Printer Friendly

Outgroup prejudice among secondary pupils in northern England: are the predictors at the individual, school or neighbourhood level?

Introduction

The aim of the present article is to examine what factors best predict levels of outgroup prejudice among Christian, Muslim and secular youth. This was done by administering the Outgroup Prejudice Index (OPI) among separate samples of the three groups in secondary schools in areas of England where these three groups are clearly visible.

The OPI is a scale of outgroup prejudice based on notions of proximity developed among Christian, Muslim and secular secondary school pupils (Brockett, Village and Francis, 2010). Previous study of a different sample of pupils in these areas had shown that notions of physical and social distance could be used to create scales for White attitudes towards Muslims (Brockett, Village and Francis, 2009). Built on this earlier work, the OPI is a scale that operates in a comparable way for both Asian/Muslim groups and for Whites who are Christian or of no religious affiliation.

We envisaged that the factors that best predict levels of outgroup prejudice among Christian, Muslim and secular youth might be related to individual pupils (their sex, age, friendships and religion), to where they live (social deprivation and ethnic make-up) and to their school (type of school, size, ethnic make-up, levels of social deprivation, academic achievement and the neighbourhood in which the school is located).

It must be said that attitudes towards outgroups in the samples were generally found to be positive or neutral rather than negative. Although outgroup prejudice was a minority position, it remains important to examine what factors predict it so that any perceived change in the future can be measured and monitored.

Method

Schools

Secondary schools in three areas of northern England (Blackburn, Kirklees and York) were asked to participate in the study, and twenty-three agreed to do so. Of these, twenty provided at least forty pupil responses, and these schools were chosen for this analysis. Of the twenty schools, twelve were state comprehensive schools, four were voluntary aided church schools (two Roman Catholic and two Church of England), and four were independent schools. All but one were mixed-sex schools. Schools were asked to provide data on their roll number (RN), percentage of ethnically White pupils (%White), percentage of pupils eligible for free school meals (%FSM) and the percentage of eligible children that passed five or more GCSEs at grade C or above (%5GCSE). The latter is a standard measure of attainment for sixteen year olds. In a few cases data were not available for 2007 and were estimated from other years.

Neighbourhoods

Statistics on levels of social deprivation and ethnic make-up were used to relate pupils and schools to their neighbourhoods. We used Lower Super Output Areas (LSOAs) data from the 2001 National Census (CASWEB, 2003) for all the school catchment areas. LSOAs are geographic areas that cover approximately 1500 residents (with a minimum of 400 households and a range of 1000 to 1800 residents). Data available at LSOA level included several key statistics, such as the proportion of various ethnic or religious groups. The percentage of Whites in the relevant LSOA was used as a proxy measure of ethnic and religious diversity within a pupil's home neighbourhood. Also linked to LSOAs was an index of multiple deprivation (IMD). The IMD is based on a range of social indicators including income, employment, health and education (Communities and Local Government, 2008). It is available through the website of the Office of National Statistics (ONS, 2011). Thus, for each pupil, the home neighbourhood data used in this analysis consisted of the IMD and percentage Whites. A similar procedure was used for the twenty schools in the sample, allowing assessment of the IMD and ethnic religious diversity of the neighbourhood in which each school was set (see neighbourhood data in Table 1).

The use of percentage Whites as a proxy for ethnic and religious diversity requires explanation. Across the three areas in which the study was based, ethnicity and religious affiliation are strongly correlated, reflecting the fact that most of the region is ethnically White and religiously Christian, and by far the largest ethnic minority in the region were Muslims of Asian origin. The catchment areas of the schools in the study comprised a possible 19, 975 LSOAs, and the religious and ethnic make-up of these are presented in Table 2a. The average ethnicity was 94.1 per cent White and 4.0 per cent Asian, though figures for individual neighbourhoods varied from virtually 0 per cent to 100 per cent in each case. A small proportion of neighbourhoods (0.7 per cent) had over 80 per cent Asians but over half (53.8 per cent) recorded none at all. Given this ethnic and religious mix in the region, there were predictably very high correlations between ethnic and religious make-up of LSOAs (see Table 2b). Given the nature of the data and the high sample size, all correlations were highly statistically significant; however, high levels of White ethnicity were most strongly predictive of low levels of Asian ethnicity (r = -0.97) and Muslim religion (r = -0.95). The correlation with Christian religion was lower (r = 0.83) because Whites were more likely than Asians to be religiously unaffiliated. Overall, these data suggest that the proportion of ethnic Whites in a neighbourhood was a strong negative predictor of the levels of Asian ethnicity and Muslim religion, and using a single indicator in this way is therefore justified.

Participants

Class teachers gave questionnaires to pupils during normal school activities between 2007 and 2008. All pupils were assured of anonymity and confidentiality and given the opportunity to opt out of the survey. Response rates were high, with nearly all pupils agreeing to complete the questionnaire. Over 95 per cent of participants indicated their religion as either 'no religion', 'Christian' or 'Muslim', and 2502 of these whose responses had no relevant missing data comprised the study sample, details of which are given in Table 1 above.

