Abstract
This study explores the beliefs and attitudes that university students enrolled in teacher education programmes in Spain, England and the US (Texas) hold about individuals who differ. A beliefs and attitudes toward difference scale (BATD) was constructed using nine dimensions of diversity; culture, language, socioeconomic status/social class, religion, gender, sexual orientation, political ideology, disability and special talent. A two‐way factorial analysis of variance indicated significant main effects due to the respondent groups on culture, religion and sexual orientation; significant main effects of worldview of difference on political ideology; and no interaction between the two factors on each of the nine domains of difference. An exploratory factor analysis was also performed in order to explore the viability of the theoretical model. The data suggest that attitudes towards people who differ include etic, emic and individual properties. These are discussed in terms of the uses of cross‐cultural data and further research opportunities.
| Original language | English |
|---|---|
| Pages (from-to) | 245-264 |
| Number of pages | 20 |
| Journal | European Journal of Teacher Education |
| Volume | 33 |
| Issue number | 3 |
| DOIs | |
| Publication status | Published - 2010 |
Bibliographical note
Taylor & Francis Online
Top
Full Article
Figures & data
References
Citations
Metrics
Reprints & Permissions
View PDF(open in a new window)
Share
Abstract
This study explores the beliefs and attitudes that university students enrolled in teacher education programmes in Spain, England and the US (Texas) hold about individuals who differ. A beliefs and attitudes toward difference scale (BATD) was constructed using nine dimensions of diversity; culture, language, socioeconomic status/social class, religion, gender, sexual orientation, political ideology, disability and special talent. A two‐way factorial analysis of variance indicated significant main effects due to the respondent groups on culture, religion and sexual orientation; significant main effects of worldview of difference on political ideology; and no interaction between the two factors on each of the nine domains of difference. An exploratory factor analysis was also performed in order to explore the viability of the theoretical model. The data suggest that attitudes towards people who differ include etic, emic and individual properties. These are discussed in terms of the uses of cross‐cultural data and further research opportunities.
Keywords:
cultural differencespre‐service teacher educationattitude change
Previous article
View issue table of contents
Next article
Introduction
Recent years have seen many changes in attitudes towards human diversity and difference across individual, institutional and policy levels. Educational and other social institutions have become progressively more diverse resulting from a range of cultural, legal, economic and societal factors that are linked to rapidly changing demographics. In many countries, increased migration and greater inclusion in education and other public services have been associated with changing attitudes to various aspects of human difference such as disability, sexual orientation, language and religious beliefs. As the forces of globalisation and the associated movement of people have changed the demographic of schooling, those who prepare teachers have become more aware of the challenges diversity issues pose for classroom teachers. At the same time, there is also an awareness that those who are preparing to become teachers bring their own attitudes and beliefs about people who are perceived to be different into the profession, and these influence how they respond to student differences. It is important, therefore, to explore the attitudes and beliefs of pre‐service teachers as an essential strand of research in teacher education (Hollins and Guzman Citation2005).
Conceptualising individual differences and diversity
As a social construct, the concept of difference is susceptible to multiple interpretations and it is important to define what is meant when referring to human diversity. Turner (Citation1990) suggests that we take into account that all people possess a group of common characteristics that are shared with other humans (universal characteristics), other characteristics that are shared only with some groups of humans (cultural characteristics) and a final set of characteristics that are unique (personal characteristics). These three aspects of the individual are exhibited in close relationship to each other at the same time, and as facets of the concept of difference, they imply that although personal beliefs are subjective, they can be interpreted at three distinct levels: macro (universal), meso (group) and the micro (individual). According to Henning‐Stout and Brown‐Cheatham (Citation1999) and Turner (Citation1990), this manner of conceptualising difference forces us to examine three phenomena: (a) the role that culture plays in one’s own development and expression of beliefs and attitudes; (b) the need to identify the dimensions of difference most relevant for education; and (c) whether or not attitudes toward difference have similar or dissimilar patterns, across countries with distinct cultures. Trans‐cultural studies on beliefs and attitudes about difference can provide insights into whether attitudes towards people who differ from the majority culture are held universally or not.
Although the concept of difference is difficult to define without recourse to theories of deviance (Gallagher Citation2004), there are alternative theories (Fine and Wong Citation1995; Greenberg et al. Citation1992; Solomon, Greenberg, and Pyszczynski Citation1991) that conceptualise difference as a dynamic construct, which is socially constructed. In other words, difference is a term that is neutral but whose meaning is established between those who define and those who are defined (Aguado Citation2004). Here differences are seen as differences in: (a) the universal symbols shared by a given group of people; (b) the significance that members of the group place in particular events and phenomena; and (c) the interpretation that these group members have about other groups and their symbols, events and phenomena. Every group can be seen as sharing significance with a determined group, which creates community and the possibility of adaptation. Contact and exchange are always possible and the aspiration to exist without stigmatising others, but to respect the identity and individualism of each person, is one of the characteristics of a postmodern society (Lubienski Citation2003; Trend Citation1996).
