Training Framework

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies are encouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Assessing teachers’ multicultural and egalitarian beliefs: The Teacher Cultural Beliefs Scale

Axinja Hachfeld a,*, Adam Hahn b, Sascha Schroeder a,c, Yvonne Anders a,d, Petra Stanat e, Mareike Kunter a,f

a Max Planck Institute for Human Development, Berlin, Germany b University of Colorado at Boulder, Boulder, CO, USA c University of Kassel, Germany d Otto Friedrich University, Bamberg, Germany e Institute for Educational Progress, Berlin, Germany f Johann Wolfgang Goethe University, Frankfurt am Main, Germany

a r t i c l e i n f o

Article history: Received 22 October 2010 Received in revised form 21 April 2011 Accepted 26 April 2011

Keywords: Multiculturalism Color-blindness Beginning teachers Teacher beliefs Immigrant students

a b s t r a c t

The article describes the newly developed Teacher Cultural Beliefs Scale (TCBS). The TCBS assesses multicultural and egalitarian beliefs about diversity, both of which reflect favorable attitudes toward immigrant students, but differ with regard to how cultural diversity is believed to be best accommodated in schools. Results from a first study with 433 beginning teachers supported the two-factor structure and the measurement invariance of the scale. Results from a second study with 340 teacher candidates and educational science students showed that proponents of multiculturalism and egalitarianism shared a motivation to control prejudiced reactions, but that they differed in their views on acculturation, prejudices, and authoritarianism.

� 2011 Elsevier Ltd. All rights reserved.

Schools in most Western countries have become increasingly culturally diverse over recent decades; and as global migration continues to rise, cultural diversity in schools will continue to grow worldwide (UNESCO, 2004).At the same time, large-scale assess- ments such as PISA have repeatedly shown that immigrant students’ school achievement lags behind that of majority students in many of the participating countriesdbut especially in Germany (see Stanat & Christensen, 2006, for OECD and partner countries). It has been claimed that schools fail to address the needs of immigrant students, and that catering to these needs is among today’s major educational challenges (Vedder, Horenczyk, Liebkind, & Nickmans, 2006). This challenge cannot be met by top-down educational reforms alone, however. Instead, all educational actors will need to work together to address it. Given their central role in educational practice worldwide, teachers have a key function in this process. However, little is yet known about teachers’ attitudes, views, and beliefs about cultural diversity in schools (Vedder et al., 2006).

This article presents the Teacher Cultural Beliefs Scale (TCBS), which we have developed to assess teachers’ beliefs about dealing with cultural diversity in the classroom. Informed by social- psychological research on intergroup contact, the TCBS focuses on two beliefs: multiculturalism and egalitarianism. This article tests the measurement model of the TCBS in different samples (study 1) and examines the construct validity of its two subscales of multi- culturalism and egalitarianism (study 2). Both studies were con- ducted in Germany. However, we drew on international research to formulate our hypotheses and argue that the key question of how today’s teachers are responding to and dealing with increasing cultural diversity is of crucial relevance for schools around the globe. Accordingly, finding ways to access teachers’ beliefs about cultural diversity in schools is a precondition for “the transformation required for teachers to implement appropriate education” in culturally diverse settings (Harrington & Hathaway, 1995, p. 275).

1. Teacher beliefs

Through structuring the learning environment, teachers play a pivotal role in student learning. A large body of evidence from various countries shows that teachers’ beliefs significantly influence

* Corresponding author. Max Planck Institute for Human Development, Center for Educational Research, Lentzeallee 94, 14195 Berlin, Germany. Tel.: þ49 30 82406 671; fax: þ49 30 82406 433.

E-mail address: [email protected] (A. Hachfeld).

Contents lists available at ScienceDirect

Teaching and Teacher Education

journal homepage: www.elsevier.com/locate/tate

0742-051X/$ e see front matter � 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.tate.2011.04.006

Teaching and Teacher Education 27 (2011) 986e996

Author's personal copy

how they plan, organize, and implement their lessons and how responsive they are to their students (e.g., Staub & Stern, 2002; Stipek, Givvin, Salmon, & MacGyvers, 2001). Much of this research has focused on teachers’ beliefs about learning and instructing (Woolley, Benjamin, & Woolley, 2004), which have also been shown to affect students’ learning outcomes (Dubberke, Kunter, McElvany, Brunner, & Baumert, 2008; Staub & Stern, 2002).

Although it has been argued that beliefs also play a fundamental role in how teachers encounter diversity in the classroom (Harrington & Hathaway, 1995), most research on teaching effec- tiveness to date has been ‘color-blind’ (Zeichner, 1992). Conse- quently, very little is known about teachers’ beliefs about cultural diversity or about how those beliefs affect immigrant students. Most research in this domain has focused on the effects of preju- dices and discrimination on immigrant students’ achievement (Schofield, 2006). Unsurprisingly, the empirical findings support the common knowledge that “a [.] favourable attitude [.] contributes to the development and learning of immigrant youth” (Vedder et al., 2006, p. 10). However, as we discuss below, social- psychological research on intergroup relations suggests that different ‘favorable’ beliefs can be distinguished and may have differing implications for social interaction. Two such distinguish- able beliefs are multiculturalism and egalitarianism.

In social psychology, these beliefs are often referred to as ideologies; in educational psychology, in contrast, the term ‘beliefs’ is used to describe “suppositions, commitments, and ideologies” (Calderhead, 1996). In line with educational research, we use the term ‘beliefs’ to refer to attitudes, views, ideologies, or models that teachers hold about students with a different cultural background from their own. Multicultural and egalitarian beliefs both reflect positive attitudes toward immigration; however, proponents of multiculturalism and proponents of egalitarianism differ in how they encounter, interpret, and respond to diversity.

Multicultural beliefs recognize that, because individuals have engaged with different socio-cultural contexts, they have legiti- mately different perspectives and beliefs. These differences are seen as difficultdif not impossibledto ignore (Park & Judd, 2005). Proponents of multiculturalism believe that these intercultural differences should be embraced and viewed as enriching. In the educational context, teachers with multicultural beliefs can be expected to incorporate students’ different cultures into everyday school practice when planning their lessons, choosing materials, and interacting with students in class.

Egalitarian beliefs emphasize the importance of treating all people equally. This implies finding similarities and common grounds between students of different ethnic and cultural back- grounds, regardless of their ethnic or cultural background (Markus, Steele, & Steele, 2002; Richeson & Nussbaum, 2004; Wolsko, Park, & Judd, 2002). In line with a ‘color-blind’ ideal, proponents of egali- tarianism often argue that categorization on the basis of ethnicity or culture is one of the sources of discrimination (Park & Judd, 2005) and that it should be avoided (Richeson & Nussbaum, 2004). In the educational context, teachers with strong egali- tarian beliefs pay less attention to the cultural backgrounds of their students, focusing instead on their similarities and seeking to treat all students equally. As a consequence, they can be expected to favor a common curriculum and to give students’ cultural back- ground less consideration in their lesson planning.

These two beliefs are not mutually exclusive (Plaut, 2010); in fact, they are conceptually independent. For example, a teacher may hold egalitarian beliefs to some degree, but at the same time try to accommodate cultural differences. However, as outlined above, social-psychological research suggests that the two beliefs have different implications for intergroup contact and that they should be distinguished. Although the requisite empirical data from

educational research are not yet available, it seems likely that the two beliefs also have different implications for instruction.

Various authors in the field of social psychology have discussed the benefits and drawbacks of the two beliefs; however, these studies have not targeted teachers (Markus et al., 2002; Park & Judd, 2005; Plaut, Thomas, & Goren, 2009; Richeson & Nussbaum, 2004; Verkuyten, 2005; Vorauer, Gagnon, & Sasaki, 2009; Wolsko et al., 2002; Wolsko, Park, & Judd, 2006; Wolsko, Park, Judd, & Wittenbrink, 2000). It should also be noted that what we describe as ‘egalitarian’ beliefs are termed ‘color-blindness’ in most social-psychological publications. Given the conceptual ambigui- ties of the term ‘color-blindness’ when applied to an educational context and its ties to U.S. history, however, we prefer to use the broader term ‘egalitarianism’ (see Park & Judd, 2005; Wolsko et al., 2002). This term can be applied to any cultural setting, indepen- dently of a country’s immigration history. We do, however, draw on findings from color-blindness research to guide our hypotheses.