Instruments

The OPI (Brockett, Village and Francis, 2010) is a six-item scale with a high score indicating greater levels of outgroup prejudice. As a measure of reliability, its Cronbach's alpha coefficiant is 0.85, and its validity has been tested against independent measures of religious and racial stereotyping. The items in the scale use social distance as an indication of prejudice against those of a different race or religion. In the first four items pupils were asked to rate their attitude towards the idea of families moving in next door who were Asian, Muslim, Black or Sikh (for White pupils) or White, Christian, Black or Sikh (for Asian pupils). Responses on a five-point scale ranged from 'I would love it' (scored 1) to 'I would hate it' (scored 5). In the remaining two items pupils were asked to respond to the following statements: 'People of a different race/colour should not hang out together' and 'People of a different religion should not hang out together'. In each case possible responses ranged from 'strongly disagree' (scored 1) to 'strongly agree' (scored 5).

The Astley-Francis Scale of Attitude Toward Theistic Faith (or TBS--Theistic Belief Scale--for short) was used to measure religious affect related to generalised theistic belief (Astley et al., in press). It consists of seven Likert items ('I know that God is very close to me'; 'God means a lot to me'; 'God helps me'; 'God helps me to lead a better life'; 'I know that God is very close to me'; 'Prayer helps me a lot'; 'I find it hard to believe in God'; and 'I think going to a place of worship is a waste of my time') with a five-level response scale that is reverse coded for the last two items. This scale was considered to be a better measure of religious salience than simple religious affiliation. The scale had a high internal consistency among pupils in this study (Cronbach's alpha = 0.96), and was positively correlated with items measuring frequency of attendance at services (r = 0.52, df = 2756, p < 0.001) and religious salience (r = 0.69, df = 2746, p < 0.001).

Other key variables were measured using various other instruments. Outgroup contact was assessed by a single question asking respondents how many friends they had who were of a different race. Answers were recoded into an ordinal scale such that 0 = none, 1 = 1, 2 = 2-5, and 3 = > 5. Sex was considered an important predictor variable to include in the model because it has been shown to be related to religious belief (Argyle and Beit-Hallahmi, 1975; Beit-Hallahmi and Argyle, 1997; Francis, 1997; Walter and Davie, 1998) and to prejudice (Hoxter and Lester, 1994; Altemeyer, 1998; Ekehammar, Akrami and Araya, 2003). It was coded as 0 = male and 1 = female. Age group (1 = 10-13 years, 2 = 14-16 years, 3 = 17-19 years) was also included because of the possibility that both prejudice and religiosity might vary during adolescence and to control for variations in age distributions between schools.

Linking neighbourhood, pupil and school data

Participants were asked to give the full postcode of their home address, and this was used to calculate demographic variables for pupils' home neighbourhoods. Some pupils were excluded from the analysis because they gave no postcodes, incomplete postcodes or postcodes that could not be found in the national dataset. Neighbourhood statistics for individual LSOAs were then matched to postcodes using data generated from the UKBorders website (Edina, 2011) and postcodes in use in 2007. This procedure was repeated for schools using school postcodes.

Analysis

The data thus included variables measured at individual, school and neighbourhood level. The standard way of assessing the effects of community characteristics on individual attitudes or behaviours is to use multilevel regression models (Snijders and Bosker, 1999; Hox, 2002; Bickel, 2007), especially if subjects are clustered into particular contexts (such as pupils in the same school). Multilevel regression models not only adjust for lack of independence between sample members who reside in the same community, but also allow the effects of contextual-level variables (such as school characteristics) to be correctly specified.

In this study, the individual-level explanatory variables were those associated with pupils, such as their sex, age, religious affiliation and religious attitude. Pupils were potentially grouped in two independent ways: by living in the same neighbourhood and by going to the same school. Although in theory there could be an element of grouping at neighbourhood level (if lots of pupils came from the same neighbourhoods), in practice very few pupils came from the same LSOA, so effects related to home neighbourhood were treated as level 1 fixed effects. Pupils were, however, strongly grouped in schools, and so the most appropriate statistical model was a linear mixed model, using school as the subject (grouping) variable.

The initial analyses were therefore based on a mixed linear model with school as the grouping variable and fixed variables related to pupils, namely sex, age, friends of a different race, religious affiliation and TBS score. This analysis indicated a marked difference between White pupils (Christians or those of no religion) and Muslim pupils, and so it was decided to treat these two groups separately. All continuous variables were grand mean centred to reduce the effects of multicollinearity (Bickel, 2007). For Muslims, variance in the OPI associated with school contextual variables was very low, and a generalised linear model fitted to pupil data only was more appropriate than a mixed level model.

Results

Basic metrics for the variables in the analysis are shown in Table 3. Mean scores on the TBS were highest for Muslims (32.9, SD = 2.7, n = 459), intermediate for Christians (21.9, SD = 6.6, n = 1216) and lowest for those of no religion (13.6, SD = 5.5, n = 781). The TBS was also significantly higher in female (22.2, SD = 8.6, n = 1380) than male (20.3, SD = 8.9, n = 1076) pupils, and in pupils aged 10-13 (22.9, SD = 8.2, n = 1195) compared with those aged 14-16 (20.1, SD = 9.1, n = 1026) or 17-19 (18.9, SD = 8.4, n = 235). The most religious pupils thus appeared to be young female Muslims, and the least religious appeared to be older males of no religion.