Diversity is a current reality
Today many societies are described as pluralistic (Henning‐Stout and Brown‐Cheatham Citation1999). In the last several years, Spain has changed from seeing its citizens emigrate to seeing an increase in immigrants (Aja Citation2000; De Lucas Citation2006). In fact, in just 10 years the population of foreign residents in Spain has quadrupled from 2% in 1995 to 8% of the population in 2005 (National Institute of Statistics 2006). Currently, the population of students enrolled in Spanish primary and secondary schools is: 41% from Europe, 19% from Africa, 35% from South America and 5% from Asia and Asia Pacific. A similar trend has been observed in the UK. Currently in England, 23.3% of all primary school pupils and 20.6% of all secondary school pupils are of black or minority ethnic (BME) heritage. In the US one of three is a person of colour, Hispanic, or Asian American (Lindsey and Beach Citation2004). Since 1965, immigration from Latin America and Asia to the US has been steadily increasing and within two decades it is predicted that only a minority of the US population will be of European extraction.
Universities also have experienced changes in the composition of their student populations as the number of women, students with disabilities, and those from different social classes, ethnic, linguistic and cultural backgrounds enrol in higher education. While there has been a steady increase in the enrolment of minority groups, the relative enrolment of these groups in continues to be low in many subjects and becomes more marked as the level of education increases. In Britain, for example, only 5.9% of teachers are of BME heritage even though ethnic minorities are now better represented in higher education than they are in the general population (Akhtar Citation2010).
But more important than the representation of minorities in higher education is the conflict that negative views about diversity can generate. For example, research on teachers and teacher education suggests that novice teachers hold pre‐existing beliefs regarding diversity that are difficult to change. Many novice teachers ‘enter teacher preparation programmes with negative or deficit attitudes and beliefs about those different from themselves’ (Hollins and Guzman Citation2005, 511). For this reason, many university programmes require students to enrol in coursework or other programmes that focus on the understanding, tolerance or acceptance of differences in others. In fact, almost all institutions of higher education are making considerable efforts to internationalise the curriculum and prepare students to participate in a pluralistic and diverse world that is characterised by heterogeneity and diversity (Henning‐Stout and Brown‐Cheatham Citation1999; Wergin Citation1989).
Cochran‐Smith (Citation2005) argues that social, intellectual, and organisational contexts shape the learning of teachers enrolled in preparation programmes and therefore attitudes regarding diversity can be addressed through the reconceptualisation of teacher education programmes. However, despite several decades of multicultural education and inclusionary coursework, findings on the effectiveness of programmes that prepare teachers for diversity have been mixed (Cochran‐Smith, Davis, and Fries Citation2003; Ladson‐Billings Citation1999). The AERA Panel on Research and Teacher Education specifically reviewed research that identifies the particular attributes, prior knowledge, and background experiences that affect how those preparing to be teachers perceive diversity. However, the majority of research in this area has been either small‐sample qualitative studies or based on surveys in which the instrument has not been validated.
This study was designed to explore the beliefs and attitudes that university students preparing to be teachers hold about individuals who differ. It considers how notions of human diversity and difference are understood by students in three countries. Nine dimensions of diversity thought to have significant implications for education (culture/ethnic origin, language, socioeconomic status/social class, religion, gender, sexual orientation, political ideology, disability and giftedness/special talents) were identified from a review of the literature (Banks and Banks Citation2003; Henning‐Stout and Brown‐Cheatham Citation1999; Jiménez‐Fernández Citation2004). These dimensions became the variables of interest in this study. Three groups of students enrolled in teacher education programmes in three different countries (Spain, England, and the US) responded to a survey designed and piloted by the authors (see Cardona et al. Citation2008; Cardona, Jiménez‐Fernández, and Chiner Citation2006). The objectives that guided the study were:
To identify common or dissimilar patterns of students’ feelings of difference across three cultures (countries).
To analyse students’ attitudes toward difference in terms of culture, language, socioeconomic status/social class, religion, gender, sexual orientation, political ideology, disability and giftedness/special talent, and determine whether or not these attitudes differed between the three groups (Spanish, English, American) and worldview of diversity (ethnocentric versus relativistic view of difference).
To explore possible differences in opinion among respondents regarding equal treatment (e.g., equal opportunities as apply to people who differ), as well as actions to be taken in schools and universities to achieve a more inclusive society.
The study set out to explore the following hypotheses: (a) the universal properties of the pattern of attitudes would be demonstrated if similar and common patterns in the beliefs and attitudes of the respondents about people who differ were found; (b) that group properties would be demonstrated if statistically significant differences in views about difference were found between the respondents from each country; and (c) that individual properties would be demonstrated if we observed differences in the personal points of view (more or less relativistic/ethnocentric) about difference. All of these depended on the identification of a similar factorial cross‐cultural structure in the instrument that was used.