Various lines of research have highlighted important differences between multiculturalism and egalitarianism or color-blindness (for an overview, see Park & Judd, 2005). Most importantly in the present context, multicultural perspectives have been shown to lead to a more outward focus in interethnic interactions (Vorauer et al., 2009), to more empathy (Burkard & Knox, 2004), and to a stronger perception and acceptance of differences between people (Wolsko et al., 2000). Treating all people equally, regardless of their cultural background, seems worthwhile for its own sake. However, research findings suggest that egalitarian or color-blind perspectives can induce a prevention orientation in interethnic encounters (Vorauer et al., 2009), which can in turn lead to more negative implicit attitudes toward minority members in high- conflict situations (Correll, Park, Judd, & Wittenbrink, 2002). Research has also shown that egalitarianism or color-blindness can reduce in-group positivity and be associated with decreased levels of expressed stereotypicality (Wolsko et al., 2000).

Two hypotheses can be formulated on the basis of previous research. First, the findings suggest that multicultural and egali- tarian beliefs may also be important for instruction: An open, empathetic, and student-oriented attitude is crucial for positive teacherestudent interactions. Moreover, the accurate and prejudice- free assessment of students’ strengths and weaknesses is a necessary condition for providing each student with the optimal support. However, due to the lack of empirical evidence, it remains unclear whether multicultural or egalitarian beliefs are more beneficial for immigrant students. Research on teacher beliefs has been hampered by the lack of instruments assessing these beliefs in an educational context.

Second, the empirical data indicate that the two beliefs need to be assessed separately. Although several attempts have been made to assess teachers’ cultural beliefs by means of questionnaire measures (Amodeo & Martin, 1982; Barry & Lechner, 1995; Easter, Shultz, Neyhart, & Reck, 1999; Pohan & Aguilar, 2001; Stanley, 1996; Tabachnick & Zeichner, 1984), none of the instruments applied to date in educational research have explicitly assessed multicultural and egalitarian beliefs side by side. Two scales that have been vali- dated and applied in studies with teachers are the Personal and Professional Beliefs About Diversity Scale by Pohan and Aguilar (2001) and the Pluralism and Diversity Attitudes Assessment (PADAA) by Stanley (1996). These scales measure appreciation of diversity, rather than distinguishing between different favorable beliefs. Additionally, they take a broader approach to diversity, including items on social class, gender, religion, sexual orientation, and language (other than English), as well as on ethnicity and race. Both scales have proven useful for investigating general attitudes toward diversity and for studying the effects of multicultural education programs on participants’ attitudes. However, they neither explicitly

A. Hachfeld et al. / Teaching and Teacher Education 27 (2011) 986e996 987

Author's personal copy

assess multicultural or egalitarian beliefs nordmore generallyd distinguish between different favorable attitudes. Against this background, we developed the TCBS to assess teachers’ multicultural and egalitarian beliefs about cultural diversity in school. The development of this instrument is a first step toward a better understanding of how teachers are responding to and dealing with the increasing cultural heterogeneity of their classrooms. We hope that this enhanced understanding will ultimately contribute to improving the learning outcomes of immigrant students.

Most research conducted to date on teachers’ perceptions of diversity has involved small-sample qualitative studies. There have recently been calls for validated quantitative instruments that can be applied in large-sample studies (Cardona Moltó, Florian, Rouse, & Stough, 2010; for an overview of studies in the educational context, see, e.g., Castro, 2010 and the report of the AERA Panel on Research and Teacher Education edited by Cochran-Smith & Zeichner, 2005).The TCBS was developed with the explicit goal of providing a validated instrument that can be easily implemented in studies of various sample sizes. However, as we discuss in the General Discussion, quantitative methods are not without limita- tions. Before presenting our empirical studies in detail, we first describe the development of the TCBS and specify the aims and research questions of the two studies.

2. The Teacher Cultural Beliefs Scale

We used a multi-step procedure to develop items tapping multicultural and egalitarian beliefs, beginning by reviewing the definitions of the constructs and screening previous social- psychological research for existing items (Park & Judd, 2005; Wolsko et al., 2000, 2002).The multiculturalism subscale was designed to measure how strongly participants endorse the belief that the cultural background of students and their parents should be taken into consideration in teaching. Based on this definition, we modified the wording of existing items (Wolsko et al., 2002) to target schools or instruction rather than society in general and developed additional items. The egalitarianism subscale was designed to measure how strongly participants focus on cultural similarities and endorse the belief that all students should be treated equally, regardless of their cultural background. Because our approach emphasized a focus on similarities rather than a disregard of cultural differences (as often is the case for color- blindness), we did not adapt existing items assessing disregard of cultural differences, but developed new items in line with our definition of the construct. The 16 items emerging from this procedure were discussed with several independent experts in social psychology (with a focus on cultural beliefs) and educational science in the United States and Germany. In response to these experts’ comments, we dropped two items from the multicultur- alism subscale that dealt with interethnic contact and conflicts (e.g., “Schools also serve the purpose of facilitating interethnic contact.”) and four items from the egalitarianism subscale that lacked a focus on similarities (e.g., “Cultural differences should not be brought up in class instruction.”), leaving a final set of 10 items. The expert comments also prompted minor changes in the wording of some of these items.

3. Aims and research questions

3.1. Study 1

The purpose of study 1 was twofold. First, it sought to test the measurement model, factor structure, and reliabilities of the newly developed instrument. In social-psychological research, multicul- turalism and egalitarianism are conceptualized as two independent

constructs, both of which reflect favorable attitudes toward minorities. Hence, we expected that a two-factor solution would provide the best fit to the data, but also that confirmatory factor analyses would show a positive correlation between the two factors. Second, as an aspect of factorial validity, study 1 compared the measurement model across groups of beginning teachers with different degrees of experience of cultural diversity. Previous research has shown that individuals’ experiences of cultural diversity affect their cultural beliefs, and several studies have highlighted the importance of classroom experience for (begin- ning) teachers’ beliefs and attitudes about cultural diversity (for an overview, see Castro, 2010). In constructing the TCBS, we aimed to develop an instrument that could be used to compare the beliefs of teachers with varying degrees of cultural experience. We therefore tested the measurement model in different samples, testing for (a) measurement invariance and (b) mean differences. Measurement invariance across groups is often regarded as a prerequisite for the comparison of group means (Brown, 2006). Although we did not expect different factor structures across groups, there is reason to hypothesize that teachers with varying degrees of cultural experi- ence differ in their support of multiculturalism and egalitarianism. As a first proxy for cultural experience in the school context, we therefore compared beginning teachers at the start and in their second year of the teacher placement phase (the samples are described in more detail in the Method section of study 1). The longitudinal findings of Bender-Szymanski (2000) suggest that beginning teachers’ endorsement of multicultural beliefs decreases significantly during the first year of teaching. At the same time, Castro (2010) has pointed out that the younger generation of beginning teachers is generally more multicultural and that higher endorsement of multicultural beliefs among younger beginning teachers may be a cohort effect. Given these results, we expected teachers in their first and second year of classroom teaching to differ in their endorsement of multiculturalism. Because we could not draw on any prior research about beginning teachers’ endorsement of egalitarianism, we did not formulate specific hypotheses about group differences in egalitarianism.

The second proxy for differing exposure to cultural diversity was school track. Especially in the United States, studies often compare urban with suburban schools. For the German context, however, we reasoned that the school track would be a better proxy for cultural diversity than the location of the school. The German secondary system differentiates college-track schools, which qualify students for university admission, from vocational-track schools. The percentage of immigrant students is considerably higher in vocational-track schools, meaning that teachers in this track gain more experience interacting with culturally diverse students than do teachers in college-track schools. This higher level of intercul- tural interaction may lead to greater acceptance and appreciation of cultural diversity (Castro, 2010) and hence to higher endorsement of multicultural beliefs. At the same time, teachers starting their career at a vocational-track school may learn that a focus on similarities is not feasible in a culturally diverse school setting, leading to lower endorsement of egalitarian views. In sum, we expected to find mean differences between beginning teachers in college-track and vocational-track schools. Again, in order to be able to compare mean differences, we first aimed at establishing measurement invariance.

