XENOPHObIA AND INTErETHNIC rElATIONSHIPS AMONg SECONDAry lEVEl PUPIlS IN SPAIN

Migratory processes in southern Europe over the last two decades have brought about substantial changes to the ethnic makeup of secondary schools. Classrooms have increased in their complexity in terms of teaching, as there are pupils with different cultural and economic backgrounds and educational needs, but also in the relationships among the peer groups of pupils. what kinds of attitudes and relations can be found among the members of this culturally diverse pupil body? what kinds of views are held by the pupils about each other? The aim of this paper is to shed light on stereotypes and xenophobic attitudes towards immigrant pupils in secondary schools in Spain, principally among peer groups, and on how teachers address this multicultural classroom dynamic and complexity and any possible prejudices towards stereotypes and racist attitudes that surface inside the classroom. relationships tend to be formed among peer groups of the same ethnicity and there is a marked rejection of pupils from Moroccan origin. Teachers are neither aware of pupils’ stereotypes nor of their own. As a consequence, they are not giving sufficient pedagogical responses to resolve arguments or disrespectful situations against immigrant pupils. This in turn is contributing, on the one hand, to racist and xenophobic societies. On the other hand, teachers’ low expectations of these immigrant pupils has had an effect on their academic achievements and on dropout rates. The present study is based on 2196 questionnaires administered to pupils at 43 Spanish secondary schools. In-depth interviews with 54 secondary school teachers were also carried out, so as to understand teachers’


AbSTrACT
Migratory processes in southern Europe over the last two decades have brought about substantial changes to the ethnic makeup of secondary schools.Classrooms have increased in their complexity in terms of teaching, as there are pupils with different cultural and economic backgrounds and educational needs, but also in the relationships among the peer groups of pupils.what kinds of attitudes and relations can be found among the members of this culturally diverse pupil body?what kinds of views are held by the pupils about each other?The aim of this paper is to shed light on stereotypes and xenophobic attitudes towards immigrant pupils in secondary schools in Spain, principally among peer groups, and on how teachers address this multicultural classroom dynamic and complexity and any possible prejudices towards stereotypes and racist attitudes that surface inside the classroom.relationships tend to be formed among peer groups of the same ethnicity and there is a marked rejection of pupils from Moroccan origin.Teachers are neither aware of pupils' stereotypes nor of their own.As a consequence, they are not giving sufficient pedagogical responses to resolve arguments or disrespectful situations against immigrant pupils.This in turn is contributing, on the one hand, to racist and xenophobic societies.On the other hand, teachers' low expectations of these immigrant pupils has had an effect on their academic achievements and on dropout rates.The present study is based on 2196 questionnaires administered to pupils at 43 Spanish secondary schools.In-depth interviews with 54 secondary school teachers were also carried out, so as to understand teachers'

INTrODUCTION
Despite the fact that there are numerous relevant studies on how ethnicity and specifically economic determinants influence the teaching-learning process (Alba, Sloan and Sperling 2011;Gillborn and Mirza 2000;Kingdon and Cassen 2010;Portes et al. 2011;Stevens, Clycq, Timmerman and Van Houtte 2011;Tyson and Darity 2005), it is also important to analyse how xenophobic and rejectionist attitudes between peer-group members of different ethnic origin generate situations that can hinder the teaching-learning process.Racist attitudes start and can be learnt in the common spaces and places where socialisation first occurs; and this usually has knock on effects.The rejection by one peer group of another has negative consequences for pupils' self-esteem and can engender a negative climate in the classroom (García López 2006;Hierrezuelo et al. 2002;Hombrados-Mendieta and Castro-Travé 2013;Thijs and Verkuyten 2011).Given this, the aims of this paper are to focus on the ways in which groups with different ethnic backgrounds interact with each other in schools, what the consequences of these negative attitudes are on pupils, and what teachers' pedagogical responses, if any, are in this regard.Moreover, we are seeking to find possible prejudices harboured by teachers towards immigrant pupils from different cultural diversities.