Method
Sample
The sample consisted of teacher education cohorts enrolled at the University of Alicante, Spain, the University of Cambridge, UK, and Texas A&M University, USA. The respondent group, composed of a convenience sample of 310 students (38.70% from the University of Alicante, Spain, 32.26% from the University of Cambridge, UK, and 29.03% from Texas A&M University, USA), represented approximately 10% of the student body at the faculty/college of education where the study took place. The majority of the sample (85%) was female (75% of the Spanish, 89% of the English and 92% of the American sample, respectively). With respect to ethnicity, the respondents comprised Spanish students (100% ethnic Spaniards), English students (93% ethnic British and 7% other), and American students (70% White or European American, 19% Hispanic, 4% Asian American, 2% African American, and 4% other). The average age of the students was 26 years with a mean teaching experience of 2.31 years (SD = 4.38) (5.68 for the Spanish students, 1.49 for English, and .35 for American students). The majority (88%) were pursuing teacher certification at the kindergarten/elementary (primary) level or in the area of special education. Fifty‐two per cent of the Spanish respondents were current teachers (17% kindergarten teachers, 28% elementary, and 7% secondary/other) while only a minority (16% and 2%) reported some kind of teaching experience in the English and American samples, as in these cases the students were on courses leading to qualified teacher status.
Respondents were asked to describe themselves in terms of their culture, language, socioeconomic status/social class, religion, gender, sexual orientation, political ideology, disability status and whether they were perceived as being gifted or having special talents. Except for language, where Spanish students were more likely to describe themselves as bilingual, all three groups identified themselves with the dominate culture at each of the three universities, namely, as being predominately middle‐class, heterosexual, female with neither disabilities nor special talents. Political ideology was mixed for all three groups. The Spanish students described themselves predominately as Catholic. All of them were studying subject areas regarding to diversity, individual differences and inclusion.
Settings and participants
The three institutions of higher education from which the participants were recruited were all large publicly funded universities that have strong teacher education and graduate education programmes.
The University of Alicante (UA) is a young institution founded in 1979. At present it has over 30,000 students. The UA was formed on the basis of the Centre for University Studies (CEU) that regained university status which had disappeared in 1834 when its predecessor, the University of Orihuela, closed after two centuries of education (1610–1808). University studies were reinitiated in the academic year 1968–1969 with just 230 students. Now the UA offers about 50 different degree programmes through 60 departments and seven faculties/colleges. The Faculty of Education is one of UA’s oldest centres and serves annually approximately 3000 students of whom approximately 1000 are graduate students (more than 70% female). Attention to the needs of students with disabilities or other individual differences is an important part of its undergraduate, masters‐level programmes, as well as doctoral degrees in education. The educational system in Spain experienced major reconstruction in the late 1970s that affected Spanish society at all institutional levels. After the death of General Francisco Franco important political and socio‐economic changes, alongside new educational policies, transformed a society until then dictatorial into a democratic one. Educational institutions in Spain are now looking at ways to support pluralism, diversity and enhance inclusive practices at all levels. In this context, the UA Faculty of Education prepares teachers for one of two specialties (kindergarten and elementary), as well as other educational professions at masters level (e.g., secondary education, educational psychology). All of these programmes include compulsory courses on special education, but most struggle to address issues of diversity and inclusive education within the time available. Study participants from the UA were enrolled in a two‐year masters degree in educational psychology and were in their second year of study. The participants represented each of the major areas of the UA current teacher education programmes (kindergarten, elementary, secondary, and special education) and at the time of the study, all had been taught the core coursework and had completed one semester of practicum mainly in elementary or secondary public schools.
The University of Cambridge is one of the world’s oldest and leading academic centres, celebrating its 800th anniversary in 2009, and is consistently ranked in the world’s top five universities. The university consists of 31 self‐governing independent colleges, which bring together academics and students from a broad range of disciplines. In addition, over 150 departments, faculties and other institutions are responsible for lectures, seminars, research and the syllabi for teaching, overseen by a General Board and university administration. There are currently over 20,000 students enrolled at the university. Teaching and research in Cambridge is organised by faculties. The Faculty of Education is part of the School of Humanities and Social Sciences. The recently integrated faculty is also one of the oldest and one of the most prestigious education departments in the UK. It offers an undergraduate BA in education and a wide range of postgraduate courses, from the Postgraduate Certificate in Education (PGCE) to PhDs. Participants in this study from Cambridge were enrolled in the PGCE training to teach in early years and primary schools. The one‐year course, leading to qualified teacher status, is taught in partnership with local schools with trainees spending at least 50% of their course working with teachers and children in settings across the local region. Students on the PGCE are all graduates with a bachelors degree, or higher. Some students enrol on completion of their first degree, whilst others are mature having decided to become teachers after a period in another occupation or career break. The participants in this study were all training to be early years and primary teachers (ages 3–11). They had completed two school placements in different classes within one school, as well as two terms of university coursework. The survey was administered at the end of the second term.