Finally, we compared mean differences in the endorsement of multicultural and egalitarian beliefs between beginning teachers with and without an immigrant background. Previous research suggests that people from an immigrant or minority background are more likely to endorse multicultural beliefs than are people from the majority culture (for the Netherlands, see van de Vijver, Breugelmans, & Schalk-Soekar, 2008; for the United States, see

A. Hachfeld et al. / Teaching and Teacher Education 27 (2011) 986e996988

Author's personal copy

Wolsko et al., 2006). We expected similar results for our sample of beginning teachers. In turn, we expected participants without an immigrant background to support egalitarian beliefs more strongly than participants with an immigrant background.

3.2. Study 2

Study 2 was conducted to test the construct validity of the two subscales of the TCBS. As discussed above, multiculturalism and egalitarianism are two distinct beliefs that both reflect favorable attitudes toward immigrant students, but differ with regard to how cultural diversity is perceived and believed to be best accommo- dated in schools. Whereas multicultural beliefs stress accommo- dating and respecting students’ backgrounds, egalitarian beliefs emphasize similarities and equal treatment of all students. We expected that these differences would also be reflected in the relationship of the constructs with (a) prejudices and the motiva- tion to control prejudiced behavior, (b) attitudes toward pluralism and acculturation, and (c) authoritarianism as one aspect of teaching style. Whereas multiculturalism and egalitarianism can both be expected to be associated with the motivation to control prejudiced behavior and with positive attitudes toward diversity, we expected proponents of the two beliefs to differ with regard to their attitudes toward acculturation and their endorsement of authoritarianism. We outline the hypotheses that guided our choice of measures in more detail in our presentation of study 2.

4. Study 1

4.1. Method

4.1.1. Participants and procedure Participants (N ¼ 433) were beginning teachers taking part in

the COACTIV-R study, which was conducted to assess professional competence and development during teacher education in Germany. Teacher education in Germany is divided into two pha- ses: a first university-based phase and a second phase of classroom teaching practice. All COACTIV-R participants were in the second phase of teacher education. During this 2-year placement phase, beginning teachers are expected to acquire the practical knowledge of learning and instruction considered crucial for their profession (Oser, Achtenhagen, & Renold, 2006). They are introduced to the

practicalities of teaching under supervision of a mentor. Depending on their degree program, they teach either in vocational-track schools (grades 5e10) or in college-track schools (grades 5e13).

The study, which conformed to APA ethical standards, was cross-sectional. Before beginning the study, we obtained the approval of the ethics committees of the two German federal states in which data were collected. Questionnaires were sent to partici- pants at the beginning of the first (cohort 1, n ¼ 332) or second (cohort 2, n ¼ 101) year of their teacher placement phase and were answered at home. Data from cohort 2 were collected in only one of the federal states. Hence, analyses comparing the two cohorts are based on data from only one state. Participation was voluntary and participants received monetary compensation. The items relevant for the present study were administered as part of a larger ques- tionnaire tapping beginning teachers’ knowledge, beliefs, socio- demographic data, and information on the teacher training program.

Participants (65.3% women, n ¼ 277) were on average 28.2 years old (SD ¼ 4.75; range: 23e57 years). Of these, 45.3% (n ¼ 196) were teaching in the vocational track, and 14.2% (n ¼ 60) were from an immigrant background.

4.1.2. Instruments As described above, the TCBS consists of six items designed to

measure multicultural beliefs and four items designed to measure egalitarian beliefs. Responses were given on a 6-point scale ranging from 1 (strongly disagree) to 6 (strongly agree). All items are listed in Table 1.

4.2. Analyses and results

4.2.1. Measurement model of the TCBS We conducted confirmatory factor analyses (CFA) using the

Mplus software (Muthén & Muthén,1998e2006) to test whether the hypothesized two-factor structure of multiculturalism and egali- tarianism emerged. Parameters were estimated using the maximum-likelihood (ML) fitting function. We evaluated model fit according to criteria suggested by Hu and Bentler (1999). There is growing consensus that imputation of missing observations and maximum-likelihood approaches are preferable to pairwise or list- wise deletion (Schafer & Graham, 2002). In the present analyses, we chose the full-information-maximum-likelihood (FIML) procedure

Table 1 Results of confirmatory factor analysis: standardized factor loadings for the TCBS in both studies.

Teacher Cultural Beliefs Scale (TCBS) CFA study 1 CFA study 1 (e) CFA study 2

Factor 1: multicultural beliefs 1. In the classroom, it is important to be responsive to differences between cultures. .770 .722 .519 2. It is important for children to learn that people from other cultures can have different values. .613 .623 .656 3. Respecting other cultures is something that children should learn as early as possible. .615 .616 e 4. In counseling parents who have a different cultural background than I do, I try to be

considerate of cultural particularities. .567 .575 .605

5. When meeting with parents of different cultural backgrounds, I spend more time trying to understand and empathize with their perspective.

.551 .547 e

6. Dealing with cultural diversity should be taught in teacher training courses. .400 .411 .595

Factor 2: egalitarian beliefs 7. Schools should aim to foster and support the similarities between students from different

cultural backgrounds. .782 .780 .650

8. In the classroom, it is important that students of different origins recognize the similarities that exist between them.

.857 .850 .824

9. When there are conflicts between students of different origins, they should be encouraged to resolve the argument by finding common ground.

.689 .683 .684

10. Children should learn that people of different cultural origins often have a lot in common. .406 .367 .740

Latent correlation between the two factors .625 .613 .736

Note: all items were administered in German. Study 1: sample of beginning teachers (N ¼ 376). Study 1 (e): study 1 sample without beginning teachers from an immigrant background (N ¼ 325). Study 2: sample of teacher candidates (75%) and educational science students (Nfull sample ¼ 339).

A. Hachfeld et al. / Teaching and Teacher Education 27 (2011) 986e996 989

Author's personal copy

that is implemented in Mplus (Graham & Hoffer, 2000). Two items from the multiculturalism subscale assessed how sensitive partici- pants were to cultural background when meeting with or counseling parents (items 4 and 5). These two similarly worded items were more similar to each other than to the remaining items. We therefore allowed for a correlation between the two items in both models.

The descriptive fit indices of the CFA testing the hypothesized two-factor model indicated a good fit, c2(33) ¼ 70.344, CFI ¼ .966, TLI ¼ .951, RMSEA ¼ .055, SRMR ¼ .039. We also estimated a one- factor model, c2(34) ¼ 216.280, CFI ¼ .834, TLI ¼ .780, RMSEA ¼ .119, SRMR ¼ .078, and tested it against the two-factor model. The results confirmed that the two-factor model provided a better fit to the data than the one-factor model, Dc2(1) ¼ 145.936, p < .001. The two- factor structure was also supported by the latent correlation between multiculturalism and egalitarianism, which was only moderately high (r ¼ .63). The results thus supported the hypothe- sized two-dimensional structure of the TCBS. The reliabilities of the two scales were also satisfactory (Cronbach’s a ¼ .75 for multicul- turalism; a ¼ .78 for egalitarianism). All factor loadings were statis- tically significant and are listed in the first column of Table 1.

4.2.2. Measurement invariance of the TCBS In a second step, we tested whether the hypothesized two-factor

structure held across groups of beginning teachers at different stages of their education (cohort 1 and cohort 2) and in different school tracks (college and vocational) as an aspect of factorial val- idity. To this end, we conducted multigroup CFAs in Mplus to test for measurement invariance (see Cheung & Rensvold, 2002).

First, we estimated overall model fit for the entire sample (equal form; Brown, 2006). We then estimated two models with addi- tional between-group constraints: (a) equality of factor loadings and (b) equality of factor loadings and intercepts. To check for significant differences, we tested each model against the one with one additional constraint (equal form vs. equal factor loadings, and equal factor loadings vs. equal factor loadings and intercepts). Because we expected measurement invariance to hold, we did not expect to find differences in the factor structure across the two cohorts or the different school forms. In view of the small-sample size, we did not test for measurement invariance between partici- pants with and without an immigrant background.

For the two cohorts, the constraint of equal factor loadings did not lead to a significant decrease in model fit, indicating that the subscales measured the same constructs for beginning teachers in both cohorts. Assuming equal intercepts in addition to equal factor

loadings did not decrease model fit either (all c2 comparisons were nonsignificant, see Table 2).We can therefore conclude that both cohorts showed comparable means.