Relationships between the OPI and the various variables were first tested using a multilevel model on the overall data. The first step was to fit a null model (that is, one without any predictor variables, but which allows for the fact that observations were grouped by school) to assess how much of the variation in the OPI was associated with factors measured at the school contextual level (Table 4). The Intraclass Correlation Coefficient (ICC) is a standard measure of the proportion of variation in the dependent variable that can be explained by the grouping variable, and for these data it indicated that around 6 per cent of the variation in outgroup prejudice among pupils was related to variations in their schools. Although this seems small, the level indicated that multilevel modelling was an appropriate method for these data because there was significant variation in the OPI related to differences between schools rather than between pupils. As expected, adding the pupil-level data did little to reduce the ICC (because this is related to school-level variation), but it did significantly improve the fit of the model by reducing the overall amount of unexplained (residual) variance.

The results of these tests on the overall data are as follows. The OPI, and thus levels of outgroup prejudice, was higher among males than females and among younger pupils (especially 14-16 years olds), but lower among those who had friends of a different race compared with those who had none. In terms of religion, the results were somewhat surprising. In the overall data, mean OPI scores were lower for Muslims (14.4, SD = 4.0) than for either Christians (15.7, SD = 4.5) or those of no religion (15.9, SD = 4.1). However, after controlling for religious salience with the TBS, Christians appeared to be significantly more prejudiced than those of no religion. The OPI was strongly negatively correlated with attitude towards theistic belief, implying that religious salience reduced prejudice. However, among White (non-Muslim) pupils, affiliating with Christianity was associated with slightly higher levels of outgroup prejudice at any given level of theistic belief compared with those of no religion (Figure 1). Muslims had generally high levels of religious salience and correspondingly low levels of prejudice. As a result of the marked differences between Muslims and others in this initial analysis, the two groups were treated separately thereafter.

[FIGURE 1 OMITTED]

Individual, neighbourhood and school effects among White pupils

Multilevel model analysis for White pupils only is shown in Table 5, starting with the null model, then successively adding individual variables, neighbourhood variables and school-level (level 2) variables. The results for individual variables were as reported for the overall dataset, with significant positive effect on the OPI for males, 14-16 year olds and Christians, and significant negative effect for having friends of a different race and positive attitudes to theistic belief. When neighbourhood variables were added, prejudice was significantly increased by both higher levels of deprivation and higher levels of ethnic Whites in the neighbourhood (0.025 and 0.024 respectively). The effects were small, however, and the significance of the latter moved to the 10 per cent level in the final model. When school level variables were added they successfully explained the variance in OPI between schools: outgroup prejudice was slightly (but significantly) higher in church schools (compared to state schools) and slightly (but significantly) lower in schools with higher proportions of pupils with either eligibility for free school meals or GCSE passes (0.843 and -0.107 and -0.105 respectively).

Individual and neighbourhood effects among Muslim pupils

Variation in the OPI between schools was not significant for Muslim pupils, and so fitting a multilevel model of school contextual variables was inappropriate. Instead a generalised linear model was fitted to individual and neighbourhood variables, using a comparable method to the analysis for White pupils, as shown in Table 6. For Muslim pupils, the main predictors of variation in the OPI were sex (higher among males) and friends of another race. The latter significantly reduced prejudice in this group (-2.099). Neighbourhood effects were negligible, although there was a small but statistically significant effect of %Whites in the neighbourhood (-0.015), suggesting that, after allowing for individual factors, Muslims living with a higher percentage of Whites were less prejudiced than those living in areas with fewer Whites. The effect of religion in this group was difficult to assess because the scores for the TBS were uniformly high and showed little variation. The generally low levels of prejudice among Muslims compared with other pupils could be linked to the high salience of their religion. However, without a higher proportion of non-religious Asians in the sample this is difficult to test. What is true is that Muslim belief/affiliation was not associated with above average levels of outgroup prejudice.

Discussion

So are the predictors of outgroup prejudice at the individual, school or neighbourhood level? Six main findings emerge from this study, some predictable from other work on prejudice, but others apparently new. The evidence to support the findings is laid out in the various tables below. The new findings implied by the results suggest that outgroup prejudice is predicted by somewhat different factors among White pupils in the region than among Muslims.

1 Girls are less prejudiced than boys. There was strong evidence for this for both Whites and Muslims.

2 Having friends of a different race reduces prejudice. There was strong evidence for this also for both Whites and Muslims.

3 Mid-teens are more prejudiced than other adolescents. Prejudice peaked among mid-teens for Whites, but not Muslims.

4 Christian affiliation is associated with increased prejudice, although greater religious salience reduces prejudice.

5 Home neighbourhood has a slight effect on Muslim pupils' prejudice.

6 School effects are only evident for White pupils, and school neighbourhood has rather little effect.

1. Sex. That females are less prejudiced than males is a well-known trend documented in much sociological research on prejudice (Ekehammar and Sidanius, 1982; Bierly, 1985; Qualls, Cox and Schehr, 1992; Altemeyer, 1998), including some specifically on adolescents (Moore, Hauck and Denne, 1984; Hoover and Fishbein, 1999). It is true that this is not a universal trend: research suggests, for example, that females have more implicit prejudice than males (e.g. Ekehammar, Akrami and Araya, 2003) and that there is no difference in terms of closer social relations (e.g. Hoxter and Lester, 1994). In our sample, the individual male score was significantly higher for males (1.553) than for females (see Table 4). It is interesting to note that Bevelander and Otterbeck (2010: 10-12) found no difference between the sexes as a whole in attitudes towards Muslims among 15-19 year olds in Sweden, although the individual year groups did show some differences between boys and girls.