Texas A&M University is one of two research‐intensive public universities in the state of Texas. Texas A&M was established as Texas’s first public institution of higher learning in 1876 as a male military institution with all students belonging to the university’s Corps of Cadets. Membership in the Corps continued to be mandatory until the early 1960s and the university became co‐educational soon after. The university expanded its focus on agriculture and engineering during the middle of the twentieth century to include the colleges of Liberal Arts, Architecture, Business, and Education, among others to become one of the largest comprehensive public universities in the US. Texas A&M currently enrols over 46,000 students at its 5000‐acre campus and has an additional branch, Galveston campus, as well as one in Qatar. The university offers more than 120 undergraduate degree programmes as well as over 240 graduate degrees and many of these programmes are ranked in the top 10 nationally in the US. Despite its growth and changes, Texas A&M has held onto many traditions, and facilitates the acculturalisation of new students as ‘Aggies’ through their involvement in student organisations, sports activities, and, after graduation, alumni activities.
The College of Education at Texas A&M University was established in 1969 and consists of four departments. Teacher certification primarily takes place at the undergraduate level and the certification programme is the largest supplier of educators in the state. Participants in this study from Texas A&M University were enrolled in a pre‐baccalaureate special education programme, two years in length, which included multiple extensive field‐based components. At the time of the survey, the Texas A&M participants had completed two semesters of field‐based teaching in which they were placed in public school classrooms with students with disabilities for two full days each week. As part of their remaining coursework, they had completed three separate 45‐hour practicum in public schools that were selected for their ethnic and linguistic diversity.
Survey instrument
The BATD (Beliefs and Attitudes Toward Difference Scale) developed by the authors as part of a larger cross‐cultural study of attitudes toward people who differ, was designed to assess varying levels of acceptance (or openness to) a range of diversity issues. The first part (self‐description and feeling of difference from others) consisted of two items (Item 1 and 2) composed of nine five‐point rating scale sub items (‘1’ = little, ‘5’ = very much) that elicited students’ self‐description and feeling of difference from others in terms of: (a) culture/ethnic background; (b) language; (c) socioeconomic status/social class; (d) religion; (e) gender; (f) sexual orientation; (g) political ideology; (h) disability; and (i) giftedness/special talent. One example of these items was: ‘To what extent do you feel different from others in your current university environment in terms of culture?’ Low scores (close to ‘1’) reflected perceptions of strong identity in regarding each dimension of diversity, whereas high scores (close to ‘5’) reflected strong perceptions or feelings of difference. The second part (beliefs and attitudes toward people who differ) included six questions to measure: (a) the participants’ degree of tolerance towards people who differ (Item 3); (b) relativism versus ethnocentrism (Item 4); and (c) beliefs about opportunities available to people who differ (Item 5). The last three items were open‐ended questions designed to identify the actions that, in the participants’ opinions, should be taken in schools and universities to achieve a more inclusive society within their respective countries, as well as other questions not reported here. For Item 3 (attitudes toward people who differ) adjectives were assigned to a numerical Likert scale to label the degree of tolerance (rejection = ‘1’, avoidance = ‘2’, indifference = ‘3’, tolerance = ‘4’, acceptance = ‘5’). In this item, composed of nine sub items, participants responded to the question ‘Which of the following terms (rejection, avoidance, indifference, tolerance, acceptance) best describe how you yourself respond to differences in culture/ethnic background, language… others?’ Low scores (close to ‘1’) reflected rejection or general intolerance for diversity, whereas high scores (around ‘5’) openness or acceptance of the diversity issues. Midrange scores reflected a degree of indifference for (or uncertainty toward) the issues included in the measure. A binomial scale (yes = ‘1’, no = ‘2’) was used to record the participant responses in Item 4 (‘Do you agree to the statement that there are some cultures, languages, etc… superior to others?’), and Item 5 (‘Do you believe that people who differ have equal opportunities to obtain good jobs?’). Because our initial focus was on individual issues of diversity, no composite scores (sum of scores by item) were computed, except for relativism versus ethnocentrism. The first page of the survey included a brief overview of the study and its purpose, as well as some demographic data (gender, ethnicity, age, major, year of study, number of years of teaching experience, if applied). The rest of the items were listed on subsequent pages. The five‐point Likert scale with its associated rating description was reproduced on the top of each page to increase understanding of the rating scale to be used. After the translation process, the instrument was subjected to a preliminary review by six professors with experience in the field of multicultural and special education (two in each participant country). It was also reviewed by graduated students (n = 3) with teaching experience in the three countries. The professors were asked to evaluate the instrument to determine if items: (a) fell within the designated domains outlined on the instrument; (b) were clear and unambiguous; and (c) were comprehensive in measuring beliefs and attitudes about a range of diversity issues. The graduate students were asked to complete the instrument and give: (a) feedback regarding the clarity of individual items and administrative directions; and (b) recommendations for the improvement of items. Data from this preliminary review led to some minor changes in the wording of items prior to the formal pilot testing (Cardona et al. Citation2006, Citation2008) that led to a reduced pool of items – eight instead of the original pool of 14. Two types of reliability were calculated with this sample: Cronbach’s internal consistency (alpha) and split‐half. Cronbach alpha coefficients for the feelings of difference set of questions ranged from .79 to .80 and averaged .79 (SD = .00); the split‐half coefficient ranged from .71 to .76 and averaged .74 (SD = .02). For attitudes toward people who differ, alpha coefficients ranged from .88 to .94 and averaged .91 (SD = .03), while split‐half coefficients ranged from .87 to .95 and averaged .92 (SD = .04). For beliefs on equal opportunities, Cronbach alpha coefficients ranged from .82 to .94 and averaged .88 (SD = .07); split‐half coefficients ranged from .80 to .88 and averaged .86 (SD = .04). Finally, for relativism versus ethnocentrism, coefficients of internal consistency ranged from .63 to .70 and averaged .67 (SD = .04); split‐half coefficients ranged from .63 to .72 and averaged .70 (SD = .05). These reliability coefficients indicated overall good reliability for the survey instrument.