For the different tracks, the constraint of equal factor loadings again did not lead to a significant decrease in model fit, indicating that the subscales measure the same constructs for teachers in the different tracks. Assuming equal intercepts in addition to equal factor loadings did, however, significantly decrease the fit of the model (Table 3), indicating that the degree to which teachers in the different tracks endorsed multiculturalism and egalitarianism differed.

To compare the factor means of the different cohorts and tracks, we used multiple group mean structure comparisons and set the latent means and variances for multiculturalism and egalitarianism to be equal across groups. In view of the small-sample size, we used the traditional t test for the comparison of participants with and without an immigrant background.

Between the cohorts, there was no significant difference in beginning teachers’ multicultural, DMK ¼ 0.004, ns, and egalitarian beliefs, DEG ¼ 0.081, ns, (Dc2(4) ¼ 2.796, ns), replicating the results of the test of measurement invariance. Beginning teachers in cohort 1 strongly endorsed multicultural (M ¼ 4.72, SD ¼ .63) and egali- tarian beliefs (M ¼ 5.06, SD ¼ .65), as did their counterparts in cohort 2 (multiculturalism: M ¼ 4.78, SD ¼ .61; egalitarianism: M ¼ 5.07, SD ¼ .58).

Between teachers in different tracks, however, there were significant differences in multiculturalism, DMK ¼ 0.487, p < .001, and egalitarianism, DEG ¼ 0.267, p < .05, with beginning teachers in vocational-track schools as the reference group,Dc2(4) ¼ 19.459, p < .001. Beginning teachers in vocational-track schools endorsed multiculturalism more strongly (M ¼ 4.86, SD ¼ .57) than did those in college-track schools (M ¼ 4.63, SD ¼ .65). At the same time, they also endorsed egalitarianism more strongly (M ¼ 5.13, SD ¼ .60) than did their counterparts in college-track schools (M ¼ 5.00, SD ¼ .65).

In view of the small-sample size, we did not test for measurement invariance between participants with and without immigrant background. However, we used t tests to compare mean differences between beginning teachers with and without immigrant back- ground. The results revealed significant differences in the endorse- ment of multiculturalism, t(396) ¼ �2.189, p < .05. Beginning teachers with an immigrant background supported multiculturalism significantly more strongly than did those without an immigrant background (immigrant background: M ¼ 4.92, SD ¼ .60; without

Table 2 Results from multigroup confirmatory factor analyses testing measurement invariance of teachers’ cultural beliefs in cohort 1 and cohort 2.

c2 df c2diff Ddf RMSEA (90% CI) SRMR CFI TLI

Equal form 103.679 66 .077 (.047e.104) .064 .941 .920 Equal factor loadings 117.630 76 13.951 10 .075 (.047e.101) .128 .935 .923 Equal factor loadings and intercepts 124.596 86 6.969 10 .068 (.039e.093) .137 .940 .937

Note. N ¼ 193. c2diff: nested c2 difference; RMSEA: root mean square error of approximation; 90% CI: 90% confidence interval for RMSEA; SRMR: standardized mean square residual; CFI: comparative fit index; TLI: TuckereLewis Index.

Table 3 Results from multigroup confirmatory factor analyses testing measurement invariance of teachers’ cultural beliefs for beginning teachers in vocational- and college-track schools.

c2 df c2diff Ddf RMSEA (90% CI) SRMR CFI TLI

Equal form 96.840** 66 .050 (.026e.070) .046 .971 .961 Equal factor loadings 108.486** 76 11.646 10 .048 (.025e.067) .085 .970 .964 Equal factor loadings and intercepts 151.100*** 86 42.614* 10 .063 (.046e.080) .130 .939 .937

Note. N ¼ 367. c2diff: nested c2 difference; RMSEA: root mean square error of approximation; 90% CI: 90% confidence interval for RMSEA; SRMR: standardized mean square residual; CFI: comparative fit index; TLI: TuckereLewis Index. ***p < .001. **p < .01. *p < .05.

A. Hachfeld et al. / Teaching and Teacher Education 27 (2011) 986e996990

Author's personal copy

immigrant background: M ¼ 4.71, SD ¼ .63). There was no significant difference between beginning teachers with and without an immi- grant background in their support of egalitarianism, t(371) ¼ �1.269, p ¼ .21 (with immigrant background: M ¼ 5.17, SD ¼ .60; without immigrant background: M ¼ 5.04, SD ¼ .63).

In a final step, we tested for mean differences between multi- culturalism and egalitarianism and compared these differences across groups (cohorts, tracks, immigrant background). On average, beginning teachers strongly endorsed both beliefs (multicultur- alism: M ¼ 4.73, SD ¼ .63; egalitarianism: M ¼ 5.06, SD ¼ .63; range for both subscales: 1e6). However, they endorsed egalitarian beliefs significantly more strongly than multicultural beliefs, t(373) ¼ �9.600, p < .001.

4.3. Discussion

The results of factor analyses confirmed the theoretically pre- dicted two-factor structure with the factors multiculturalism and egalitarianism. Reliabilities were satisfactory for both scales. We conducted tests for measurement invariance and group comparisons for teachers of different cohorts and in different school types as proxies for differential exposure to cultural diversity in the class- room. Results showed equal factor loadings and equal intercepts for beginning teachers in the first and second cohort, confirming measurement invariance. Contrary to our hypotheses, beginning teachers in the two cohorts did not differ in their endorsement of multicultural beliefs. Likewise, we found no significant differences in their egalitarian beliefs. For beginning teachers in college- vs. voca- tional-track schools, we found equal factor loadings but different intercepts, suggesting that some items may function differentially for these two groups. This difference may be attributable to differ- ences in the degree of cultural diversity of the respective teaching environments. As expected, beginning teachers in vocational-track schools endorsed multicultural beliefs significantly more strongly than did their counterparts in college-track schools. However, and contrary to our hypotheses, they also endorsed egalitarian beliefs more strongly. Possible interpretations of our findings of higher multicultural beliefs in vocational-track teachers are that greater exposure to cultural differences raises these teachers’ awareness of cultural differences. On the other hand, teachers with multicultural beliefs may be more likely to decide to teach at more culturally diverse schools. Their heightened awareness of cultural differences may also make these teachers more ambitious to achieve ideals of equality, as reflected in their endorsement of egalitarian beliefs.

Descriptive results on the manifest level showed thatdwhereas beginning teachers tended to endorse egalitarian beliefs more strongly than multicultural beliefsdthis pattern was reversed for beginning teachers with an immigrant background, who showed significantly higher endorsement of multicultural beliefs than did beginning teachers without an immigrant background. This finding replicates results from various (non-teacher) samples in the Netherlands and the United States, which have shown that people with an immigrant or minority background endorse multicultur- alism more strongly than do majority members of society (van de Vijver et al., 2008; Wolsko et al., 2006).

In sum, the aim of study 1 was to test the newly developed TCBS instrument and to confirm the hypothesized two-factor structure of multiculturalism and egalitarianismda first but necessary step in establishing the new scale for use in future research. In a second step, we tested the construct validity of the two subscales.

5. Study 2

Study 2 tested the construct validity of the two subscales of the TCBS: multiculturalism and egalitarianism. To this end, we derived

several hypotheses from social-psychological research on multi- culturalism and egalitarianism. The following hypotheses guided our choice of measures.

5.1. Prejudices and motivation to control prejudiced behavior

Research has shown that the motivation to act without preju- dice moderates the relation between implicit and explicit stereo- types about gender (Banse & Gawronski, 2003). Two aspects of prejudiced behavior can thus be differentiated: expression of prejudices and the motivation to act without prejudice.

Based on previous research, we expected multiculturalism and egalitarianism to show different patterns of relations to the expression of prejudices. Social-psychological research suggests that multicultural beliefs are associated with fewer prejudices (Park & Judd, 2005; Wolsko et al., 2000, 2006). Findings for egalitari- anism are not as clear-cut, however. On the one hand, research on color-blindness has shown that color-blind perspectives, although initially promoting an anti-prejudice stance, can lead to more prejudicial behaviors in actual interactions (Richeson & Nussbaum, 2004). On the other hand, a study by Wolsko and colleagues showed that, relative to a control condition, White participants primed with a color-blind message made more positive evaluations of African Americans (Wolsko et al., 2000). However, as both beliefs theoretically reflect a positive regard for immigrant students, we expected both subscales to be positively related to the motivation to control prejudiced behavior.