Firstly, this paper analyses the degree to which there are stereotypes and even xenophobia among secondary school pupils, specifically among 15-and 16-year-olds in their final year of compulsory education within the Spanish system, as such an analysis could lead to an educational intervention strategy combatting the future social development of racist attitudes.In cases where there were significant levels of xenophobia and/or racism, the conditions that foster such attitudes have also been explored, as well as the extent to which the distribution of pupils from different origins tends to engender levels of xenophobia and racism (Waghid and Davids 2013).Secondly, this paper attempts to identify which ethnic groups are more susceptible to stigmatisation, and by which groups.Thirdly, it explores whether previous factors, such as a lack of social relations, social stereotypes, prejudices and the stigmatisation of certain ethnic groups influence xenophobic attitudes among pupils.Finally, it examines how teachers' perceptions are influenced by their own negative stereotypes regarding some minority groups, what their preconceptions are, and how they act in front of immigrant pupils when faced with such situations (Dovemark 2013;O'Connor 2014).

SCHOOl XENOPHObIA AND STErEOTyPES IN SPANISH SECONDAry lEVEl CONTEXT
The concepts of racism and race originate in a categorisation according to biological differences, and are characterised by power as an instrument of control based on such differences and on a perception of the inferiority of an ethnic group (Gillborn 2008;Guillaumin 2002).Today, racism is based not only on skin colour, but on cultural and economic differences, which form the basis for rejection and stigmatisation.Although skin colour serves to identify a certain group, attitudes based on it tend to become diluted when the economic component is lacking (Pérez -Yruela and Desrues 2006;2007).Thus, our analysis will focus on the perception of ethnicity, on the bases of cultural stereotypes, rather than on the influence of race.The term "ethnicity" is thus preferred to "race" (Hill et al. 2007).
There has been an increase in racist attitudes against minority groups in Spain.According to the Ministry of Interior Affairs, 25 per cent of the victims are minors (ED 2015;SOS Racismo 2015).The presence of Muslim, or Arab, communities in Europe is not welcome.There is collective imagery mistrust towards the Muslim community, especially after 9/11, and the Madrid attack, and because of recent adolescent involvement as jihadists carrying out attacks on behalf of the Islamic State (ISIL) (Banks 2011;Deusdad 2013).In the case of Spain, there is even an historical rejection of Muslim communities based on the conquest of the "infidels" from the Medieval Ages up to the Spanish Civil War (Preston 2006;Deusdad 2009;2010;2013).
A sentiment of anti-immigration implies an antagonism towards public programmes or policies that promote equality, coupled with a denial of the existence of discrimination (Evans-Winters and Twyman Hoff 2011;Flecha 1999;Hogan and Mallott 2005;Pérez-Yruela and Desrues 2006;Wieviorka 1992).As regards the concept of stereotype, understood as prejudice against others, it is an appraisal applied in indiscriminate fashion, resulting from an a priori categorisation without rational or scientific foundation or a basis in personal experience through contact with members of the group in question.It consists of a homogenous evaluation of an ethnic group, gender or other sorts of identities, made quickly in an immediate cognitive response (Campbell 2015;Hilton and von Hipple 1996).In turn, stereotypes and social labelling represent a restriction on individuals, who become pigeonholed and thereby less free (Berger and Luckmann 1966).
Despite the fact that our analysis focuses on the relation between autochthonous pupils and those of different ethnicities, it should also be noted that teachers' perceptions of the abilities and potential of immigrant pupils are in the majority of cases negative and stigmatising (Deusdad 2009).This activation of teachers' negative stereotypes has a negative Pygmalion effect and can lead to a self-fulfilling prophecy.In other words, it can have negative effects on pupils' attainments (Burgess and Greaves 2009;Campbell 2015;Earp 2010).
The distribution of foreign pupils in classrooms has clear repercussions for the relations established by members of the peer group with those of their own ethnicity and with those of other origins (Sánchez-Hugalde 2009).Pupils' rejection of peers from a different ethnic group is greater the more their own ethnicity dominates in the classroom, though this occurs to a lesser extent in the case of native pupils.At the same time, it is among Afro-Americans that there is the most rejection of one's own group (Bellmore, Nishina and Witkow 2007).This would appear to be due to the internalisation of the stigma suffered by this group and their scarce "social success" as a community (Goffman 1990).