Procedure
Cross‐cultural research involves at least four procedural challenges; translation, participant recruitment, data collection, and data analysis (Harkness, Van de Vijver, and Mohler Citation2003). We attempted to overcome these challenges through the following steps.
Translation
The survey instrument was developed initially in English and then translated into Spanish. Two types of equivalence were the focus in the instrument’s translation: conceptual (or content) equivalence through the use of a panel rating process (Shin Citation2002) and linguistic (or semantic) equivalence using back‐translation techniques (Brislin Citation1986; Shultz and Whitney Citation2005). A three‐step translation was used. First, two bilingual professionals, who were also professors of education translated the English version into Spanish. Second, two other bilingual professionals, who were also teachers of English, translated the Spanish version back into English. Third, the total group of four checked the translation and rectified any discrepancies found. Country‐specific peculiarities in language usage were considered during the entire translation process.
Participant recruitment
Each of the authors contacted colleagues at their respective universities who taught teachers or teachers‐in‐training and asked if they would be willing to distribute or permit one of the researchers to administer the survey at the end of the class period. All instructors agreed. The universities’ schedule of classes was used to identify the complete cohort of students.
Data collection
All participants were administered a questionnaire during class sessions. In each of the classrooms where data were collected, the study was described along with its purpose. Instructions were given to the participants regarding completion of the survey instrument and estimated time for completion. There was no time limit imposed for completing the survey, but actual administration took no more than 15 minutes. Completed surveys were then sent to the first author for compilation and analysis. Upon receipt, each survey was verified for completeness.
Data analysis
Cross‐cultural research poses two important issues regarding data analysis. The first relates to testing the construct to ensure comparability across cultures/language groups (e.g., its etic properties). The second issue relates to testing for a social, culture‐specific component that reflects cultural/language group differences (e.g., its emic properties). We used three analyses to evaluate the construct’s etic (universal) and emic (cultural) properties: (a) response profile analysis; (b) analysis of variance of the mean in attitudes toward difference; and (c) exploratory factor analysis.
The SPSS statistical package for social sciences (SPSS Citation1998) was used for all quantitative data entry and analyses, while the analysis of qualitative data (open‐ended questions) involved the use of a method of thematic analysis (e.g., method of constant comparison; Lincoln and Guba Citation1985): (a) all the students’ words, phrases, and sentences to the open‐ended questions were read; (b) these students’ responses were then unitised; (c) these units of information were used as a basis for extracting a list of non‐repetitive, non‐overlapping significant statements; and finally (d) clusters of themes were organised with each cluster consisting of units that were deemed similar in content; therefore, each cluster representing a unique emergent theme.
Results
Response profile analysis
The average scores on feeling of difference across cultural groups showed a general skewedness with mean scores lower than 2 on the Likert scale in almost all dimensions of diversity (see Table 1). These results provide evidence of identity of the respondents within their respective cultural groups. Spanish participants showed a stronger identity in culture/ethnic background than the English and American participants (p <.05), whereas participants’ feeling of difference because of language was stronger for the English and Americans than for the Spanish participants (p <.05). Identity in socioeconomic status was also stronger for the Spanish than for the American respondents (p <.05).
Table 1. Participants’ feelings of difference across cultural groups.
Download CSVDisplay Table
When computing responses across the three cultural groups in attitudes toward people who differ, we obtained a significant association between cultural group and rating values. As shown in Table 2, the response distribution across cultural groups showed difference in attitudes based on the percentages of respondents indicating scores of ‘1’ (rejection), ‘2’ (avoidance), ‘3’ (indifference), ‘4’ (tolerance), and ‘5’ (acceptance). A significant association between cultural groups and ratings values in culture, χ2 (6) = 13.89, p <.05; religion, χ2 (8) = 18.41, p <.05; and sexual orientation, χ2 (8) = 42.67, p <.01 was found. These findings, based on calculating the Pearson chi‐square and the likelihood ratio chi‐square, suggests that the respondents varied in their attitudes toward people who differed from them in terms of culture, religion, and sexual orientation, but these attitudes did not vary with regards to perceived differences in terms of language, socio‐economic status/social class, gender, political ideology, disability, or giftedness/special talents.
Table 2. Attitudes toward people who differ across cultural groups.
Download CSVDisplay Table
Differences in attitudes across cultural groups and view of difference
Mean scores and SD across the attitudes ratings for the nine core dimensions of diversity were calculated as a function of cultural group (Spanish, English, and American) and worldview of difference (relativistic versus ethnocentric) (Table 3). A two‐way 3 × 2 factorial analysis of variance on these data indicated: (a) significant main effects due to the respondent groups on culture, religion, and sexual orientation; (b) significant main effects of worldview of difference on political ideology; and (c) no interaction between the two factors on each of the nine domains of difference. Regarding cultural groups, post hoc tests showed that Spanish and American student scores were significantly lower (less positive) than the English student attitudes with regard to difference in culture, F(2,263) = 3.14, p <.05, and religion, F(2,262) = 5.03, p <.01, while the American student attitudes were less positive than the Spanish and English student attitudes regarding sexual orientation, F(2,262) = 15.51, p <.01. Concerning relativism versus ethnocentrism, post hoc tests showed that students with a relativistic point of view of difference, independent of their cultural group, showed a more positive attitude toward people who differed in political ideology, F(1,255) = 5.48, p <.05 than students with a more ethnocentric point of view.
Table 3. Mean and SD in attitudes by cultural groups and worldview of difference.
Download CSVDisplay Table
Differences in opinion regarding equal treatment
The proportion of respondents across cultural groups that believed that people who differ have equal opportunities as applied to obtain good jobs was not the same (see Table 4). In contrast to English and American respondents, Spanish participants strongly felt that persons who differ in culture (p <.01), religion (p <.01), gender (p <.01), sexual orientation (p <.01) and disability (p <.01) were more likely to have fewer opportunities to obtain good jobs than those from the mainstream.
Table 4. Differences in opinion among respondents regarding equal opportunities.1
Download CSVDisplay Table
Nevertheless, the three cultural groups show similarities in their opinions when asked about what they thought should be done to achieve more inclusive schools and universities. The common themes emerging from their answers are included in Table 5. All of them reflect high sensitiveness and very important issues to reflect on.
Table 5. Summary of emergent themes of whole group respondent opinions about the question of things to be done to achieve a more inclusive society.
Download CSVDisplay Table
Factor analysis
An exploratory factor analysis was performed to identify the factors that explain most of the variance observed. The nine sub‐items on attitudes (Item 3) and beliefs on equal opportunities (Item 5) were included in the analysis. We employed principal component analysis method of extraction, collapsing over the three groups. A prior selection of two factors to extract was established, in order to explore the viability of the theoretical model (identification of common/dissimilar patterns of beliefs and attitudes toward people who differ in the nine dimensions of diversity across the three groups). Communalities ranged from .49 (Item 3.7) to .71 (Items 3.8 and 3.9). Initial Eigen values ranged from 5.43 (Factor 1) to 4.25 (Factor 2), with a cumulative percentage of 53.74. The two‐factor solution grouped the 18 variables as follows: (a) attitudes toward people who differ (Items 3.1 to 3.9); and (b) beliefs on equal opportunities (Items 5.1 to 5.9) with Factor 1 contributing 28.99% and Factor 2, 24.79% of the variance, respectively. Loadings for each of the items in the factors were higher than .63 in every case.
In order to confirm the proposed factor structure, we conducted repeated factor analyses across the three university samples. If similar factor structures emerged, there would be a strong argument for the cross‐cultural universality of the nine core dimensions of diversity. First, analyses by groups were performed, with the Spanish sample being the first respondent group analysed. Principal component analysis with Varimax rotation showed a two factors solution that together account for the 49.80% of the variance, with Eigen values from 5.20 to 3.77. The factors reproduced the proposed model with Factor 1 and 2 contributing 28.86% and 20.94% of the variance, respectively. Loadings ranged from .55 to .83. The second analysis involved the ratings from the English sample. Initial Eigen values ranged from 6.49 (Factor 1) to 5.48 (Factor 2), with a cumulative percentage of 66.35. The rotated component matrix replicated the two factors solution from the total sample with Factor 1 contributing 34.95% of the variance and Factor 2, 31.41%. Items loadings ranged .60 to .94. Finally, the third analysis was conducted with the American sample. Again, the factor solution resulted in two factors that together explained 47.04% of total variance. Initial Eigen values ranged from 5.08 (Factor 1) to 3.39 (Factor 2). The rotated component matrix resulted in two factors whose items grouped according to the proposed model with Factor 1 explaining 26.60% of the variance and Factor 2, 20.47%, and with items loadings from .50 to .79. Given that the analyses revealed similar factor structures in the three samples, we conclude that there was cross‐cultural consistency in the underlying structure of the two‐factor solution of the scale – (the beliefs and attitudes in the nine dimensions of difference examined in our survey). In other words, data from the samples provided enough evidence supporting our two‐factor‐multidimensional model of diversity across the three cultural groups. As a test of factor independence, we also constructed a three‐ and a four‐factor solution, but the strength of the correlation between the two‐factor solution (p <.01) provided further evidence that the BATD is the common two‐factor solution operationalised by our Beliefs and Attitudes toward Difference Scale. In addition, the factor loadings again revealed that the items loading on Factor 1 and 2 reflected the beliefs and attitudes components of the scale with its nine dimensions.
Discussion
Support was obtained in the present study for our three hypotheses. However, interpreting the etic (universal), emic (cultural), and individual properties of the attitudes toward difference depends on the type of data and the corresponding analysis. Specifically, in the current study, response profiles based on mean scores were quite similar, supporting the etic and universal properties of the assessed dimensions of difference. However, significant respondent and group differences were obtained when analysing the percentage of responses. Thus, Hypothesis 1 was only partially supported. Regarding the analysis of mean scores (Hypothesis 2), significant respondent and cultural groups differences were obtained, reflecting the construct’s emic properties. In reference to the individual properties significant respondent differences were obtained in relativism versus ethnocentrism for all dimensions of diversity here considered, as well as for the identification of the most problematic dimensions and barriers to face, reflecting the construct’s personal and unique properties. Finally, in reference to the proposed factor structure, the clusters and factors on attitudes and equal opportunities clearly grouped into the two specified domains, again suggesting the construct’s etic and universal properties. These findings have at least two important implications for the cross‐cultural study of attitudes toward diversity: (a) understanding its etic, emic, and individual properties; and (b) the approach, purpose and use of cross‐cultural data.
Etic, emic, and individual properties
These data suggest that attitudes toward people who differ include etic, emic, and individual properties. Two results from the current study support its etic properties: (a) the similar response profiles across cultural groups; and (b) the two‐factor structure of the BATD scale. However, in reference to the construct’s emic properties, two results also suggest that one must be sensitive to the cultural properties of the construct: (a) the significant difference found across cultural groups in the attitudes response categories (rejection, avoidance, indifference, tolerance and acceptance); and (b) the significant difference in mean scores between the diverse domains of diversity (see Tables 2 and 3). Finally, with regard to the individual properties of the construct, these results also suggest that we have to pay close attention to personal views given the significant differences between individuals with regard to the relativistic versus ethnocentric view of difference.
Use of cross‐cultural data
The data from this study can be employed for a number of purposes at the micro (individual) and macro (cultural group) levels. For example, at the micro level, ratings among the three respondent groups for the nine domains of diversity can be determined based on the proportions and mean scores reported in Tables 2 and 3. Two trends are apparent in the data: (a) acceptance is rated higher than tolerance and indifference in all dimensions of diversity (see Table 2); and (b) there are differences regarding culture, religion and sexual orientation across the three different groups. These differences may well account for conflicting value‐related priorities across the three groups. For example, English students were more likely to accept differences in culture and religion than were the Spanish and American students. Similarly, English and Spanish students were more likely to report attitudes of acceptance toward differences in sexual orientation than were the American students. Overall, students’ attitudes toward people who differ were predominantly acceptance regarding disability, gender and special talents, which were the dimensions of diversity that students accepted the most (83%, 82%, and 80%, respectively), with differences in political ideology, religion and sexual orientation being the domains with less acceptance (39%, 57%, and 66%, respectively). Similarly, the different mean scores (Table 3) reflect cultural variability and unique experiences of differences (e.g., students with a relativistic view of difference were more tolerant toward people who differed in political ideology when compared to students with a more ethnocentric view). Thus, multiple perspectives and different samples should be incorporated into cross‐cultural research of this type and specific scores should be viewed within the context of the respondent group.
At the macro level, data in Table 3 clearly showed differences in main effects among the respondent groups with regard to culture, religion and sexual orientation, as well as main effect differences in political ideology due to relativism versus ethnocentrism. In other words, students with a relativistic point of view across groups showed more positive attitudes toward difference than those with a more ethnocentric view of difference.
However, the question that subsumes all others is ‘Why undertake such comparisons?’ Increasingly, diversity is viewed positively and measures of attitudes toward diversity have become tools for change, not simply as tools for comparison or judgement (Lubienski Citation2003). Teacher educators are interested both in teaching tolerance and acceptance and preparing pre‐service teachers to do the same. Subsequent investigators may want to explore why the attitude toward people who differ (Table 2) includes acceptance to a greater extent in some cultural contexts than in others. Is it an issue of ‘dominant’ versus ‘minority’ culture, a heightened sense of democracy, or something else? To what extent is tolerance linked to social and political history? Are there cultural factors which facilitate or inhibit positive attitudes, acceptance and respect toward people who differ?
Although the current study focused on the subjective viewpoints of respondents, future cross‐cultural studies should also consider the life experiences and circumstances of individuals. In that regard, we suggest the following guidelines. First, determine whether people feel as comfortable with their own diversity within the general societal context as do other population subgroups within that context. If these ratings are different, personal or environmental factors that might explain such differences should be investigated. Second, assess attitudes toward difference within other social, professional or population groups. Cross‐cultural differences in attitude ratings and scores may reflect differences due to the professional, legal and statutory position of people. For example, in this study can it be assumed that teachers have similar professional, legal and statutory positions across the three countries? Variations in beliefs and attitudes may be related to variations in the availability and accessibility of services, differences in societal attitudes, contrasting outlooks and expectations or simply be an artefact of the research methods employed (Buck et al. Citation1999).
Although exploratory, the present study is not without limitations. First, the sample size for some cultural groups was small. Second, issues related to cultural equivalence and/or cultural bias were not addressed directly. As with any study of attitudes, investigations of the construct are mediated by the specific measures used. Thus, the cross‐cultural variation obtained may be due to a lack of cultural equivalence and/or cultural bias in the measures used or a lack of reliability or validity in the measures used. Third, the sampling strategy, in this case convenience samples, places limits on the conclusions and generalisations about national variations in attitudes and beliefs toward difference. There is a need for subsequent studies to incorporate a follow‐up strategy that would allow more qualitative explanations of each domain. Fourth, the factor structure of BATD scale domains needs to be confirmed on a larger sample through confirmatory factor analysis. The etic and emic properties also need to be validated on a larger sample and through a more robust statistical procedure (e.g., structural equation modelling). And finally, it is necessary to consider the extent to which the respondents perceived that there were ‘right’ and ‘wrong’ answers to these questions.
As previously argued, reference is made to the universal characteristics (etic properties), cultural (emic properties), and individual characteristics of the phenomena or concept studied in trans‐cultural research. The universal characteristics of the construct of ‘difference’ have been studied and have demonstrated that the concept of ‘diversity’ has often been associated with the idea of deviance (Artiles Citation1998). However, our work suggests that there is a factorial structure in the instrument used to measure the beliefs and attitudes towards those who differ. The cultural properties of the construct are reflected in: (a) close variations in the trans‐cultural evaluation by respondents about diversity with respect to culture/ethnic origin, language, socioeconomic status/social class, religion, gender, sexual orientation, political ideology, disability and giftedness/special talents; and (b) with the observation that the stability of the instrument in measuring attitudes is not perfect (variations are observed in some of the dimensions). Finally, the individual connotations of the concept make themselves evident in observing differences from personal views (relativist versus ethnocentric) of difference.
In conclusion, through cross‐cultural studies such as this one, investigators can begin to answer a number of fundamental questions regarding attitudes toward diversity. Over the past decades, the concept of diversity has been used as a sensitising notion, a social construct and as a unifying framework for its conceptualisation, measurement, and application. In addition, in many countries, the concept of difference has been part of social movements with purposes related to civil rights, deinstitutionalisation, normalisation, mainstreaming and inclusion. If diversity is to continue to be a viable social construct as well as a unifying framework that guides international efforts, future cross‐cultural studies need to be designed to address at least four questions: How can these concepts be integrated into the major social forces impacting human behaviour? What is the proper use of attitude data toward difference? Does the measurement and change of attitudes make a difference in people’s lives? Can measuring attitudes be a useful part of university‐based courses in education by stimulating thought and discussion about matters of difference and diversity? In the present study we have attempted to address some of these key issues in a small but important step in creating more tolerant professionals to work in inclusive settings.
Notes on contributors
M. Cristina Cardona Moltó is professor of Special and Inclusive Education in the Faculty of Education at the University of Alicante, Spain.
Lani Florian is professor of Social and Educational Inclusion in the School of Education at the University of Aberdeen, Scotland.
Martyn Rouse is professor of Social and Educational Inclusion and director of the Inclusive Practice Project in the School of Education at the University of Aberdeen, Scotland.
Laura M. Stough is associate professor in the Department of Educational Psychology and interdisciplinary training director at the Center on Disability and Development, Texas A&M University.
Acknowledgements
This study was supported by a grant from the Generalitat Valenciana Department of Education and Science, Valencia (GV‐2004‐B‐692), and the Ministry of Education and Science, Madrid (Ref. PR2005‐0414), Spain.
Keywords
- cultural differences
- pre-service teacher education
- attitude change
Fingerprint
Dive into the research topics of 'Attitudes to diversity: a cross‐cultural study of education students in Spain, England and the United States'. Together they form a unique fingerprint.Cite this
- APA
- Standard
- Harvard
- Vancouver
- Author
- BIBTEX
- RIS