5.2. Attitudes toward pluralism and acculturation

Empirical evidence on the relationship between multicultur- alism and egalitarianism, on the one hand, and attitudes toward pluralism and acculturation, on the other, is still scarce.

Because both subscales of the TCBS assess an appreciation of cultural diversity in school, we expected both subscales to be positively related to attitudes toward pluralism. Given the two beliefs’ differing emphasis on students’ cultural backgrounds and their relevance for teaching practice, however, we expected to find a stronger relationship for multiculturalism than for egalitarianism. Likewise, multiculturalism and egalitarianism entail differing views on intergroup contact and on how acculturation should take place (Park & Judd, 2005). Teachers with multicultural beliefs can be expected to be more accepting of cultural differencesda necessary condition for the integration of other cultures into mainstream societydand hence to have more positive attitudes toward accul- turation than teachers with egalitarian beliefs, whose focus is more on equality.

5.3. Authoritarianism

Social-psychological research in the U.S. context has shown that authoritarianism is correlated with political conservatism and with opposition to societal change and progress (Jost, Glaser, Kruglanski, & Sulloway, 2003). Accepting different cultural influences in schools requires a certain level of openness to changing societal structures, but also to the shortcomings of one’s own culture and openness to new points of view. Petzel, Wagner, van Dick, Stellmacher, and Lenke (1997) found that high teacher authoritarianism correlated nega- tively with what they called “pedagogically useful” conflict solving (e.g., discussion). The authors argued that authoritarian teachers show higher approval for hierarchical school structures and are less open to emancipated teacherestudent relationships. Moreover, authoritarianism often seems to be associated with ethnic preju- dices, rather than with respectful attitudes toward other cultures (Petzel, Wagner, van Dick, et al., 1997; Zick & Six, 1997).

A. Hachfeld et al. / Teaching and Teacher Education 27 (2011) 986e996 991

Author's personal copy

We therefore expected scores on the multiculturalism subscale to be negatively correlated with authoritarianism. Egalitarian beliefs, in contrast, are largely compatible with traditional Western beliefs of equality and thus do not require openness to different viewpoints and to societal changes. Hence, we did not expect scores on the egalitarian subscale to show any relationship with authoritarianism.

5.4. Method

5.4.1. Participants and procedure A total of N ¼ 340 students (233 women) sampled from

a German university participated in this questionnaire study. Of the participants, 79% (n ¼ 254) were of German nationality and 21% (n ¼ 68) had an immigrant background. Ages ranged from 19 to 55 years (M ¼ 25, SD ¼ 5). The sample comprised 266 (81%) teacher candidates; the remaining 19% (n ¼ 62) were studying educational science. Overall, 55% (n ¼ 182) had already taught at schools or in comparable settings (63%, n ¼ 166, of the teacher candidates, and 23%, n ¼ 13, of the educational science students answered ‘yes’ when asked whether they had teaching experience). Of those with teaching experience (n ¼ 175), 83% (n ¼ 146) reported having worked with immigrant students (88%, n ¼ 142, of the teacher candidates and 31%, n ¼ 4, of the educational science students). About half of the participants (47%, n ¼ 145) had worked with immigrant children beyond the academic setting (e.g., as a camp counselor), and 61% (n ¼ 195) reported having discussed issues of intercultural education in their professional education courses.

Students were recruited in their university seminars, but participation was voluntary. APA ethical standards were followed in the conduct of the study. The study questionnaire was administered in the last 10 min of the class, and participants received chocolate bars as an incentive.

5.4.2. Instruments The questionnaire included a shortened version of the TCBS

containing four multiculturalism items and four egalitarianism items. The questionnaire also included items assessing prejudices and motivation to control prejudiced behavior, attitudes toward pluralism and acculturation, and authoritarianism. Additionally, socio-demographic data were assessed.

5.4.2.1. Prejudices and motivation to control prejudiced behav- ior. The behavior control subscale from the German version of the Motivation to Control Prejudiced Reactions scale (Banse & Gawronski, 2003) was used to assess participants’ motivation to control prejudiced behavior. Participants gave their responses on a 5-point agreeedisagree scale, with higher scores reflecting higher self-control and higher motivation to control prejudiced behavior.

Participants’ prejudices toward foreigners living in Germany were assessed with four prejudiced statements taken from the German General Social Survey (Ganter, 2001). Again, participants gave their responses on a 5-point agreeedisagree scale, with higher scores reflecting more prejudiced views toward foreigners.

5.4.2.2. Attitudes toward pluralism and acculturation. Two scales assessed attitudes toward pluralism and acculturation: a translated and slightly adapted and shortened version of the PADAA (Stanley, 1996) and the German Acculturation scale (van Dick, Wagner, Adams, & Petzel, 1997). The original items of the PADAA were designed for physical education teachers. For the purposes of the present study, physical education was replaced by lessons in general. Participants gave their responses on a 6-point agreee disagree scale, with higher scores reflecting higher endorsement of pluralism. We further administered the 10 items of the integration and assimilation (reverse-recoded) subscales from the German Acculturation scale (van Dick et al., 1997). Participants gave their responses on a 6-point agreeedisagree scale, with higher scores reflecting more favorable views on cultural integration.

5.4.2.3. Authoritarianism. We used a teacher-specific authoritari- anism scale (Petzel, Wagner, Nicolai, & van Dick, 1997) to measure the tendency toward an authoritarian teaching style. Responses were made on 5-point agreeedisagree scales, with higher scores indicating higher agreement with authoritarianism.

The psychometric properties of all scales are reported in Table 4. All scales had acceptable or good reliabilities.

5.5. Analyses and results

5.5.1. Measurement model of the TCBS As in study 1, we used Mplus to perform CFA to test for the

hypothesized factor structure of the TCBS. The two-dimensional model showed good fit, c2(19) ¼ 26.789, CFI ¼ .990, TLI ¼ .985, RMSEA ¼ .035, SRMR ¼ .024; more importantly, it fared better than a one-dimensional model, c2(20) ¼ 78.783, CFI ¼ .924, TLI ¼ .893, RMSEA ¼ .093, SRMR ¼ .051, and the chi-square difference test was significant: Dc2(1) ¼ 51.994, p < .001. All factor loadings were statistically significant (see Table 1, column 3). The correlation between the two subscales was r ¼ .54, p < .001. This replication of the results of study 1 in a different sample is evidence for the stability and generalizability of the instrument.

5.5.2. Validation of the TCBS To validate the multiculturalism and egalitarianism scales, we

estimated a fully identified path model with the two subscales of the TCBS as predictors and the other scales as outcome variables. The advantage of a path model over simple correlations was that

Table 4 Descriptive statistics for scales administered in study 2.

Scale Example item Number of items

Mean SD Cronbach’s alpha

MC In the classroom, it is important to be responsive to differences between cultures. 4 4.91 .78 .68 EG In the classroom, it is important that students of different origins recognize the

similarities that exist between them. 4 4.95 .87 .81

MCPRa One should never say anything derogatory about minorities in public. 7 3.68 .72 .76 Prejudicesa Foreigners living in Germany should seek their spouses within their own ethnic group. 4 1.76 .57 .64 PADAA Cultural diversity is a valuable resource and should be preserved. 12 5.04 .54 .75 ACS It would be good if all ethnic groups in Germany maintained their cultural identity. 10 3.96 .61 .74 AUTa The most important things for learning are discipline and obedience. 8 2.60 .61 .73

MC: multiculturalism; EG: egalitarianism; MCPR: Motivation to Control Prejudiced Reactions scale (behavior control subscale); prejudices: German General Social Survey; PADAA: Pluralism and Diversity Attitude Assessment; ACS: Acculturation scale; AUT: authoritarianism.

a Scale ranged from 1 to 5; all other scales ranged from 1 to 6. SD: Standard deviation.

A. Hachfeld et al. / Teaching and Teacher Education 27 (2011) 986e996992

Author's personal copy

we were able to control for the effects of the respective other subscale and to analyze the relationships between the two subscales and the outcomes simultaneously. The results of the path model are presented in Table 5.

We tested for differential effects by imposing equality constraints on the unstandardized paths of the effects of multi- culturalism and egalitarianism on the respective outcomes and comparing the fit of the models.1 In the following, we report results from the path model for (a) prejudices and the motivation to control prejudiced behavior, (b) attitudes toward pluralism and acculturation, and (c) authoritarianism, separately.

5.5.2.1. Prejudices and the motivation to control prejudiced behavior. Based on previous research, we expected multicultur- alism and egalitarianism to be equally associated with the moti- vation to control prejudiced behavior. This hypothesis was supported by the results of the path analysis: multiculturalism and egalitarianism were both positively and significantly related to the motivation to control prejudiced behavior. To test for differential effects, we set the coefficients of multiculturalism and egalitarianism to be equal for motivation to control prejudiced behavior. In line with our expectations, there was no significant decrease the fit of the model, Dc2(1) ¼ 0.302, ns. Hence, the multicultural and the egalitarian subscales both assess beliefs that involve the motivation to end discrimination.

We expected multiculturalism and egalitarianism to be differ- entially related to prejudices, and the relationship between multiculturalism and prejudices to be negative. We indeed found significant negative relations of prejudices with multiculturalism, but no significant relations with egalitarianism. Again, we tested for differential effects. Imposing equality constraints on the effects of multiculturalism and egalitarianism on prejudices decreased the model fit significantly, Dc2(1) ¼ 11.554, p < .001. Hence, the multi- culturalism subscale assesses a belief that is negatively related to prejudices, whereas the egalitarian subscale does not. Participants scoring high on multiculturalism opposed prejudiced statements significantly more strongly than did participants scoring high on egalitarianism.

5.5.2.2. Attitudes toward pluralism and acculturation. We expected both multiculturalism and egalitarianism to be positively related to

the appreciation of pluralism. Indeed, our results showed positive correlations of the two scales with the PADAA. The more strongly participants endorsed multiculturalism or egalitarianism, the more strongly they supported pluralism.

At the same time, given its emphasis on appreciating diversity, we expected multiculturalism to show a stronger relationship with pluralism than egalitarianism, which is more concerned with equal rights than with the benefits of diversity. Consistent with this expectation, model fit decreased significantly when we imposed equality constraints on the effects of multiculturalism and egalitari- anism on the PADAA, Dc2(1) ¼ 5.826, p < .05. As expected, the rela- tionship was stronger for multiculturalism than for egalitarianism.

For attitudes toward acculturation, we hypothesized a positive relationship with multiculturalism but not with egalitarianism. As expected, multiculturalism was significantly and positively related to acculturation, whereas the results for egalitarianism were not significant. The difference between multiculturalism and egalitar- ianism was significant. In line with our expectations, model fit decreased significantly when equality constraints were imposed on the effects of multiculturalism and egalitarianism on the Accul- turation scale, Dc2(1) ¼ 10.875, p < .001.

5.5.2.3. Authoritarianism. Research has shown that proponents of multiculturalism are more appreciative and tolerant of differences than are proponents of egalitarianism. Consequently, we expected authoritarianism to be negatively related to multiculturalism. Indeed, results showed a significant negative correlation between authoritarianism and multiculturalism, but not egalitarianism. Imposing equality constraints decreased the model fit significantly, Dc2(1) ¼ 4.219, p < .05. We can therefore conclude that endorse- ment of multiculturalism is significantly associated with lower authoritarianism.

5.6. Discussion

In sum, the results of study 2 were consistent with our expec- tations that the multiculturalism and egalitarianism subscales of the TCBS measure two beliefs about cultural diversity that share a motivation to control prejudiced behavior but that are distinctly related to beliefs about pluralism and acculturation, prejudices toward immigrants, and authoritarianism.

Both subscales were related to openness to pluralism in schools and to a higher motivation to control prejudiced behavior. Partici- pants with multicultural and egalitarian beliefs shared the moti- vation to control prejudiced behavior, but appreciation of pluralism was significantly more strongly related to multiculturalism than to egalitarianism. We also found differential relationships between the two subscales and several other outcomes: prejudices, attitudes about acculturation, and authoritarianism. Multiculturalism was associated with all three outcomes, whereas egalitarianism was not, with proponents of multiculturalism being less likely to agree with prejudiced statements, having more integrative views on acculturation, and showing a lower tendency toward an authori- tarian teaching style.

These results suggest thatdalthough often promoted as the preferable strategydegalitarian beliefs may in fact be the less effective response to culturally diverse classrooms.

6. General discussion

The aim of the present studies was to establish the newly developed Teacher Cultural Beliefs Scale (TCBS) for use in future research. To this end, we tested and validated the TCBS in two different samples. The TCBS was designed to measure two beliefs about how to deal with cultural diversity in the classroom.

Table 5 Standardized path coefficients of multiculturalism and egalitarianism on control and avoidance of prejudices, attitudes toward diversity, and authoritarian teaching style.

R2 Multiculturalism Egalitarianism

Estimate SE t Estimate SE t

Control and avoidance of prejudices MCPRa 12.7% .224 .057 3.933* .190 .058 3.272* Prejudicesa 15.6% �.368 .054 �6.795* �.051 .057 �0.893

Attitudes toward diversity PADAA 26.3% .389 .050 7.741* .196 .053 3.782* ACS 10.3% .321 .057 5.662* 0 .059 �0.003

Authoritarian teaching style AUTa 7.1% �.242 .059 �4.132* �.044 .060 �0.734

MCPR: Motivation to Control Prejudiced Reactions scale (behavior control subscale); prejudices: German General Social Survey; PADAA: Pluralism and Diversity Attitude Assessment; ACS: Acculturation scale; AUT: authoritarianism.

a Scale ranged from 1 to 5; all other scales ranged from 1 to 6. N ¼ 339, *p < .001 (two-tailed).

1 We re-ran the analyses using z transformation of multiculturalism and egali- tarianism. The pattern of results remained qualitatively similar; hence, we report only the results for the unstandardized paths coefficients.

A. Hachfeld et al. / Teaching and Teacher Education 27 (2011) 986e996 993

Author's personal copy

Multiculturalism, on the one hand, combines a general appreciation of cultural differences with the belief that such differences should be accommodated in teaching practice. Egalitarianism, on the other hand, emphasizes cultural similarities and stresses the importance of treating all students equally, regardless of their cultural background.

Study 1 confirmed the hypothesized two-factor structure of the scale and showed that both beliefs can be reliably assesseddtwo necessary prerequisites for the implementation of the TCBS in future research. Tests of measurement invariance confirmed both subscales to have equal factor structure across beginning teachers in different cohorts and school tracks. Nonetheless, we found mean differences in the endorsement of multiculturalism and egalitarianism between participants with differing experience of cultural diversity (in terms of school track and immigrant background). In Germany, most vocational-track schools have a higher proportion of immigrant students than do college-track schools, meaning that teachers in vocational-track schools experience more cultural diversity in their everyday school practice than do teachers in college-track schools. Our results show that beginning teachers in vocational-track schools endorsed multiculturalism significantly more strongly than did beginning teachers in college-track schools. Similarly, participants with immigrant backgrounds endorsed multiculturalism more strongly than did participants without immigrant backgrounds.

The restriction of the study 1 sample to beginning teachers in the teaching placement phase is certainly a limitation of the study. Further research is needed to test the factorial structure of the TCBS in samples of teachers with more teaching experience. Another limitation is the use of school track as a proxy for experience of cultural diversity. Future studies should aim at assessing personal experience of cultural diversity more directly and in more detail to investigate how this experience influences cultural beliefs. Finally, further research should consider how multiculturalism and egali- tarianism relate to negative beliefs about cultural diversity (e.g., assimilationist beliefs). A further limitation is the quantitative nature of the study, which we discuss below.