2. Friendships. That having friends of a different race reduces outgroup prejudice is also an expected and well-known trend (e.g. Allport, 1958; Jackman and Crane, 1986; Hamberger and Hewstone, 1997; Pettigrew, 1997, 1998; Levin, van Laar and Sidanius, 2003; Pettigrew and Tropp, 2006). This trend is also present in studies with children and adolescents specifically (Aboud, Mendelson and Purdy, 2003; McGlothlin, Killen and Edmonds, 2005; McGlothlin and Killen, 2006; White et al., 2009). There is also a correlation between having friendships with people of a different religion and reduced prejudiced towards people of different religions (e.g. Paolini et al., 2004; Pettigrew and Tropp, 2006; Pettigrew et al., 2007; Lee et al., 2009).

3. Age. The literature on outgroup prejudice suggests that racial and ethnic bias is present from about the age of four and begins to decline from around the ages of five to seven (e.g. Aboud, 1988; Doyle, Beaudet and Aboud, 1988; Powlishta et al., 1994). For outgroup prejudice during the adolescent stage, however, the findings vary. In our sample, prejudice peaked among mid-teens for Whites (although not Muslims), but that mid-teens in particular are more prejudiced than other adolescents is not evidenced in literature. There is research that suggests a pattern of lower prejudice in earlier than in later teenage years (White et al., 2009), and in particular increasing between the ages of ten and fourteen to sixteen (Black-Gutman and Hickson, 1996; Rutland, 1999), remaining stable between fourteen and eighteen and then increasing again between eighteen and twenty-three (Hoover and Fishbein, 1999). Conversely, Robinson, Witenberg and Sanson (2001: 86; Bevelander and Otterbeck, 2010: 3) claim that research shows that as adolescents get older they become more tolerant of people with different beliefs. And there is also research that finds no age-related differences between eleven and sixteen (Moore, Hauck and Denne, 1984), and research that shows explicit prejudice declining from age ten while implicit prejudice remains stable from age six to adulthood (Baron and Banaji, 2006).

4. Religion, specifically Christian affiliation. This is the most interesting finding--and the most complicated. Our data shows that Christian affiliation is associated with increased prejudice, although greater religious salience--a more positive attitude towards theistic religion--reduces the prejudice. The three previous individual factors (sex, friendship and age) are well evidenced by other research, but this finding is relatively new. The analysis was complicated by the fact that Muslim pupils are very different from non-Muslim pupils with respect to religious salience and so had to be treated separately for some analyses.

A curvilinear relationship between Christianity and prejudice, with Christians being generally more prejudiced than non-religious people but prejudice declining with increased salience, is a pattern to emerge from other studies (e.g. Gorsuch and Aleshire, 1974; Batson, Schoenrade and Ventis, 1993; Altemeyer, 1996). A more limited number of studies have also reported this pattern in Europe (Bagley, 1970; Scheepers, Gijsberts and Hello, 2002) and our research has added significant new evidence to support these interesting findings. However, there is evidence from elsewhere that seems to challenge this pattern. There have been two recent studies on prejudice towards Muslim practices among European Christians that found that while a majority of Christians are no more prejudiced than non-Christians (Fetzer and Soper, 2003; Saroglou et al., 2009), those with 'orthodox' beliefs are more prejudiced (Saroglou et al., 2009). Additionally, research seeking to explain such findings has concentrated on the factors within religiosity that encourage or discourage prejudice and has identified religious fundamentalism as a factor that increases prejudice (Altemeyer and Hunsberger, 1992; Wylie and Forrest, 1992; Kirkpatrick, 1993; Laythe, Finkel and Kirkpatrick, 2001). Such findings suggest a complex relationship between religiosity and prejudice and provide impetus for further research.

5. Home neighbourhood (social deprivation and ethnic make-up). These findings were found to have a slight but not very strong effect. There seems be consensus in the literature that greater neighbourhood deprivation equates with less social cohesion (e.g. Tolsma, van der Meer and Gesthuizen, 2009; Laurence, 2011) but findings from research on the effects of ethnic diversity are more mixed. Some argue that more diversity equates with less prejudice (e.g. Oliver and Wong, 2003; Semyonov and Glikman, 2009), while others argue that it equates with less cohesion (e.g. Alesina and Ferrara, 2002; Putnam, 2007).

The first finding from our sample was that for White pupils higher prejudice was associated with living in more socially deprived areas. Other research has found that Swedish children in higher socio-economic homes had more positive attitudes towards Muslims (Bevelander and Otterbeck, 2010: 15). This accords with the 'power-threat hypothesis' whereby negative attitudes arise towards groups seen as economic, social and/or political competitors (Bevelander and Otterbeck, 2010: 4, citing Blalock, 1967). The second finding from our sample was that for Muslim pupils there was a slight indication that those living in areas with more White residents were less prejudiced than those who lived in areas with more Asians and/or Muslims. This accords with Allport's (1958) contact hypothesis, which suggested that outgroup attitudes become more positive with increased contact between groups of similar socio-economic background.

6. School (type of school, size, ethnic make-up, levels of social deprivation, academic achievement and the neighbourhood where it is located). School effects are only evident for White pupils and may be associated with going to church schools (higher prejudice), having a bigger mix of disadvantaged children (lower prejudice) and better academic standards (lower prejudice). School neighbourhood has rather little effect.