The context of immigration in Spain is characterised, as in other southern European countries, by relatively recent immigration in two waves mostly since 1996; after that, immigration increased exponentially until the global economic crisis in 2008.The peculiarity of the Spanish case lies in the rapidity of the process: in just 5 years -from 2000 to 2005 -the number of foreign inhabitants increased by 4 million.This had significant repercussions in schools, where there was a massive influx of non-Spanish nationals, but without the necessary planning or educational resources (Miralles, Prats and Tatjer 2012;Prats, Barca and López-Facal 2014).Studies on the Spanish case have highlighted the fact that negative perceptions of ethnic-minority pupils increased from 10 per cent in 2000 to 31 per cent in 2005 (Oller, Vila and Zufiaurre 2012;Zufiaurre and Peñalva 2007) In the autonomous regions (comunidades autónomas) with the highest percentages of immigrant pupils, such as Cataluña, Murcia or Madrid, it was decided to set up "Sheltered Catalan or Spanish Immersion Classrooms" (aulas de acogida).These constituted a temporary measure whereby all those pupils who did not speak the official language or languages of the region in question were grouped together for a period of one to two years maximum so as to provide them with "total immersion" language tuition and keep all the new arrivals together (Alegre and Subirats 2007;Besalú 2002;Deusdad 2009).
The establishment of these classrooms began in 2004, and represented a significant step in the right direction.They were provided with the necessary teaching staff and resources, the teachers involved were highly motivated, and intercultural methodology was employed.These classrooms had the support of a specialist in interculturality (Molina and Casado 2014).Nevertheless, the transition to ordinary classes was not so carefully implemented, and the new pupils suffered from a certain degree of "invisibility" (Deusdad 2009;2010;2013).
As far as the distribution of pupils from immigrant families in schools is concerned, it was quite unequal in Spain.Particularly in Catalonia, the majority of them (84%) were in public schools, with only 16 per cent in grant-assisted or private schools, according to official data for the school year 2011-12 (Consell Superior d'Avaluació del Sistema Educatiu 2013).This has brought about problems of segregation and a concentration of pupils from the same ethnic group in certain primary and secondary schools.
The unequal distribution of pupils across classrooms, together with segregation in schools and neighbourhoods, made it difficult for the new arrivals to learn the languages, and above all to acquire the requisite academic skills and knowledge (Sánchez-Hugalde 2009).On the other hand, recent studies at a European level have stressed the importance, for academic success, of factors such as management and leadership style and provision of resources in schools with high levels of ethnic diversity (Gillborn 1995;1997;2008;Stevens et al. 2011).

STUDy DESIgN AND METHODOlOgy
The general goal of the research was to explore the educational and social reality of pupils from an immigrant background in compulsory secondary education (Educación Secundaria Obligatoria [ESO]) in Spain, including first, second and "1.5" generation children.The aim of this paper is to shed light, on the one hand, on how integration occurs in schools and classrooms with a mix of ethnic origins, how interethnic relationships developed and whether they were marked by prejudices, the role played by stereotypes, and whether or not there were xenophobic attitudes.On the other hand, another part of the research set out to analyse the acquisition of social sciences knowledge and the differences found according to ethnic origin.
This section of the paper focuses on the part of the research in which we analyse the factors that influence an increase in racist attitudes within the peer group.The objective is to explore whether xenophobic attitudes are generated, and how social factors and aspects related to the makeup of schools and classrooms affect the increase in such attitudes.Among our important objectives is that of explaining the correlation between racism and ethnic groups and the incidence of stereotypes.Moreover, teachers' stereotypes concerning immigrant pupils and their discourses on immigrant peer-group relationships could have effects on pupils' attitudes and influence negatively their attainment, which could provoke a self-fulfilling prophecy (Campbell 2015).Teachers' low expectations of these ethnic groups have negative consequences for their selfesteem and achievement (Apple and Beane 1995;Deusdad 2009).
The methodology employed in the present study was primarily quantitative, though this was complemented by some qualitative research involving structured observations in classrooms and at break times (22 observations), together with exploratory and semistructured interviews with teachers and other education professionals (54 interviews).Around 2000 questionnaires (N=1985) were administered at 43 Spanish secondary schools to fourth-grade pupils (4º de la ESO; approximately age 15).
In all, the interviewees were asked about their teaching, their relationships with foreign pupils, the special needs, strengths and positive aspects of their pupils, and other questions related to the theoretical framework described above.As well as interviews, extensive classroom observations were carried out, during which data was checked as to what teachers had said in their interviews regarding pupils and class atmosphere, while pupils' behaviour were also observed in multicultural classrooms.The responses were also underlined and contrasted in a hermeneutic analysis of the interviews and classroom observations.Those aspects which were different in the two cases were highlighted and looked at deeply from an emic and etic perspective.
The sample included those autonomous regions with the highest percentages of immigrants (Catalonia 12%, Murcia 13%, Madrid 12% and Aragón 8%) for the period 2010-2012.In these regions schools were selected according to the following diversity criteria: type (grant-assisted and public), diverse percentages of immigrant pupils, a range of socio-economic levels of pupils' families, and location (urban, suburban and rural).The schools chosen had between 25 per cent and 75 per cent of immigrant pupils, the majority of whom were Latin American or Moroccan; there were also smaller numbers of pupils with a Chinese background.The breakdown of ethnic groups can be explained by the greater presence of Latin American communities in Spain as a whole and of Moroccans in Catalonia, and the isolation and difficulties of communication presented by pupils of Chinese origin when it came to the qualitative part of the study.
In order to analyse whether there is a relationship of dependence between the degree of xenophobia and the different variables (sex, percentage of immigrants at the school, and social class), the categorical variables to binary variables were converted.The general xenophobia variable (xenophobia1) was obtained through the questionnaire item: "The presence of pupils from other countries and relations with them:…", which has the following categories: "contribute nothing", "are enriching", and "are harmful or detrimental to us", and was recoded assigning the value "no" to the first category ("contribute nothing") and "yes" to the second and third categories.
Likewise, the xenophobia against nationalities variable (xenophobia2) was obtained based on the choice of one of the four categories for each of the nationalities: "I like them as friends that I could spend a lot of time with", "They are people I don't mind being with", "I would prefer not to spend too much time with them", and "I would prefer not to have anything to do with them in my whole life".In the first two cases the option that indicated xenophobia against the nationality in question was "no", in the third and fourth cases it was "yes".

PrEVAlENCE OF STErEOTyPES
In order to measure pupils' stereotypes according to nationality, we coded the questionnaire items that used positive and negative attributes to refer to pupils of foreign origin, such as: serious, suspicious or mistrustful, stupid, lazy, violent, and cocky, among others.These were classified in two categories: positive and negative, which we used as the basis for measuring whether or not there existed stereotypes in relation to a particular cultural group.
As can be seen in Table 1, the nationalities to which the most stereotyped views apply are Moroccans and Peruvians.There is marked rejection of Moroccans from all the groups analysed, except of course from Moroccans themselves.We also found a stereotyped view of Germans on the part of most groups.Also notable are the stereotyped views (93.8%) about Peruvians (and most other Latin American groups) held by Moroccans, and the highly stereotyped view of Moroccans held by Dominicans (95%).Note: The stereotype variable is obtained through a questionnaire item in which respondents assigned two adjectives to each nationality.These variables were coded in binary fashion according to whether they were positive adjectives (serious, hard-working, friendly, trustworthy, kind, cheerful, fun, etc.) or negative adjectives (suspicious/mistrustful, stupid, strange, lazy, violent, uneducated, mean with money, cocky, racist, etc.).The analysis corresponds to the percentage of negative adjectives assigned by each nationality in relation to the total number of adjectives assigned by each nationality.