In study 2, we investigated the construct validity of the two subscales by investigating their correlations with (a) prejudices and the motivation to control prejudiced behavior, (b) attitudes toward pluralism and acculturation, and (c) authoritarianism. Both subscales theoretically imply favorable attitudes toward immigrant students, which should be reflected in a motivation to control prejudiced behavior and a general appreciation of pluralism. However, previous research in non-educational contexts has shown that the two beliefs may differentially affect prejudices. For example, focusing on equality and trying to ignore cultural differ- ences can lead to racial bias and to negative interracial interactions (Richeson & Nussbaum, 2004; Vorauer et al., 2009; Wolsko et al., 2000). In line with these findings, the results of study 2 showed that multiculturalism and egalitarianism were equally related to the motivation to control prejudiced behavior but not to opposition to prejudiced statements. Multiculturalism was negatively related to prejudiced statements, whereas egalitarianism was not related to this scale at all. This finding again shows that motivation to control prejudiced behavior is not necessarily synonymous with not harboring prejudicial thoughts. Although our results do not warrant the conclusion that focusing on equality can lead to prej- udice, they show that an egalitarian perspectiveddespite its correlation with the motivation to control prejudicesdis not at odds with prejudicial thoughts. Multiculturalism, on the other hand, was strongly negatively related to prejudicial thoughts. In other words, holding multicultural beliefs in fact means disagree- ing with prejudicial sentiments, whereas holding egalitarian beliefs solely reflects a will to control them, but is unrelated to their content.

The results showed positive and significant correlations of pluralism with both multiculturalism and egalitarianism. As ex- pected, however, the relationship was significantly stronger for multiculturalism than for egalitarianism. Our results for views on acculturation were similar. Theoretically, the two beliefs entail differing views on how acculturation should take place. The data confirmed that, relative to proponents of egalitarianism, propo- nents of multiculturalism were more strongly in favor of the inte- gration of minority cultures into mainstream society.

We also investigated correlations with authoritarianism. Previous research has shown that authoritarianism stands in fundamental contrast to an open and tolerant attitude (Zick & Six, 1997). Because multiculturalism is defined as open and tolerant, we expected it to be negatively related to authoritarianism, and our results confirmed this hypothesis. Egalitarianism showed no rela- tionship with authoritarianism, again confirming that an egali- tarian perspective, unlike multicultural beliefs, is not at odds with authoritarianism.

Further research is needed to investigate how different beliefs translate into teaching practices. The finding that multicultural and egalitarian beliefs are differentially related to authoritarian views on teaching gives a first hint of their importance for teaching.

We acknowledge that teachers’ beliefs about diversity can be manifold and that assessing only multicultural and egalitarian beliefs cannot provide a full picture. Qualitative approaches may provide deeper insights into the complexity and interrelatedness of beliefs about cultural diversitydfor example, by exploring how beliefs are constructed in discourse and interaction. As discussed above, however, previous social-psychological research demon- strating the relevance of multicultural and egalitarian beliefs for interethnic interaction has implications for the educational context. We therefore drew on instruments that have been successfully applied in quantitative social-psychological research to develop the TCBS. Additionally, our aim was to provide a validated instrument that can be used in future studies with larger samples. The present studies were not conducted to investigate how multicultural and egalitarian beliefs develop or how they translate into behavior in certain situations. Rather, the aim was to validate the new scale for application in future research that might, for example, investigate the role of personal experiences or compare the beliefs of different groups of teachers, such as majority and minority teachers (Castro, 2010). To further the understanding of teachers’ belief systems in general and beliefs about cultural diversity in particular, future studies should incorporate both quantitative and qualitative approaches.

In sum, the results of our two studies suggest that multicultur- alism and egalitarianism are two distinct and fundamentally different approaches to dealing with cultural diversity and to accommodating immigrant students in the classroom. In investigating teachers’ multicultural and egalitarian beliefs, we have crossed the border between social-psychological and educational research. The two beliefs have previously been studied primarily in the field of social psychology; instruments assessing the multicultural and egalitarian beliefs of teachers, in particular, were lacking. Our studies thus represent the first necessary steps toward research assessing multi- cultural and egalitarian beliefs in the educational context. With increasing diversity in schools around the globe, teachers in Western societies need to be prepared to teach in culturally heterogeneous schools and to challenge their own beliefs about cultural diversity. Our findings provide support for the idea that teachers’ cultural beliefs directly impact their teaching practices. However, research on the implications of multicultural and egalitarian beliefs in the educational context and research on how teachers’ beliefs manifest themselves in everyday classroom behavior is still scarce or nonex- istent. First evidence for the importance of cultural beliefs comes

A. Hachfeld et al. / Teaching and Teacher Education 27 (2011) 986e996994

Author's personal copy

from the employment context. Plaut and colleagues found that White co-workers’ multiculturalism positively predicted their minority co- workers’ psychological engagement (Plaut et al., 2009). In a similar vein, educational research should probe for differential effects of teachers’ multicultural and egalitarian beliefs on immigrant students’ engagement, learning progress, and school achievement. The newly developed TCBS provides a useful tool for assessing teachers’ multi- cultural and egalitarian beliefs. Future research should investigate whether favorable attitudes toward immigrant students are suffi- cient to create anatmosphere inwhich students from all backgrounds can excel or whether multicultural and egalitarian beliefs are asso- ciated with distinct strategies that are differentially beneficial for immigrant students. The ultimate aim must be to improve the learning outcomes of immigrant students and thereby reduce the achievement gap between students with and without an immigrant backgroundda challenge that needs to be addressed by the majority of the OECD countries (see Stanat & Christensen, 2006).

Acknowledgement

The COACTIV-R research project at the Max Planck Institute for Human Development was funded by the Max Planck Society’s Strategic Innovation Fund (2008e2010).

The authors thank Bernadette Park and Chick Judd for helpful discussions on multiculturalism and color-blindness, and for their help with adapting the constructs for the educational context. Many thanks go to Susannah Goss for her editorial assistance.

References

Amodeo, L. B., & Martin, J. (1982). A study of the effects of multicultural training on the factual knowledge and stereotypic attitudes of elementary and secondary teachers. Paper presented at the annual meeting of the American Educational Research Association, New York, NY.

Banse, R., & Gawronski, B. (2003). Die Skala Motivation zu vorurteilsfreiem Ver- halten: Psychometrische Eigenschaften und Validität. [The scale motivation to act without prejudice: psychometric properties and validity]. Diagnostica, 49, 4e13. doi:10.1026//0012-1924.49.1.4.

Barry, N. H., & Lechner, J. V. (1995). Preservice teachers’ attitudes about and awareness of multicultural teaching and learning. Teaching and Teacher Education, 11, 149e161. doi:10.1016/0742-051X(94)00018-2.

Bender-Szymanski, D. (2000). Learning through cultural conflict? A longitudinal analysis of German teachers’ strategies for coping with cultural diversity at school. European Journal of Teacher Education, 23, 229e250. doi:10.1080/ 02619760120049120.

Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York, NY: Guilford.

Burkard, A. W., & Knox, S. (2004). Effect of therapist color-blindness on empathy and attributions in cross-cultural counseling. Journal of Counseling Psychology, 51, 387e397. doi:10.1037/0022-0167.51.4.387.

Calderhead, J. (1996). Teachers: beliefs and knowledge. In D. C. Berliner, & R. C. Calfee (Eds.), Handbook of educational psychology (pp. 709e725). New York, NY: Macmillan.

Cardona Moltó, M. C., Florian, L., Rouse, M., & Stough, L. M. (2010). Attitudes to diversity: a cross-cultural study of education students in Spain, England and the United States. European Journal of Teacher Education, 33, 245e264. doi:10.1080/ 02619768.2010.495771.

Castro, A. J. (2010). Themes in the research on preservice teachers’ views of cultural diversity: implications for researching millennial preservice teachers. Educa- tional Researcher, 39, 198e210. doi:10.3102/0013189X10363819.

Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modelling, 9, 233e255.

Cochran-Smith, M., & Zeichner, K. M. (2005). Studying teacher education. The report of the AERA panel on research and teacher education. Mahwah, N.J.: Lawrence Erlbaum.

Correll, J., Park, B., Judd, C. M., & Wittenbrink, B. (2002). The police officer’s dilemma: using ethnicity to disambiguate potentially threatening individuals. Journal of Personality & Social Psychology, 83, 1314e1329. doi:10.1037/0022- 3514.83.6.1314.

Dubberke, T., Kunter, M., McElvany, N., Brunner, M., & Baumert, J. (2008). Lern- theoretische Überzeugungen von Mathematiklehrkräften: Einflüsse auf die Unterrichtsgestaltung und den Lernerfolg von Schülerinnen und Schülern. [Mathematics teachers’ beliefs and their impact on instructional quality and student achievement]. Zeitschrift für Pädagogische Psychologie, 22, 193e206. doi:10.1024/1010-0652.22.34.193.