That the four church schools showed slightly increased prejudice despite higher Christian affiliation and salience may be explainable by the increasing sense of religious identity they foster. Little empirical research has been published on prejudice in faith schools, although some research into Catholic schools published in the 1990s suggests that they do not encourage students to be more prejudiced than others and are not necessarily divisive (Grace, 2003). However, this issue needs further investigation. That attending a school with more socially disadvantaged children (as evidenced by numbers of free school meals) reduced levels of prejudice for White pupils while living in more socially deprived areas increased it seems to fit with Allport's (1958) hypothesis that contact engenders tolerance. The reduction in prejudice associated with better academic standards was also found by Bevelander and Otterbeck in Sweden where higher grades at school correlated with higher socio-economic background and with more positive attitudes towards Muslims (2010: 13).

Conclusion

This study set out to examine the Outgroup Prejudice Index and assess what factors best predict levels of outgroup prejudice among eleven- to sixteen-year-old adolescents living in northern England. We envisaged that these factors would be related to individual pupils (their sex, age, friendships and religion), to where they live (social deprivation and ethnic make-up) and to their school (type of school, size, ethnic make-up, levels of social deprivation, academic achievement and the neighbourhood where it is located).

Although the effects were small the results were statistically significant and six main conclusions emerged from the data.

First, in terms of individual factors, it is clear that sex is significant in shaping levels of outgroup prejudice among young people in the present study: girls were less prejudiced than boys.

Second, also in terms of individual factors, it is clear that having friends of a different race reduced young people's outgroup prejudice. There was strong evidence for this for both Whites and Muslims.

Third, again in terms of individual factors, it is clear that age played a significant role in shaping levels of young people's outgroup prejudice. Those in their mid-teens had less positive attitudes towards outgroups. Both before and after this period their outgroup prejudice was significantly less.

Fourth, again in terms of individual factors and perhaps most interestingly, it is clear that being Christian plays a more significant role in raising levels of young people's outgroup prejudice than being Muslim or secular. Overall, Muslim and secular youth growing up in northern England show lower levels of outgroup prejudice than is the case among young Christians. When religious salience is taken into consideration, however, Christian prejudice is significantly reduced. This finding deserves further research.

Fifth, the data shows that after taking these more significant individual factors into account, home neighbourhood factors (social deprivation and ethnic make-up) have some slight effect in shaping outgroup prejudice within the overall model. In this limited sense home neighbourhood is important.

Sixth, the effects of school factors (type of school, size, ethnic make-up, levels of social deprivation and academic achievement) are only evident for White pupils and may be associated with going to church schools (higher prejudice), having a bigger mix of disadvantaged children (lower prejudice) and better academic standards (lower prejudice). School neighbourhood has rather little effect.

Taken together these findings demonstrate that individual differences in outgroup prejudice among young people are a complex function primarily of individual factors (their sex, age, friendships and religion), and secondarily of aspects of residence and school type. The residential aspects were social deprivation and ethnic make-up and the school aspects were type of school, size, ethnic make-up, levels of social deprivation, academic achievement and the neighbourhood where it is located. Building on previous research, the present study has contributed to knowledge by succeeding in measuring outgroup prejudice among secular, Christian and Muslim and young people within the same study and thereby confirming the usefulness of the Outgroup Prejudice Index in extending the scope of studies in the psychology of prejudice within a multi-faith context.

Note

Help with the data from Dr Andrew Village, York St John University, is very gratefully acknowledged.

References

Aboud, F. E. (1988), Children and Prejudice, New York: Basil Blackwell.

Aboud, F. E., Mendelson, M. J. and Purdy, K. T. (2003), 'Cross-race peer relations and friendship quality', International Journal of Behavioural Development, 27, 165-73.

Alesina, A. and Ferrara, E. (2002), 'Who trusts others?', Journal of Public Economics, 85, 207-34.

Allport, G. W (1958), The Nature of Prejudice, Reading, MA: Addison-Wesley.

Altemeyer, B. (1996), The Authoritarian Specter, Cambridge, MA: Harvard University Press.

--(1998), 'The other "authoritarian personality"', Advances in Experimental Social Psychology, 30 (1), 47-92.

Altemeyer, B. and Hunsberger, B. (1992), 'Authoritarianism, religious fundamentalism, quest, and prejudice', The International Journal for the Psychology of Religion, 2, 113-33.

Argyle, M. and Beit-Hallahmi, B. (1975), The Social Psychology of Religion, London: Routledge and Keegan Paul.

Astley, J., Francis, L. J. and Robbins, M. (in press), 'Assessing attitude toward religion: the Astley-Francis Scale of Attitude toward Theistic Faith', British Journal of Religious Education.

Bagley, C. (1970), 'Relation of religion and racial prejudice in Europe', Journal for the Scientific Study of Religion, 9 (3), 219-25.

Baron, A. S. and Banaji, M. R. (2006), 'The development of implicit attitudes. Evidence of race evaluation from ages 6 and 10 and adulthood', Psychological Science, 7, 53-8.

Batson, C. D., Schoenrade, P. and Ventis, W L. (1993), Religion and the Individual: A Social Psychological Perspective, New York: Oxford University Press.

