The literature examining the relation between ethnic school composition and pupils competences present several theoretical frameworks and explanations. The main theoretical frameworks referred to in studies examining the relationship are the reference theory, the resource theory and the (language)contact theory, which will be briefly discussed. The research questions are theoretically addressed and five hypotheses are derived from the main theoretical frameworks and additional arguments. One has to notice that the aim of the study is not to examine the mechanisms that are underlying within these specific theories. These theories will only form a foundation for the argumentation of the hypotheses. This chapter closes with two conceptual models (figure 2.1 and 2.2) which give an overview of the expected effects that are central in this research.
2.1| Theoretical frameworks
2.1.1 Reference-group theory
Within classes several comparison processes between pupils take place (van Haalen, 2008). According to the reference-group theory of Richer (1976) two forms of reference processes can be distinguished: a normative function and a comparative function. According to the normative reference theory, the group determines the norms, attitudes and standards which the person has to comply to. The standard of the group is a comparison point that the person can use while evaluating him/herself and others when using the comparative reference theory (Guldemond & Meijnen, 2000; Belfi, de Fraine & van Damme, 2010). In a heterogeneous class the low-scoring student can choose the group of high-scoring students as normative reference group. The norms and values that prevail in this group will positively influence the school achievements. However, the low-scoring student can choose the group of high-scoring students as a comparative reference group, which then will constitute a negative self-image and relative deprivation. Whether or not a weak student chooses the normative or the comparative comparison, is according to Richer (1976) dependent of the possible opportunities for upward mobility. If upward mobility is perceived as feasible, normative reference group processes will appear, if not, comparative processes will appear.
According to the reference theory, immigrant students will compare themselves with non-immigrant students and be motivated for achieving higher school results. One assumes that ethnic minorities profit when their fellow students are of native origin. Positive values and attitudes will be transferred to immigrant students. Since in general, native students obtain higher school results than non-native students (Inspectie van het Onderwijs, 2011).
2.1.2 Resource theory
Another explanation for disparities in pupils' educational performance and the influence of school composition can be found in the concept of 'capital'. It is argued that the so-called cultural and social capital, but also variants as the linguistic capital, available to individuals are relevant sources for educational performance (Driessen, 2002). Bourdieu's (1977) cultural capital thesis argues, in terms of education and schooling, that some students possess the social class, family, cultural background and dispositions that enable them to benefit from the school environment and its facilities more efficiently than those who do not. Pupils themselves can be seen as resources since they learn from each other (Driessen, 2002). Ethnic minority pupils generally have fewer resources available to them than non-minority pupils. When numerous minority children like this attend a single school, the pupils in the school are obviously placed in a position to benefit less from the resources or 'capital' of other pupils. Pupils' language development in such a situation can also be inhibited by a lack of contact with the majority language or the presence of extreme ethnic and linguistic diversity.
2.1.3 (language) Contact theory
According to the contact theory of Allport (1954), contact with out-group members reduces negative out-group attitude. A cognitive mechanism underlying the relationship between intergroup contacts and intergroup attitudes is that intergroup contacts lead to more knowledge about and more accurate perceptions of people from other ethnic groups, resulting in less prejudice (Vervoort, Scholte & Scheepers, 2011). This more accurate perception would be the result of more mutual information about norms and values, lifestyle and experiences (Havekes & Uunk, 2008). This positive effect would only occur in contact-situations marked by four key conditions: equal group status within the situation, having common goals, intergroup cooperation and support of relevant institutions and authorities.
Follow-on the contact theory is the language-contact thesis which assumes that native students are a source of help for immigrant students. When immigrant students are keeping contact with the Dutch language, e.g. a significant part of the student body is native, their language development will be stimulated. A better control of the Dutch language will eventually lead to higher scores on other school achievements. The language-contact theses assumes that pupils in classes with a high percentage of immigrant students will perform worse, since there is little opportunity to get in touch with the Dutch language, or rather, with children that originally control the language at a good level. This lack of language contact, negatively influence the Dutch language development, and thus also their performance on other disciplines.
2.1.4 Overview of theories
The above described three theories are mostly mentioned in research of the relationship between ethnic school composition and pupils' competences. Although the underlying mechanisms found in each theory differ from each other, they all predict the outcome to be in the similar direction. All predict that natives and immigrants learn from each others' values and attitudes; immigrant students will be stimulated in their language development and benefit from the knowledge and resources when attending a school with native students; native students do not suffer from negative influences when attending a school with immigrant students. Since the used data provide not enough variables to examine the mechanisms of the above described theories, I choose to elaborate more on the three components of ethnic school composition: ethnic concentration, ethnic diversity and majority/minority position regarding ethnicity. Therefore, hypotheses will be formulated per element of school composition and their relation with pupils competences.