Easter, L. M., Shultz, E. L., Neyhart, T. K., & Reck, U. M. (1999). Weighty perceptions: a study of the attitudes and beliefs of preservice teacher education students regarding diversity and urban education. Urban Review, 31, 205e220. doi:10.1023/A:1023287608741.

Ganter, S. (2001). Zu subtil? Eine empirische Überprüfung neuerer Indikatoren zur Analyse interethnischer Beziehungen. [Too subtle? An empirical test of new indicators for the analysis of inter-ethnic relations]. Kölner Zeitschrift für Soziologie und Sozialpsychologie, 53, 111e135. doi:10.1007/s11577-001- 0006-5.

Graham, J. W., & Hoffer, S. M. (2000). Multiple imputation in multivariate research. In T. D. Little, K. U. Schnabel, & J. Baumert (Eds.), Modelling longitudinal and multilevel data: Practical issues, applied approaches, and specific examples (pp. 201e218). Mahwah, NJ: Lawrence Erlbaum.

Harrington, H. L., & Hathaway, R. S. (1995). Illuminating beliefs about diversity. Journal of Teacher Education, 46, 275e284. doi:10.1177/0022487195046004006.

Hu, L.-T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Structured Equation Modeling, 6, 1e55. doi:10.1080/10705519909540118.

Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. J. (2003). Political conservatism as motivated social cognition. Psychological Bulletin, 129, 339e375.

Markus, H. R., Steele, C. M., & Steele, D. M. (2002). Color blindness as a barrier to inclusion: assimilation and nonimmigrant minorities. In R. Shweder, M. Minow, & H. R. Markus (Eds.), Engaging cultural differences. The multicultural challenge in liberal democracies. New York, NY: Russell Sage Foundation.

Muthén, L. K., & Muthén, B. O. (1998e2006). Mplus: The comprehensive modeling program for applied researchers. User’s guide (4th ed.). Los Angeles, CA: Muthén & Muthén.

Oser, F. K., Achtenhagen, F., & Renold, U. (2006). Competence oriented teacher training: Old research demands and new pathways. Rotterdam, The Netherlands: Sense Publishers.

Park, B., & Judd, C. M. (2005). Rethinking the link between categorization and prejudice within the social cognition perspective. Personality and Social Psychology Review, 9, 108e130. doi:10.1207/s15327957pspr0902_2.

Petzel, T., Wagner, U., Nicolai, K., & van Dick, R. (1997). Ein kurzes Instrument zur Messung der Autoritarismus-Neigung. [A short instrument to measure authoritarian tendency]. Gruppendynamik, 28, 251e258.

Petzel, T., Wagner, U., van Dick, R., Stellmacher, J., & Lenke, S. (1997). Der Einfluss autoritaristischer Einstellungen von Lehrerinnen und Lehrern auf ihr Verhalten in konflikthaften interkulturellen Situationen in der Schule. [The impact of authoritarianism of schoolteachers on their dealing with problematic inter- cultural situations]. Gruppendynamik, 28, 291e303.

Plaut, V. C. (2010). Diversity science: who needs it? Psychological Inquiry, 21, 168e174. doi:10.1080/1047840X.2010.492753.

Plaut, V. C., Thomas, K. M., & Goren, M. J. (2009). Is multiculturalism or color blindness better for minorities? Psychological Science, 20, 444e446. doi:10.1111/ j.1467-9280.2009.02318.x.

Pohan, C. A., & Aguilar, T. E. (2001). Measuring educators’ beliefs about diversity in personal and professional contexts. American Educational Research Journal, 38, 159e182. doi:10.3102/00028312038001159.

Richeson, J. A., & Nussbaum, R. J. (2004). The impact of multiculturalism versus color-blindness on racial bias. Journal of Experimental Social Psychology, 40, 417e423. doi:10.1016/j.jesp.2003.09.002.

Schafer, J. L., & Graham, J. W. (2002). Missing data: our view of the state of the art. Psychological Methods, 7, 147e177. doi:10.1037/1082-989X.7.2.147.

Schofield, J. W. (Ed.). (2006). Migration background, minority-group membership, and academic achievement research. Evidence from social, educational, and develop- mental psychology. Berlin, Germany: Social Science Research Center.

Stanat, P., & Christensen, G. (2006). Where immigrant students succeed. A comparative review of performance and engagement in PISA 2003. Paris: OECD.

Stanley, L. S. (1996). The development and validation of an instrument to assess attitudes toward cultural diversity and pluralism among preservice physical educators. Educational and Psychological Measurement, 56, 891e897. doi:10.1177/ 0013164496056005017.

Staub, F. C., & Stern, E. (2002). The nature of teachers’ pedagogical content beliefs matters for students’ achievement gains: quasi-experimental evidence from elementary mathematics. Journal of Educational Psychology, 94, 344e355. doi:10.1037/0022-0663.94.2.344.

Stipek, D. J., Givvin, K. B., Salmon, J. M., & MacGyvers, V. L. (2001). Teachers’ beliefs and practices related to mathematics instruction. Teaching and Teacher Educa- tion, 17, 213e226.

Tabachnick, B. R., & Zeichner, K. M. (1984). The impact of the student teaching experience on the development of teacher perspectives. Journal of Teacher Education, 35, 28e36. doi:10.1177/002248718403500608.

UNESCO. (2004). Changing teaching practices: Using curriculum differentiation to respond to students’ diversity. Paris: UNESCO.

van de Vijver, F. J. R., Breugelmans, S. M., & Schalk-Soekar, S. R. G. (2008). Multi- culturalism: construct validity and stability. International Journal of Intercultural Relations, 32, 93e104. doi:10.1016/j.ijintrel.2007.11.001.

van Dick, R., Wagner, U., Adams, C., & Petzel, T. (1997). Einstellungen zur Akkultu- ration: Erste Evaluation eines Fragebogens an sechs deutschen Stichproben. [Attitudes toward acculturation: first evaluation of an acculturation scale in six German samples]. Gruppendynamik, 28, 83e92.

Vedder, P., Horenczyk, G., Liebkind, K., & Nickmans, G. (2006). Ethno-culturally diverse education settings: problems, challenges and solutions. Educational Research Review, 1, 157e168. doi:10.1016/j.edurev.2006.08.007.

A. Hachfeld et al. / Teaching and Teacher Education 27 (2011) 986e996 995

Author's personal copy

Verkuyten, M. (2005). Ethnic group identification and group evaluation among minority and majority groups: testing the multiculturalism hypothesis. Jour- nal of Personality and Social Psychology, 88, 121e138. doi:10.1037/0022- 3514.88.1.121.

Vorauer, J. D., Gagnon, A., & Sasaki, S. (2009). Salient intergroup ideology and intergroup interaction. Psychological Science, 20, 838e845. doi:10.1111/j.1467- 9280.2009.02369.x.

Wolsko, C., Park, B., & Judd, C. M. (2002). The measurement and consequences of interethnic ideology. Boulder, CO: University of Colorado.

Wolsko, C., Park, B., & Judd, C. M. (2006). Considering the tower of Babel: correlates of assimilation and multiculturalism among ethnic minority and majority groups in the United States. Social Justice Research, 19, 277e306. doi:10.1007/ s11211-006-0014-8.

Wolsko, C., Park, B., Judd, C. M., & Wittenbrink, B. (2000). Framing interethnic ideology: effects of multicultural and color-blind perspectives on judgments of groups and individuals. Journal of Personality and Social Psychology, 78, 635e654. doi:10.1037/0022-3514.78.4.635.

Woolley, S. L., Benjamin, W.-J. J., & Woolley, A. W. (2004). Construct validity of a self- report measure of teacher beliefs related to constructivist and traditional approaches to teaching and learning. Educational and Psychological Measure- ment, 64, 319e331. doi:10.1177/0013164403261189.

Zeichner, K. M. (1992). Educating teachers for cultural diversity (NCRTL special report no. 41). Michigan State University, National Center for Research on Teacher Learning.

Zick, A., & Six, B. (1997). Autoritarismus, Vorurteile und Einstellungen zur Akkul- turation. [Authoritarianism, prejudices, and attitudes toward acculturation]. Gruppendynamik, 28, 305e320.

A. Hachfeld et al. / Teaching and Teacher Education 27 (2011) 986e996996