DEgrEE OF XENOPHObIA AND STIgMATISATION OF grOUPS
Another aspect analysed in this study was the degree of xenophobia, measured through the appraisal of cultural enrichment contributed by the relationship with "others".To this end we designed a questionnaire item inquiring whether the experience of interacting with people from other cultures was enriching or detrimental, analysed in accordance with three variables: sex, educational level of parents or guardians and socio-economic status.
The results for general xenophobia (xenophobia1) reveal that 30 per cent of pupils thought the presence of pupils from other countries was harmful or detrimental, whilst 70 per cent considered that this was not the case.As regards differences by sex, there was less xenophobia among girls than among boys, with a difference of almost four percentage points.With regard to parents'/guardians' educational level, we found that as their educational level increases, xenophobia tends to decrease slightly; however, such a difference is not appreciable in the case of socio-economic status.Even so, despite the difference being subtle, the group considered "low class" does show slightly over two percentage points more xenophobia than the group considered "higher-middle and high class".This is undoubtedly due to the fact that it is in this "low class" group that most of the pupils with immigrant backgrounds are situated.Table 3 shows the differences in the degree of xenophobia by nationalities, and includes the native (ethnic Spanish) pupils together with those of non-EU immigrant background (Latin American and Moroccan).From the data obtained it is clearly observed that Spanish pupils are those that present the most xenophobic attitudes (38.2%), whilst in the rest of the groups it is considerably lower, descending to levels of just 3.4 per cent in the case of Colombians and 5.1 per cent in that of Dominicans.The level of xenophobia presented by Moroccans and Bolivians is around 23.1 per cent in either case.The Chinese pupil community is also strongly rejected (though less so than the Moroccans), and this contrasts with the lack of negative stereotypes of the Chinese.This is undoubtedly due to the fact that there is a positive view of the Chinese community in relation to values such as work, seriousness and placidity, but a rejection of them as members of the community and as friends within the peer group.In general, pupils with immigrant backgrounds present fewer stereotypes and less xenophobia than the native, ethnic Spanish pupils.As regards Moroccan pupils -the group most rejected by all the other nationalities -paradoxically, they are not at all reluctant to interact with Spanish and with European pupils in general, as well as (to a lesser extent) with those from a Latin American background.
Another of the correlations made with the degree of xenophobia concerns the percentage of foreign pupils in the schools.We set out to discover whether segregation at the secondary-school level led to xenophobic attitudes and what repercussions this had for the relationships within groups comprising pupils with different ethnic backgrounds.Table 5 shows how schools with greater cultural diversity and larger numbers of foreign pupils present lower rates of xenophobia among the pupil body (25.9%) than schools with zero per cent to 25 per cent of pupils from immigrant backgrounds (39%).

TEACHErS' DISCOUrSES AND EXPECTATIONS
One of the main problems pointed out by teachers and educators is pupils' lack of schooling in their countries of origin, in general as well as regards language skills and school habits, as the following excerpt from an interview shows: The basic education they have received is very poor in general terms.They do not understand the vocabulary we use in the social sciences; for instance, pupils in the first year do not distinguish between north and south in a map.We have a map in class and we play games so that they can learn such basic orientation.We do things as easy as that.Their basic knowledge is very bad, very bad.We ask them how many years they went to school in their countries.It is unlikely that we will be given all the documents.Yesterday, one pupil told me: "I rarely went to school" (iba muy poco).So we know that this pupil will know nothing.(Interviewee 24) This view is partly based on teachers' perceptions and preconceptions and partly the consequence of the pupils' negative academic results and the high dropout rates among immigrant pupils, in particular among Moroccan (94%) and Dominican pupils (91%) (Deusdad 2009, 34;De Witte and Rogge 2013;Portes et al. 2011).Teachers stereotype and classify immigrant pupils as unable to continue studying because of family circumstances and the immigration process in general.Their academic expectations of these immigrant pupils are negative and low.Some arguments demonstrating this view can be seen in the following excerpt from an interview: A Moroccan pupil is doing precollege studies.I am her class teacher (tutor) and she is thinking of dropping out because her family want her to.For them, studying is not useful.In their world it is worthless.What did her mother and grandmother do?Like them, she will get married and will be kept by her husband.Why should she be studying