prejudice neglects the diversity within social groups but accentuate the differences between social groups. In this sense, the ingroup provides a system of values and information about the world and others (as a shared and socially validated reality), and the outgroup represents something different and threatening (Hogg, 2000). Pettigrew and Meertens (1995) differentiated blatant (manifest or direct) prejudice from subtle (latent or indirect) prejudice. Blatant prejudice is related to a perception of threat towards the outgroup that induces feelings of contempt for intimate relationships with outgroup members and a desire to avoid outgroup members to safeguard social identity. Subtle prejudice is instead studded with attitudes and beliefs in defense of the traditional values of the ingroup, the exaggeration of cultural differences between groups, and the denial of positive emotions towards the outgroup. Subtle prejudice is more common than blatant prejudice in democracies that do not consider the open expression of hostile attitudes against minority groups as legitimate (Bratt et al., 2016; De Caroli et al., 2013). Homophobia is, on the other hand, the set of negative feelings and attitudes towards people who belong to sexual minorities (Lal & Garg, 2020; Petruccelli et al., 2015). The literature underlines how it can compromise the implementation of inclusive practices, since it hinders the open-mindedness of teachers (D'Urso et al., 2023).

The Role of Moral Disengagement

Moral disengagement processes are activated when a person behaves in contrast to internalized moral standards. Studies show that the majority of adults, including teachers, endorse human dignity and are committed to human rights (Florian & Camedda, 2020; Osler, 2016; Sison et al., 2016), yet still a substantial number of teachers are also holding prejudice about immigrants or sexual minorities (D’Urso & Symonds, 2021; Strohmeier & Gradinger, 2021). Hence, it can be assumed that holding ethnic prejudice and homophobic attitudes create a cognitive dissonance, because they are in sharp contrast to these internalized moral standards. According to Bandura's theorizing this cognitive dissonance can be reduced by using self-justification processes, which allow people to morally disengage from actions that harm others (Bandura, 2002b). Bandura proposed that individuals develop moral standards about what is right and wrong and that these standards act as moral guides for social conduct. When a distance exists between an individual's moral standards and their attitudes (e.g., prejudice), the person might experience moral emotions such as guilt and shame as outcomes of this dissonance. Moral disengagement operates to avoid these negative feelings and related self-censure by cognitively reframing the situation and one's own actions, so that they appear congruent with internalized standards. The moral self is embedded in a broader social-cognitive self-theory encompassing self-organizing, proactive, self-reflective, and self-regulative mechanisms. These self- referent processes provide the motivational as well as the cognitive regulators of moral conduct (Bandura, 1999). Importantly, moral disengagement mechanisms operate at four levels: (1) the interpretation of harmful practices (i.e., minimizing personal responsibility), (2) the interpretation of harmful effects (i.e., cognitive restructuring), (3) the representation and interpretation of the results (e.g., to minimize, ignore, or distort the consequences), and (4) the representation of victims (e.g., to dehumanize and/or attribute blame to the victim).

Because people tend to use moral disengagement habitually across a variety of situations, moral disengagement can also be understood as a personal characteristic or trait, capable of influencing individual beliefs enabling harmful actions (Bandura, 2018). A large body of research has previously demonstrated the disinhibitory power of moral disengagement and its strong associations with several forms of aggressive behaviour and deviant conduct (e.g., Almeida et al., 2009; Bandura, 1999, 2002a; Detert et al., 2008; Fida et al., 2015; Moore et al., 2012). Furthermore, studies with young adults have found that moral disengagement processes are also positively associated with homophobic attitudes (Camodeca et al., 2019; Maftei & Holman, 2022). Studies on the moral disengagement of teachers are rather rare. However, a recent study found that moral disengagement mechanisms related to the representation of victims (e.g., the tendency to dehumanize victims and attribute blame to them) were positively associated with homophobic attitudes among Italian primary and secondary teachers (D’Urso & Symonds, 2021). Similarly, Strohmeier and Gradinger (2021) found that teachers with high levels of moral disengagement were more likely to ignore episodes of peer harassment towards ethnic and religious minorities. Thus, there is already evidence suggesting that moral disengagement processes are related to ethnic and homophobic prejudice that undermine openness to diversity and acceptance of others.