With regard to the ethnic concentration, the expectation is that pupils which are attending a school with a majority of immigrants and pupils of lower socioeconomic origins, perform qualitatively worse on language and mathematics, compared to pupils, which are attending schools of which the student body consists of many native high-milieu students (Driessen, 2002). This prediction is based on the idea that the latter group introduces the possibility of resources, of which the less privileged pupils could profit as well. Besides that, the consequences of comparison processes are relevant: immigrant and lower class pupils which are attending a school with mostly native pupils from higher origin have contact with the other group and therefore increase their performance. Moreover is the following differential effect expected: the achievements of high-milieu pupils at concentration schools do not abide under the presence of low-milieu immigrant pupils. The following hypothesis is formulated considering ethnic concentration:
1. 'Native and immigrant pupils will perform better on cognitive and social competences when attending a school with a higher percentage of native pupils'.
When regarding the ethnic diversity, the expectation arises that pupils on ethnic heterogeneous schools are dealing with lower achievements, than ethnic homogeneous schools. This prediction is based on the idea that the teaching on homogeneous schools is less problematic and therefore leads to better results than on heterogeneous schools (Driessen, 2002). The class composition has major implications for the context in which teachers work. Teachers adjust their teaching methods, discussion of the subject matter and evaluation techniques on the average level of the class. Homogenous classes require an education design that fits the specific interests and needs of the class. In contrast, in heterogeneous classes exclusive frontal instruction is less designated. In these classes some degree of internal class differentiation must take place (Belfi, et. al., 2010).
Moreover, high levels of diversity are expected to enrich students' morale, social and citizenship education (Driessen, 2007). Divers classroom contexts are expected to increase opportunities to teach students to cooperate with students from different backgrounds, to discuss differences, stimulate creativity and to cope with different norms and values (Banks, 2008). From a citizenship of multicultural society perspective, diverse classroom contexts are expected to benefit students attitudes towards classmates' ethnic-cultural backgrounds. In addition, the Dutch government once departed from the assumption that desegregation (i.e. the increase of ethnic diversity within schools) enhances minority students' achievements, and promotes favorable intergroup attitudes which are seen as a precursor for societal integration. The following two hypotheses are formulated considering ethnic diversity:
2. 'The more ethnic homogenous the composition of the school is, the better native and immigrant pupils perform on cognitive competences'.
3. 'The more ethnic heterogeneous the composition of the school is, the better native and immigrant pupils perform on social competences'.
When taking a majority/minority position, it is at first expected that consequences will possibly arise on the non-cognitive effect-measurements. Furthermore, it is expected, that pupils, which obtain the majority position on basis of comparison processes, feel faster socially approved; they position themselves more favourable within class, feel better and are more confident. Pupils, that obtain the minority position in terms of their ethnicity, will achieve lower scores on social-emotional aspects. Earlier research has found that attending a school with a high proportion of your ethnic group has a positive effect on school attachment (Fekjaer & Birkelund, 2007). Therefore, attending a school with a high proportion of immigrant pupils would be positive for an immigrant pupil, but not for a native pupil and vice versa. The following hypothesis is formulated considering majority/minority position:
4. 'Native and immigrant pupils will perform better on cognitive and social competences when they obtain the majority position at school in terms of their ethnicity'.
It is expected that attending a school with a higher ethnic concentration and ethnic diversity negatively influences pupils' cognitive competences. However, it can be expected that the home environment of native pupils acts as a buffer. Since native pupils speak Dutch at home they develop their language in this way. The language problems of immigrant pupils will have in less extent an impact on their school performance than for immigrant pupils (Karssen, et. al., 2011). (Nog opzoek naar betere argumenten/literatuur, etc.) (Ik heb een tweetal studies over rol van taalvaardigheid op prestaties met ook argumenten voor verschillen tussen autochtone en allochtone leerlingen ïƒ nog verwerken
5. 'The effects of ethnic school composition (ethnic concentration, ethnic diversity and majority/minority position) on pupils' competences is stronger for immigrant pupils than for native pupils'.
2.3| Conceptual models
All independent variables and their expected influences on pupils' competences are shown in figure 2.1 and 2.2. A plus sign indicates a positive effect and a minus sign indicates a negative effect. At the individual level four control variables are included: origin, immigrant status (no immigrants vs. first generation immigrants), sex and age.