if it is not worth it?Obviously, she doesn't want to waste her time and finish college.She knows she is in Spain and she tries to integrate and take advantage of the possibilities open to her. (Interviewee 18) This situation is not only the result of the lack of schooling: a major recognition of alterity is required, as well as higher academic expectations.One of the problems that faces these pupils is that they need all teachers to understand and appreciate their cultural diversity while they are in the process of learning.This can also been seen in the way school activities were organised.For instance, a meeting with parents at the beginning of the year was placed during Ramadan celebrations, as is explained in the following extract from an interview: Well, we could not hold the meeting on another day, because it would have been too late.What we decided was to start the meeting at 6 pm instead of 7 pm so that they would have time to prepare the food and eat it after sunset.We also asked the Arabic mediator to help us.Even though we made the announcement in Arabic and Hassan also came, too, and despite calling all the families by telephone and announcing the meeting to our pupils, only one family attended the meeting.We invested a considerable amount of time and effort on it, we had many compliments, but we were not successful.I think it was because of Ramadan.(Interviewee 15) As the interview shows, the school staff felt disappointed because their attempts to involve immigrant families had been a waste of time.If the school had scheduled a meeting at Easter or Christmas it's sure no one would have attended either.In other words, what this example shows is that it is important to recognise the cultural diversity of the community and even more important to understand their culture and traditions if you want to interact and transform their reality.The host institutions and native population believe that Muslim communities do not want to integrate with the host society, when what they really want is to preserve their traditions and reaffirm family bonds.
There was a constant thread running through all the interviews.Teachers often consider attitudes of rejection, disrespect and contempt towards immigrant pupils as typical of the pupils' age and do not recognise racism in such attitudes.For them, teenagers generally hate those who are different: children who are fat or who wear glasses, for instance.This can be seen in the next interview extract: No, there is no xenophobic conflict.When someone is fighting with a "Paki" he will call him a "shitty Paki" (paki de mierda) or when he is fighting with a Moroccan he will call him a "shitty Arab" (moro de mierda).But I have heard one Moroccan call another a "shitty Arab" and two or three, who have been in the country longer than others, say "Bloody hell!These shitty Arabs" (Coño!Estos moros de mierda) to a chap whose skin was even darker than your skirt.(Interviewee 1) In contrast to this tendency, although they are a minority, there are also teachers aware of these situations of disrespect and xenophobia among pupils, as can be seen in the following interview excerpt: Yes, the conflict is serious, although for a time I think there was no conflict.However, you have to help and mediate.If someone trusts you and says, "They call me, 'moro'" I answer: "Who calls you that?"And then you start talking with them and with others.This kind of conflict exists.(Interviewee 41)

DISCUSSION
"Social tolerance" -that is, the acknowledgement and acceptance of difference -has greater presence among individuals with higher levels of education and greater means to acquire economic power.Likewise, we find differences in terms of sex, girls being less xenophobic in general, and the greater the cultural diversity, the lower the levels of xenophobia.In other words, ethnic and gender diversity reduce the amount of xenophobia (Hadler 2012;Hjerm 2005).In turn, it has been observed that the effect of education is increased when it is complemented by help from international organisations whose work reinforces and aids understanding of teachers' discourse (Hadler 2012).Countries with high levels of xenophobia would include India, France and Turkey, among many others.
The translation of these situations into the classroom context brings with it rejection, disdain and indifference.This has even been observed in higher education and brings to light the necessity of a specific pedagogical treatment in terms of intercultural relationships among peers inside their classrooms (Mabin 2007;Moorea and Lemmerb 2010;Ordóñez and Ordóñez 2004).This fact links us to the importance of further research in this regard.Moreover, there is still a need to evaluate educational centres to prove their pedagogical attitudes against racism and xenophobia.
These xenophobic relationships are more emphatic in the case of the Muslim community, and are found in different European countries where there is a presence of such communities (Deusdad 2009;2013;Jiménez-Gámez 2010;Tomlinson 2005).In our study the Muslim community emerges as the most strongly rejected, regardless of the type of school, and there is strong stigmatisation of this cultural group who have arrived in Spain for reasons of economic migration.
All the studies in this field coincide in affirming that preferences in terms of friendship and relations among late-primary and secondary-school pupils are for boys and girls from one's own ethnic group, socio-economic status and sex.The more culturally diverse the school, the more interethnic relations within the pupil body increase, whilst they decrease when there is a small ethnic minority.At a more specific level, it is observed that the two pupil groups with the most difficulties in terms of making friends with those from other ethnic groups are African-Americans and whites in North American schools.Those pupils who do have friends from other ethnic groups present different characteristics, such as leadership attributes -and this would apply in the particular case of whites who strike up relationships with Afro-Americans in the United States (Graham and Cohen 1997;Hallinan and Williams 1989;Quillian and Campbell 2003).
In this study, Moroccan pupils, contrary to what might be expected, show high levels of interest in making a connection with Spaniards.Despite their widespread rejection by Spaniards, Moroccans are more open to striking up relationships with them than with Latin Americans.If the degree to which Moroccans make friends with others is low, it is, as we see it, because of the reciprocity factor: such friendships are more likely to flourish when kindness and openness are expressed by the other party (Goulner 1960).Despite the rejection experienced by the Moroccan community, this favourable attitude is noteworthy, and contradicts the widespread view among ethnic Spanish teachers and pupils that Moroccans "don't want to integrate", as reflected in the exploratory interviews that took place before respondents filled out the questionnaires.
Studies on "the burden of acting white" for Afro-American pupils or for white/ European pupils from poor backgrounds show that it represents a limitation that can be mitigated through school organisation, but it is often reinforced by a school context with unfavourable characteristics (Hallinan and Smith 1985;Hallinan and Williams 1989;Tyson and Darity 2005).
With a view to promoting inter-racial relations within the pupil body it is crucial to avoid segregation by communities in schools.Segregation of multicultural pupil bodies has consequences for the acquisition of knowledge and for how the pupils act when they become adults.In order to reverse such tendencies it is important to organise schools responsibly, which would include planning class composition and numbers of foreign pupils in the schools carefully, as well as avoiding a concentration of boys and girls from "other" cultures in particular schools (Burgess and Wilson 2005;Clotfelter 1999;Hallinan and Smith 1985;Moody 2001;Sánchez-Hugalde 2009).
As regards school organisation, in the schools analysed here, pupils are not grouped in classrooms according to the criterion of cultural diversity, but rather according to knowledge levels and behaviour.The effects of these divisions are exacerbated by xenophobia and the stereotypes held about the groups in question, not only making it difficult to strike up friendships with classmates of other ethnicities, but also, in the worst cases, fomenting attitudes of rejection.
Despite the fact that some studies argue for the irrelevance of the ethnic group, compared to other variables such as socio-economic disadvantage or poor academic results (Kingdon and Cassen 2010), in this research ethnic diversity emerges as a relevant factor both for the degree of xenophobia and knowledge acquisition.In the results of the PISA Study (OECD 2012), in maths ability Spanish pupils attained 492 points, whilst pupils with immigrant backgrounds scored just 439 points: a significant difference of 53 points.This difference is not appreciated in other European countries, such as Hungary or Slovakia, where those from immigrant backgrounds even score higher than native pupils (OECD 2012).This leads us to conclude that there are other variables, as well as ethnic origin, to be taken into account in relation to academic results and differences, such as the stigmatisation of and xenophobia towards certain ethnic groups in schools.
Studies on mutual perceptions between immigrant pupils and teaching staff highlight the fact that if teachers' perceptions of such pupils do not coincide with the pupils' perceptions of themselves, then racist attitudes can emerge on the part of teachers (Oller, Vila and Zufiaurre 2012).In such cases, teachers normalise xenophobic attitudes against immigrant pupils as something common among their age -as typical features of teenage peer groups.Furthermore, in-depth exploration of teachers' narratives show that they have negative and low expectations about immigrant pupils' achievements and easily stereotype their skills and capacities according to their nationalities.
This research not only looks at the perceptions of native pupils in multicultural classrooms, but also at teachers' pedagogical responses in facing these new classroom situations.Teachers still need more self-reflection and awareness about cognitive mechanisms and need to pre-empt against negative stereotyping of different pupils' nationalities.
Furthermore, the degree of xenophobia itself reflects the classroom climate, and the racial tension liable to emerge from a negative and stigmatised representation of the "other".The results of the qualitative research also reveal a failure of schools to apply anti-racist and intercultural programmes that would help to dilute and transform such racist attitudes.

CONClUSIONS
Xenophobic attitudes within the peer group against pupils of other ethnic origin are determined by economic and ethnic factors associated with certain groups, such as the Moroccan community, and by the degree of school segregation.The level of xenophobia is determined by two variables: negative stereotypes towards different ethnic communities and the number of pupils with an immigrant background in the classroom.Xenophobia in schools does not increase as the proportion of immigrant pupils in a school rises, rather the opposite occurs.There is a proportionally inverse relationship: the greater the cultural diversity, the less the xenophobia.In the schools studied in the present research, once the proportion of pupils with first-and second-generation immigrant backgrounds rises above 25 per cent, the levels of xenophobia begin to decrease in accordance with the ratio of natives to immigrants.The most rejected groups are Moroccans and Chinese.
On the other hand, it is surprising how the Moroccan community, the group most widely rejected by all the other ethnic groups, shows no reluctance to make friends and establish relationships with ethnic Spaniards.This is in contrast to the case of Latin American pupils, in whom such a favourable disposition cannot be so clearly seen.This leads us to highlight the importance of the education system and of the involvement of the community as a whole in efforts to, first of all, improve perceptions based on stereotypes and stigmatisation, and second, remain alert to this rejection, disdain and contempt towards Moroccans.
The distribution of pupils from different ethnic backgrounds in schools is a complex issue in relation to peer groups.It is not just that pupils prefer to make friends with pupils of their own ethnic origin; rather, there is a rejection and negative appraisal of potential relationships with "others".This implies the need for educational intervention in this area, as regards the organisation of both schools and classrooms, and this issue cannot be left at the mercy of the pupils themselves following a laissez-faire attitude.
Neither the organisation of the schools concerned nor the educational methodologies employed can alone mitigate these xenophobic attitudes, as has been seen in the qualitative work and from the data collected.Moreover, teachers are not aware of the existence of stereotypes and xenophobia among pupils nor of their own stereotyping perceptions.They consider these attitudes common among teenagers and as a result no pedagogical responses are applied.Furthermore, they have low expectations concerning immigrant pupils and this could have a self-fulfilling prophecy that results in increased dropouts.
The educational system has a substantial responsibility for setting up the mechanisms best suited to fostering a reduction of xenophobia and racism in schools.Such mechanisms and strategies might include social policies, which promote teachers as reflective agents, skilful experts and lifelong learners, as well as intercultural and citizenship education (Banks 2011;Caena 2014;Deusdad 2013;Osler and Starkey 2002) that would involve pupils' contact with NGOs and other types of international associations, and the incorporation into the curriculum of topics or subjects based on an anthropological perspective that provides knowledge and understanding of other cultural and historical practices.

Table 2 :
general xenophobia (xenophobia1) according to sex, educational level and socio-economic status.
Note: The socioeconomic variable is obtained from the mother's/father's or guardian's work situation and the resources the pupils have for studying.Both types of information are provided in the questionnaire.
Table4reveals which nationalities are most rejected, and which groups are those that exercise such disdain.As in the case of stereotypes, it is the Moroccan community that is subjected to the highest levels of xenophobia, from all the ethnic groups analysed.

Table 4 :
Degree of xenophobia (xenophobia2) by nationality School Xenophobia and Interethnic RelationshipsThe categorical variable Xenophobia2 is obtained through two pupil questionnaire items.The first inquired about the degree of acceptance (or rejection) they perceived from each one of the nationalities; the second asked which nationality they disliked (if any).

Table 5 :
School segregation Note: The category 75-100% is not shown, because there were no such schools in the sample.