The Role of Religiosity

The literature has also highlighted the critical and controversial role of religiosity in relation to prejudice and negative attitudes toward sexual minorities (Burch-Brown & Baker, 2016). Indeed, some studies have found higher levels of prejudice against ethnic and sexual minority groups among individuals who hold strong religious feelings or beliefs (Batson et al., 1993; Hall et al., 2010; Johnson et al., 2012; Whitley, 2009). Many religions including the Roman Catholic Church openly oppose romantic same-sex relationships, and in some religious communities ethnic minorities might be constructed as believers to the wrong god. As such, religiosity might play an important role in the implementation of social hierarchies and social disparities (Fredrickson, 2002; Harvey, 2016) that are maintained by holding prejudice towards ethnic and sexual minorities. Although some research has not found a link between religiosity and homophobic prejudice (Petruccelli et al., 2015), others have suggested that religiosity can increase helping behaviours towards the ingroup (Preston et al., 2010). However, a study conducted by Johnson et al. (2012) showed that religion among adult participants may simultaneously promote derogation towards out-groups and an increase in favouritism within the in-group. Moreover, other research has found that religiosity was associated with lower levels of discriminatory attitudes, indicating that not all believers share the same strong religious dogmas (Hall et al., 2010).

Although some previous work has examined the links between prejudiced attitudes and either religiosity or moral disengagement processes, few studies have investigated how the relationship between religiosity and moral disengagement interact to influence ethnic prejudice and homophobic attitudes. In this sense, the existing literature highlights that among those who follow religious precepts, dehumanization, a strategy of moral disengagement that implies the perception of a person as devoid of humanity, can be perceived as contradicting one of the core premises of many religions (i.e., that God has created humans in his image). Among non-religious people, however, dehumanization contradicts a universalistic perspective of human dignity and human rights. Therefore, moral disengagement processes can have different meanings, yet the same effects on prejudice for religious or non-religious people (Sverdlik & Rechter, 2020).

The Current Study

Teachers have an important role to establish diversity-friendly and inclusive school environments that offer equal chances to all learners independent of their gender, ethnic or religious background, or sexual orientation. To not discriminate students based on any group membership is a basic human right and ideally schools should be places where these core values are put into practice. However, some teachers do hold prejudice against ethnic and sexual minorities even though most of them also endorse democratic societal values of human dignity and human rights for all. Moral disengagement is a cognitive process that functions to reduce the cognitive dissonance that arises when core moral values (e.g., human dignity) and attitudes (e.g., prejudice and negative attitudes) are incompatible. The current study investigates (1) whether moral disengagement is associated with blatant ethnic prejudice, subtle ethnic prejudice, and homophobic prejudice (main effects), and (2) whether religiosity moderates these relationships (interaction effects). These two hypotheses want to bring out how morality and religiosity may explain prejudice and negative attitudes in order to design targeted training courses.

Overall, we expect positive associations between moral disengagement and the three forms of prejudice, because blatant and open ethnic prejudice and homophobic prejudice are in contrast with human dignity and human rights, values that are highly endorsed by most teachers (Florian & Camedda, 2020; Sison et al., 2016). Based on Bandura’s theorizing all three forms of prejudice should induce feelings of guilt and shame, as well as cognitive dissonance. Thus, we expect that it will be easier for people who are habitually high on moral disengagement to cope with these moral emotions and cognitions (Hypothesis 1). Because of the limited evidence to date, we are not in the position to formulate more fine-grained hypotheses regarding the importance of sub-mechanisms of moral disengagement (e.g., blaming and dehumanizing the victim) instead we analyze the importance of four sub-levels (e.g., harmful practice, harmful effect, result, and victim) in an exploratory way.