Beit-Hallahmi, B. and Argyle, M. (1997), The Psychology of Religious Behaviour, Belief and Experience, London: Routledge.

Bevelander, P and Otterbeck, J. (2010), 'Young people's attitudes towards Muslims in Sweden', Ethnic and Racial Studies, 33 (3), 404-25.

Bickel, R. (2007), Multilevel Analysis for Applied Research: It's Just Regression!, New York: The Guilford Press.

Bierly, M. M. (1985), 'Prejudice toward contemporary outgroups as a generalized attitude', Journal of Applied Social Psychology, 15, 189-99.

Black-Gutman, D. and Hickson, F. (1996), 'The relationship between racial attitudes and social cognitive development in children: an Australian study', Developmental Psychology, 32, 448-56.

Blalock, H. M. (1967), Toward a Theory of Minority-Group Relations, New York: Wiley.

Brockett, A., Village, A. and Francis, L. J. (2009), 'Internal consistency reliability and construct validity of the Attitude toward Muslim Proximity Index (AMPI): a measure of social distance', British Journal of Religious Education, 31, 241-9.

--(2010), 'Assessing outgroup prejudice among secondary school pupils in northern England. Introducing the Outgroup Prejudice Index', Research in Education, 83, 67-77.

CASWEB (2003), 2001 National Census, Output Area Boundaries, Census Dissemination Unit: University of Manchester. Available at: http://casweb.mimas.ac.uk. Census output is Crown copyright and is reproduced with the permission of the Controller of HMSO and the Queen's Printer for Scotland. Last accessed July 2011.

Communities and Local Government (2008), The English Indices of Deprivation 2007, Wetherby: Communities and Local Government Publications. Available at: www.communities.gov.uk/publications/communities/indiciesdeprivation07. Last accessed July 2011.

Doyle, A. B., Beaudet, J. and Aboud, E. E. (1988), 'Developmental patterns in the flexibility of children's ethnic attitudes', Journal of Cross-Cultural Psychology, 19, 3-18, doi: 10.1177/0022002188019001001.

Edina (2011), UKBorders. Available at: http://edina.ac.uk/ukborders/. Last accessed July 2011.

Ekehammar, B., Akrami, N. and Araya, T. (2003), 'Gender differences in implicit prejudice', Personality and Individual Differences, 34 (8), 1509-23.

Ekehammar, B. and Sidanius, J. (1982), 'Sex differences in sociopolitical attitudes: a replication and extension', British Journal of Social Psychology, 21, 249-57.

Fetzer, J. S. and Soper, J. C. (2003), 'The roots of public attitudes toward state accommodation of European Muslims' religious practices before and after September 11', Journal for the Scientific Study of Religion, 42, 247-58.

Francis, L. J. (1997), 'The psychology of gender differences in religion: a review of empirical research', Religion, 27 (1), 81-96, doi: 10.1006/reli.1996.0066.

Gorsuch, R. L. and Aleshire, D. (1974), 'Christian faith and ethnic prejudice: a review and interpretation of research', Journal for the Scientific Study of Religion, 13 (3), 281-307.

Grace, G. (2003), 'Educational studies and faith-based schooling: moving from prejudice to evidence-based argument', British Journal of Educational Studies, 51 (2), 149-67.

Hamberger, J. and Hewstone, M. (1997), 'Inter-ethnic contact as a predictor of blatant and subtle prejudice: test of a model in four West European countries', British Journal of Social Psychology, 36, 173-90.

Hoover, R. and Fishbein, H. D. (1999), 'The development of prejudice and sex role stereotyping in White adolescents and White young adults', Journal of Applied Developmental Psychology, 20 (3), 431-48.

Hox, J. (2002), Multilevel Analysis: Techniques and Applications, London: Lawrence Erlbaum Associates, Publishers.

Hoxter, A. L. and Lester, D. (1994), 'Gender differences in prejudice', Perceptual and Motor Skills, 79, 1666.

Jackman, M. R. and Crane, M. (1986), '"Some of my best friends are Black . . . ": interracial friendship and Whites' racial attitudes', Public Opinion Quarterly, 50, 459-86.

Kirkpatrick, L. A. (1993), 'Fundamentalism, Christian orthodoxy, and intrinsic religious orientation as predictors of discriminatory attitudes', Journal for the Scientific Study of Religion, 32 (3), 256-68.

Laurence, J. (2011), 'The effect of ethnic diversity and community disadvantage on social cohesion: a multi-level analysis of social capital and interethnic relations in UK communities', European Sociological Review, 27 (1), 70-89, doi: 10.1093/esr/jcp057.

Laythe, B., Finkel, D. and Kirkpatrick, L. A. (2001), 'Predicting prejudice from religious fundamentalism and right-wing authoritarianism: a multiple-regression approach', Journal for the Scientific Study of Religion, 40 (1), 1-10.

Lee, S. A., Gibbons, J. A., Thompson, J. M. and Timani, H. S. (2009), 'The Islamophobia scale: instrument development and initial validation', The International Journal for the Psychology of Religion, 19, 92-105.

#Levin, S., van Laar, C. and Sidanius, J. (2003), 'The effects of ingroup and outgroup friendships on ethnic attitudes in college: a longitudinal study', Group Processes and Intergroup Relations, 6, 76-92.

McGlothlin, H. and Killen, M. (2006), 'Intergroup attitudes of European-American children attending ethnically homogeneous schools', Child Development, 77, 1375-86.