Regarding the moderating role of religiosity, it is difficult to formulate clear hypotheses, because the evidence to date is conflicting. Given the location of our sample (see Participants), it is possible that it will be easier for more religious teachers to hold high levels of homophobic prejudice because the Roman Catholic Church has a clear stand against same-sex romantic relationships. If this is true, religious teachers would need lower levels of moral disengagement to cope with their moral emotions and cognitions compared to less religious teachers. However, this is not necessarily the case for individuals who understand their religiosity in a less conservative way, as such people might be less inclined to follow the dogmatic view of their church. A similar argument could be put forward regarding ethnic or religious minorities. Only if religious people perceive them as believers to the wrong God, they might need less moral disengagement to cope with their moral emotions and cognitions. Thus, based on these theoretical elaborations we expect that the strength of associations between moral disengagement and the three types of prejudice might differ between teachers with low versus high levels of religiosity (Hypothesis 2).

Method

Participants and Procedure

The sample is composed of n = 550 teachers (76.4% women) aged 20–70 years old (Mage = 46.18; SD = 9.80) from Italian primary and secondary schools. Their teaching experience ranged between 1 and 44 years (M = 17.22; SD = 9.90). It can be assumed that the vast majority of the participants are members of the Roman Catholic Church, because 88% of the Italian population are Roman Catholics (IPSOS, 2017). Participants were recruited online via various social media platforms (e.g., emails, facebook groups, blogs, etc.). and they were asked to distribute the link of the online survey to colleagues who might also be interested in participating. Participation was voluntary and confidential, and active informed consent was obtained prior to participation. The data were collected between November 2020 and January 2021. All procedures performed in studies involving human participants were compliant with the ethical standards approved by the Helsinki Declaration of 1964 (and subsequent amendments) and was approved by the ethics’ committee of the Social and Forensic Psychology Academy of Rome.

Measures

Demographic Information

Teachers were asked to provide information related to their gender, age, current relationship status, country of birth, city of residence, and professional experience (e.g., years of teaching).

Religiosity

Three items were constructed for this study: “How important is religion to you?” “How important is it to follow the precepts of your religion?” and “How important is it to follow religious rites for you?” Participants were able to answer these items using a Likert scale ranging from 1 (in no way) to 6 (very important). The Cronbach’s alpha suggests adequate reliability (α = 0.77).

Homophobic Prejudice

The Teacher Attitudes Towards the Representation of Homosexuality in Film and Television scale (D’Urso & Symonds, 2021) is a self-report questionnaire comprising 15 items composing one factor. The items cover a range of emotional and social aspects of teachers’ homophobic attitudes regarding film and television (e.g., feel anger when I see a film or advertisement featuring homosexuals, same-sex families and/or references to them; homosexual issues should never be referred to, as they are morally wrong). Each item was measured using a 5-point Likert type scale (1 = strongly disagree, 5 = strongly agree). The Cronbach’s alpha suggests good reliability (α = 0.88).

Ethnic Prejudice

The subtle and blatant prejudice scale (Arcuri & Boca, 1996) is a self-report questionnaire comprising 20 items that can be answered using a Likert scale ranging from 1 (absolutely disagree) to 6 (absolutely agree). The blatant prejudice subscale included items related to threat, rejection, and anti-intimacy with ethnic minorities (e.g., it would be annoying if a loved one or a relative would get married to a non-EU person). The subtle prejudice subscale measured attitudes related to defense of traditional values, exaggeration of cultural differences, and denial of positive emotions towards ethnic minorities (e.g., foreign people who live in our country pass on to their children values and skills that are not those necessary to be successful in Italy). The Cronbach’s alpha suggests good reliability in both subscales (blatant prejudice α = 0.90); subtle prejudice α = 0.87).

Moral Disengagement

The Italian version of moral disengagement scale (Caprara et al., 1996) is composed by 32 items evaluated on a 5-point Likert-type scale (1 = completely disagree, 5 = completely agree). The scale measures eight moral disengagement mechanisms, with four items in each subscale: moral justification, attribution of blame to victim, euphemistic labelling, advantageous comparison, displacement of responsibility, diffusion of responsibility, distortion of consequences, and dehumanization of victim. These eight mechanisms were collapsed into four levels (Bandura, 2006): (1) harmful practices (comprised of the three mechanisms of moral justification, advantageous comparison, and euphemistic labelling); (2) harmful effects (comprised of displacement of responsibility and diffusion of responsibility); (3) result (comprised of one mechanism, namely distortion of consequences); (4) victim (comprised of two mechanisms of dehumanization and attribution of blame to the victim). The Cronbach α of these four sub-scales ranged between 0.80 and 0.90.

Analysis Plan

The statistical analyses were carried out in two steps. In the first step, we established the measurement model for moral disengagement. We used the four sub-scales of moral disengagement (harmful practice, harmful effect, result and victim) as parcelled indicators to reduce model complexity. Parcels are preferred for the current analyses because, compared with individual items, parcels have superior psychometric quality that reduce both Type I and Type II sources of error but do not bias or otherwise inflate construct relations (for details see Little et al., 2003). To test Hypothesis 1, a structural equation model (SEM) using Mplus was conducted. The latent factor moral disengagement was regressed on blatant and subtle ethnic prejudice and homophobic prejudice controlling for age, gender, religiosity, and teaching experience (see Fig. 1). To test Hypothesis 2, we computed a multiple group structural equation model to test whether the relationship between the latent factor of moral disengagement on the three outcome variables (blatant and subtle prejudice, and homophobic prejudice) differed between teachers with low and high levels of religiosity.

Fig. 1
figure 1

Structural equation model (SEM) for the effects of moral disengagement as latent variable on blatant and subtle ethnic prejudice and homophobic prejudice. The model is controlled by age, gender, religiosity, and teaching experience. Note: pathways with STDYX standardized estimates are displayed. *p < 0.05, **p < 0.01, ***p < 0.001

Full size image

Goodness of fit was assessed with indices that are less sensitive to sample size (Kline, 2005). The comparative fit index (CFI), the Tucker-Lewis index (TLI), standardized root mean square residual (SRMR), and the root mean square error of approximation (RMSEA). CFI and TLI range from 0 to 1; values greater than 0.90 and 0.95 are indicative of acceptable and good model fit. The RMSEA ranges from 0 to 1 (< 0.05 indicates good fit; < 0.08 indicates acceptable fit, with associated 95% confidence intervals (CIs), SRMR < 0.08 gives an indication of adequate fit. Missing data was handled using Full information maximum-likelihood (FIML), as it offers less biased estimates even when the pattern of missingness cannot be ignored (Baraldi & Enders, 2010).

Results

Correlations and descriptive statistics among the main variables are shown in Tables 1 and 2, respectively. Moderate to strong correlations were observed among the four moral disengagement subscales, and moderate associations were found among the three forms of prejudice. Moral disengagement indices were significantly correlated with subtle ethnic prejudice, but not blatant ethnic prejudice. Dehumanizing the victim was the only moral disengagement mechanism that positively correlated with homophobic prejudice. Religiosity was also positively related to homophobic prejudice, but not ethnic prejudice, whereas teaching experience was positively associated with religiosity and all forms of prejudice. Significant gender differences were found in religiosity, where women scored higher compared to men. No gender differences were observed in homophobic prejudice, blatant ethnic prejudice, or subtle ethnic prejudice.

Table 1 Correlations among variables
Full size table
Table 2 Means and standard deviations for the main variables by gender
Full size table

Measurement Model

The latent variable moral disengagement was built with the for parceled indicators (harmful practice, harmful effect, result and victim). The model fit was good, χ2 (2) = 20.46, p = 0.000, CFI = 0.984, TLI = 0.951, RMSEA = 0.13 (90% confidence interval = 0.083–0.184). Harmful effect loaded with 0.887, harmful practices loaded with 0.831, result loaded with 0.61 and victim loaded with 0.84 on the latent factor moral disengagement (standardized estimates, STDXY standardization).

Associations Between Moral Disengagement and Prejudice (Hypothesis 1)

The structural equation model (SEM) is presented in Fig. 1, STDXY standardized estimates are shown, and only significant paths are displayed. The standardized associations between moral disengagement and blatant ethnic prejudice, subtle ethnic prejudice, and homophobic prejudice partly confirmed Hypothesis 1: moral disengagement as latent factor was positively associated with subtle ethnic prejudice (β = 0.18, SE = 0.04, p < 0.001), while homophobic prejudice was predicted by religiosity (β = 0.14, SE = 0.04, p < 0.001), and age (β = 0.36, SE = 0.09, p < 0.001). No significant associations were found between moral disengagement and blatant ethnic prejudice, and homophobic prejudice.

We also explored whether the four sub-scales of moral disengagement (harmful practices, harmful effect, result, victim) were individually associated with blatant ethnic prejudice, subtle ethnic prejudice, and homophobic prejudice (see Table 3). The blatant indicators of the four sub-scales in addition to religiosity, age, gender, and teaching experience, were regressed on blatant ethnic prejudice, subtle ethnic prejudice, and homophobic prejudice, χ2(18) = 56.88, p < 0.001, CFI = 0.983, TLI = 0.947, RMSEA = 0.062 (90% confidence interval = 0.043–0.081). Results (see Table 3) showed that harmful practice and victims were the two subscale that predicted homophobic prejudice (β = − 0.22, SE = 0.12, p = 0.003 and β = 0.15, SE = 0.13, p = 0.048, respectively), along with religion (β = − 0.22, SE = 0.12, p = 0.003), and age (β = − 0.22, SE = 0.12, p = 0.003).

Table 3 Latent variables of moral disengagement on blatant, subtle ethnic, and homophobic prejudice
Full size table

Multigroup Comparison: Religiosity as a Moderator Between Moral Disengagement and Prejudice (Hypothesis 2)

To test whether the level of moral disengagement is moderated by the level of religiosity for blatant ethnic prejudice, subtle ethnic prejudice and homophobic prejudice, a multiple group comparison was carried out. A median split was conducted to be able to compare teachers with high and low levels of religiosity. Teachers with means for religiosity lower or equal than 3.00 formed the low-level group, while teachers with means above 3.01 formed the high-level group.

The model fit of the multiple group SEM was good, χ2(52) = 118.23, p < 0.001, CFI = 0.979, TLI = 0.952, RMSEA = 0.069 (90% confidence interval = 0.052–0.085). As shown in Fig. 2, after controlling for age, gender, and teaching experience, we found that moral disengagement positively predicted subtle ethnic prejudice both for teachers with low (β = 0.20, SE = 0.06, p = 0.001) and high levels of religiosity (β = 0.16, SE = 0.06, p = 0.016). Age positively predicted homophobic prejudice among teachers with low levels of religiosity (β = 0.47, SE = 0.13, p < 0.001), and high levels of religiosity (β = 0.27, SE = 0.11, p = 0.016). Gender (being male) and teaching experience was not significantly associated, in neither of the two groups.

Fig. 2
figure 2

Multiple group comparisons for religiosity separately for the low and high religiosity. For factor loadings and intercepts (freely estimated but constraint to be equal between groups). The model is controlled by age, gender, religiosity, and teaching experience. Note: pathways with STDYX standardized estimates are displayed. *p < 0.05, **p < 0.01, ***p < 0.001

Full size image

To find out which of the simple slopes differ between groups (low vs. high religiosity), 12 difference scores (6 slopes × 2 groups) were calculated. The inspection of the simple slopes revealed no significant differences indicating that the strength of the predictors did not differ depending on low versus high religiosity.

Discussion

The aim of this study was to analyse the relationship between moral disengagement and ethnic and homophobic prejudice, considering also the role of religiosity in a group of teachers. Studying this topic among teachers is highly relevant, because their subtle and blatant prejudice towards minority students severely impact the quality of their teaching (D’Urso & Petruccelli, 2022; Kennedy, 2010; Vervaet, et al., 2016). Subtle and blatant prejudice of authority figures like teachers also convey the message that students belonging to ethnic and sexual minorities do not have the same right to be treated with dignity and respect. Because schools have a particularly high societal responsibility to put into practice the basic values of democratic societies that are based on human dignity and human rights, it is important to understand the factors that contribute to high levels of ethnic and homophobic attitudes among teachers. Such knowledge helps with defining strategies to improve inclusive didactic actions.

Moral disengagement is a represented by a collection of self-justification processes that allow humans to hold negative attitudes towards immigrants and sexual minorities even if they are in contrast with their own (or societies) moral standards (Maftei & Holman, 2022), In the current study, religiosity was hypothesised to add to the complexity of these processes. The results of our analyses suggest that moral disengagement indeed increases subtle ethnic prejudice among teachers. However, in contrast to our hypotheses no associations were found for blatant ethnic prejudice and homophobic prejudice. Higher levels of homophobic prejudice were instead found for older and more religious teachers. Importantly, and in contrast with our predictions, high versus low level of religiosity did not moderate the strengths of associations between moral disengagement and the three forms of prejudice and negative attitudes.

The finding that moral disengagement was not associated with blatant ethnic prejudice but was associated with subtle ethnic prejudice might be explained by the fact that people who are openly racist and have little compunction against displaying their antipathies and hatred against immigrants might not feel a sense of moral conflict, as they do not perceive their prejudice and negative attitudes as morally wrong. In contrast people who hide their prejudice against ethnic minorities and therefore display them in a latent way might be aware that their attitudes are clearly in contrast to basic democratic values, human dignity and human rights. Consistent with the theory of moral agency (Bandura, 1999), the results of the current study suggest that individuals are motivated to employ moral disengagement strategies to resolve their discomfort when they are aware that society’s moral values conflict with their unjustified, personally held prejudicial attitudes.

This study also demonstrated that teachers who are older and those who are more religious, have higher levels of homophobic attitudes. It does mean younger teachers represent a different generation that is more open and accepting of diversity (from exclusion to social inclusion). Furthermore, in line with the literature (e.g., Harvey, 2016; Preston et al., 2010), although among the precepts of religion there is acceptance and considering the other as a brother or sister, sometimes this may not be corroborated if teachers is faced with sexual minority groups. In this sense, religiosity may prevent the creation of inclusive attitudes and emotions towards the LGBT community, probably because the precepts of the Catholic Church has not consistently promoted an unconditional accepting view of homosexuality, same sex relationships, and same sex parenting (Adamczyk & Pitt, 2009; Jäckle & Wenzelburger, 2015). Religiosity, which leads to the internalization of ideas, precepts, and ways of seeing the world over time, may help maintain a state of internal coherence in values and attitudes that would increase this form of prejudice and avoid cognitive dissonance. This could compromise the didactic actions of teachers who could get stuck in a vicious circle that leads them not to act when they are faced with an episode of bullying towards minority groups, for instance (Strohmeier & Gradinger, 2021).

Study Limitations and Future Research

Although the study extends the literature, it should be considered in light of its limitations. First, the use of self- report questionnaires may increase the social desirability of teachers who may attempt to show themselves in a more positive and socially acceptable way. To address this issue, future studies could make use of structured interviews and other implicit measures such as the Implicit Association Test (IAT) and the Implicit Relational Assessment Procedure (IRAP) to reduce the social desirability bias. Building on the present findings, additional research is also needed that includes other potentially important individual variables (e.g., personality characteristics, political orientation) that may be configured as predisposing and/or protective factors for the aetiology of prejudice and negative attitudes, considering also the particularities of different national contexts.

Conclusion

The study highlights that there is a the need to implement human rights education for teachers to make sure that the basic values of democracy are adequately transmitted to students in schools. Teachers are a highly important group that should undergo such an education, because they will only be able to think about diversity as something that enriches democratic societies if their mindsets (such as social, cognitive, and value processes) are open to accepting it as an unconditional consequence of human rights. In order to structure specific interventions, teachers should be encouraged to reflect on their own strengths and potential biases, as well as their cognitive and value resources, and should be empowered to make the changes that will benefit their students (and themselves). Teachers should benefit from proper training to work on awareness of their limitations and resources to counter prejudice. Furthermore, the debate between colleagues on moral and ethical issues relating to minority groups, in a continuous training course, may encourage inclusive attitudes. It also demonstrates that working on teachers’ awareness of their own biases and self-justifying strategies, and implementing human rights education for teachers are necessary actions towards building more inclusive schools.