McGlothlin, H., Killen, M. and Edmonds, C. (2005), 'European-American children's intergroup attitudes about peer relationships', British Journal of Developmental Psychology, 23, 227-49.

Moore, J. W, Hauck, W E. and Denne, T. C. (1984), 'Racial prejudice, interracial contact, and personality variables', Journal of Experimental Education, 52, 168-73.

Oliver, J. E. and Wong, J. (2003), 'Intergroup prejudice in multiethnic settings', American Journal of Political Science, 47, 567-82.

ONS (Office for National Statistics) (2011), Neighbourhood Statistics. Available at: www.neighbourhood.statistics.gov.uk/dissemination/. Last accessed July 2011.

Paolini, S., Hewstone, M., Cairns, E. and Voci, A. (2004), 'Effects of direct and indirect cross-group friendships on judgments of Catholics and Protestants in Northern Ireland: the mediation role of an anxiety reduction mechanism', Personality and Social Psychology Bulletin, 30, 770-86.

Pettigrew, T. F. (1997), 'Generalised intergroup contact effects on prejudice', Personality and Social Psychology Bulletin, 23, 173-85.

--(1998), 'Intergroup contact: theory, research and new perspectives', Annual Review of Psychology, 49, 65-85.

Pettigrew, T. F., Christ, O., Wagner, U. and Stellmacher, J. (2007), 'Direct and indirect intergroup contact effects on prejudice: a normative interpretation', International Journal of Intercultural Relations, 31, 411-25.

Pettigrew, T. F. and Tropp, L. (2006), 'A meta-analytic test of intergroup contact theory', Journal of Personality and Social Psychology, 90, 751-83.

Powlishta, K. K., Serbin, L. A., Doyle, A. B. and White, D. R. (1994), 'Gender, ethnic, and body type biases: the generality of prejudice in childhood', Developmental Psychology, 30, 526-36.

Putnam, R. D. (2007), 'E Pluribus Unum: diversity and community in the twenty-first century. The 2006 Johan Skytte Prize Lecture', Scandinavian Political Studies, 30, 137-74.

Qualls, R. C., Cox, M. B. and Schehr, T. L. (1992), 'Racial attitudes on campus: are there gender differences?', Journal of College Student Development, 33, 524-9.

Robinson, J., Witenberg, R. and Sanson, A. (2001), 'The socialisation of tolerance', in M. Augoustinos (ed.), Understanding Prejudice, Racism, and Social Conflict, London: Sage, pp. 73-88.

Rutland, A. (1999), 'The development of national prejudice, in-group favouritism and self stereotypes in British children', British Journal of Social Psychology, 38, 55-70.

Saroglou V., Lamkaddem, B., Van Pachterbeke, M. and Buxant, C. (2009), 'Host society's dislike of the Islamic veil: the role of subtle prejudice, values, and religion', International Journal of Intercultural Relations, 33 (5), 419-28.

Scheepers, P., Gijsberts, M. and Hello, E., (2002), 'Religiosity and prejudice against ethnic minorities in Europe: cross-national tests on a controversial relationship', Review of Religious Research, 43, 242-65.

Semyonov, M. and Glikman, A. (2009), 'Ethnic residential segregation, social contacts, and anti-minority attitudes in European societies', European Sociological Review, 25 (6), 693-708.

Snijders, T. A. B. and Bosker, R. J. (1999), Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling, London: Sage.

Tolsma, J., van der Meer, T. and Gesthuizen, M. (2009), 'The impact of neighbourhood and municipality characteristics on social cohesion in the Netherlands', Acta Politica, 44 (3), 286-313.

Walter, T. and Davie, G. (1998), 'The religiosity of women in the modern west', British Journal of Sociology, 49 (4), 640-60, doi: 10.2307/591293.

White, F. A., Wootton, B., Man, J., Diaz, H., Rasiah, J., Swift, E. and Wilkinson, A. (2009), 'Adolescent racial prejudice development: the role of friendship quality and interracial contact', International Journal of Intercultural Relations, 33, 524-34.

Wylie, L. and Forest, J. (1992), 'Religious fundamentalism, right-wing authoritarianism and prejudice', Psychological Reports, 71, 1291-8.

http://dx.doi.org/10.7227/RIE.88.1.2

Adrian Brockett and Kate Wicker York St John University

Address for correspondence

Adrian Brockett, Faculty of Education and Theology, York St John University, Lord Mayor's Walk, York, YO31 7EX, United Kingdom. E-mail a.brockett@yorksj.ac.uk
Table 1 Data for the twenty schools in the sample

                                       Mean     Range

School roll                            897.6   114-1733
% Ethnically White (school roll)       75.8      0-98
% Eligible for free school meals       12.4      0-39
% With 5+ GCSE passes                  62.3     36-98
School neighbourhood:
Index of Multiple Deprivation (IMD)    20.4      8-60
% Ethnically White (school
  neighbourhood)                       91.1     21-100

Table 2 Ethnic and religious 2001 Census Data for the 19,975
Lower Super Output Areas within the catchment areas of all the
schools in the study

(a) Mean percentages

Percentage of      Mean    SD    Minimum   Maximum
area population:

White              94.1   13.3     0.5      100.0
Asian               4.0   12.0     0.0       97.9
Other ethnicity     1.9    3.4     0.0       82.3
Christian          74.4   13.6     0.0       97.7
Muslim              3.4   11.1     0.0       93.8
Other religion      0.9    2.4     0.0       72.6
No religion        13.3    6.4     0.0       50.5

(b) Correlation matrix

                   Asian       Other     Christian   Muslim
                             ethnicity

White              -0.97       -0.47        0.83     -0.95
Asian                           0.24       -0.80      0.98
Other ethnicity                            -0.42      0.25
Christian                                            -0.79
Muslim
Other religion

                   Other        No
                  religion   religion

White              -0.21        0.09
Asian               0.16       -0.17
Other ethnicity     0.22        0.25
Christian          -0.30       -0.38
Muslim              0.08       -0.17
Other religion                  0.07

Table 3 Summary of individual-level variables

Nominal                 Value       Coded   Number       %
variables

Sex                     Male          0      1076      43.8
                        Female        1      1380      56.2
Age group               10-13         1      1195      48.7
                        14-16         2      1026      41.8
                        17-19         3       235       9.6
Friends of a            Some          0      1419      57.8
different race          None          1      1037      42.2

Religion                Christian     1      1216      49.5
                        Muslim        2       459      18.7
                        None          3       781      31.8

Continuous variables    Mean         SD     Minimum   Maximum

Theistic Belief Scale   21.3         8.8       7        35.0
Neighbourhood IMD       22.0        16.3      1.6       75.3
Neighbourhood %White    87.1        22.5      4.6      100.0

Table 4 Multilevel linear model fitted to all data for individual
variables

Variable            Value         Null         Individual

Intercept                        15.375 ***      14.590 ***
Sex (Female)      Male                            1.553 ***
Age (17-19)       10-13                           0.800 *
                  14-16                           1.027 **
Friends of        Some                           -1.259 **
  different
  race (none)
Religion (No      Christian                       0.739 **
  religion)       Muslim                          0.411
TBS                                              -0.141 ***
-2loglikelihood               14028.564       13711.997
Deviance                                        316.567 ***
Parameters                        3              10
Residual                         17.406 ***      15.274 ***
Intercept                         1.154 **        1.295 **
ICC                               6.2%            7.8%

Note For nominal variables, reference value is given in parentheses.
ICC = Intraclass

Correlation Coefficient. * p < 0.05; ** p < 0.01; *** p < 0.001.

Table 5 Multilevel linear models fitted to data for White pupils
only

Variable                     Value          Null         Individual

Intercept                                  15.676 ***      14.301 ***
Individual:
Sex (Female)              Male                              1.573 ***
Age (17-19)               10-13                             0.708
                          14-16                             1.151 **
Friends of                Some                             -1.129 ***
  different race (none)
Religion (No religion)    Christian                         0.815 ***
TBS                                                        -0.146 ***
Neighbourhood:
IMD
%White
School:
Type (State)              Church
                          Independent
Number on roll
%White pupils on roll
% Eligible FSM
% 5+ GCSEs
School neighbourhood:
  IMD
  %White
-2loglikelihood                         11395.7         11146.218
Deviance                                                  249.457 ***
Parameters                                  3               9
Residual                                   17.239 ***      15.202 ***
Intercept                                   1.698 **        1.6622 **
ICC                                         9.0%            9.9%

Variable                  Neighbourhood      School

Intercept                    14.292 ***      13.902 ***
Individual:
Sex (Female)                  1.593 ***       1.592 ***
Age (17-19)                   0.681           0.609
                              1.118 **        1.160 **
Friends of                   -1.120 ***      -1.087 ***
  different race (none)
Religion (No religion)        0.842 ***       0.860 ***
TBS                          -0.145 ***      -0.147 ***
Neighbourhood:
IMD                           0.025 **        0.025 **
%White                        0.024 *         0.021
School:
Type (State)                                  0.843 *
                                              1.017
Number on roll                                0.000
%White pupils on roll                         0.009
% Eligible FSM                               -0.107 *
% 5+ GCSEs                                   -0.105 ***
School neighbourhood:
  IMD                                         0.018
  %White                                      0.005
-2loglikelihood           11135.394       11104.295
Deviance                     10.824 **       31.099 ***
Parameters                   11              19
Residual                     15.134 ***      15.166 ***
Intercept                     1.468 *         0.070
ICC                           8.8%            0.5%

Table 6 General linear models fitted to data for Muslim pupils
only

Variable         Value    Null        Individual   Neighbourhood

Intercept                 14.41 ***   14.626 ***   14.464 ***
Individual:
Sex (Female)     Male                  1.51 ***     1.47 ***
Age (17-19)      10-13                 0.356        0.35
                 14-16                -0.239       -0.26
Friends of       Some                 -2.099 ***   -2.031 ***
  different
  race (none)
TBS                                    0.036        0.026
Neighbourhood:
                 IMD                               -0.017
                 %White                            -0.015 *
COPYRIGHT 2012 Sage Publications, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Brockett, Adrian; Wicker, Kate
Publication:Research in Education
Article Type:Report
Geographic Code:4EUUE
Date:Nov 1, 2012
Words:7018
Previous Article:An evaluation of the science education component of the Cross River State Science and Technical Education Project.
Next Article:Sources of stress among undergraduate students in the University of Benin, Benin City, Nigeria: implications for counselling.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters