Learning to Teach for Social Justice as a Cross Cultural Concept: Findings from Three Countries

All over the world, countries are paying close attention to how teachers are recruited, selected, and prepared for the nation’s schools. Increasingly, teachers are expected to teach all students to high standards at the same time that they play a major role in meeting rising expectations regarding social equity. Preparing teachers for these challenges is among the most pressing and complex tasks in teacher education. In response to these and other challenges, some initial teacher education programs now include among their major goals preparing teachers to teach for social justice, work toward equity and access for all students, and/or challenge inequities in existing educational systems and policies. This article focuses on three initial teacher education pro-grams—one each in the United States, New Zealand, and Ireland. Although these programs differ from one another in many ways, they also share some goals related to teaching for social justice and equity. The article examines longitudinal survey data regarding teacher candidates’ scores on the “Learning to Teach for Social Justice-Beliefs” scale, which was designed to measure candidates’ endorsement of beliefs consistent with the concept of teaching for social justice. For each of the three research sites, the article analyzes: (a) demographic and teacher quality contexts, (b) initial teacher education program goals related to social justice/social equity, and (c) the results of surveys administered to teacher candidates at entry to and exit from the programs. The article concludes with discussion of learning to teach for social justice as a cross-cultural concept.

depends on the quality of teachers, is to produce a workforce that can meet the demands of the competitive global market.
Although widely shared, this point of view has also been questioned. Furlong and colleagues (Furlong, et al, 2009), for example, point out that critics in a number of countries have questioned the idea that education alone determines the economic health of a nation and challenged the assumption that teachers are the over-arching factor in improving educational outcomes. Similarly Spring (2010) has questioned the basic logic of the human capital argument itself, pointing out that it could be that economic growth makes it possible for nations to provide education to more people rather than the reverse-that growth in school attendance leads to greater economic growth. Despite important critiques of the human capital argument, however, in short and globally, teachers have been identified as one of the major determinants, if not the key factor, in the quality of education, which in turn is assumed to be tied to the economic health of nations (OECD, 2005). Schooling-and especially teachers-are seen as having a key role in helping nation states respond to rising social expectations and achieve greater equity.
At the same time that teacher quality has become the focus of attention at the highest levels of government, a second major global trend is shaping initial teacher education as well as how teaching and learning are structured in the schools. In many nations, there is increasing diversity in the school population as well as increasing recognition of the challenges diversity poses (Banks, 2009a;Castles, 2009;OECD, 2006). Although the situation has changed in some countries since the global economic recession that began in 2008, many nations have experienced changes in migration flow over the last two decades, with the result that in a number of countries, including the three that are the focus of this study, 1 the total number of people entering the country has far exceeded the number leaving. Countries in this category include, but are not limited to the United States, Canada, England, Scotland, most of the countries in western and northern Europe, Australia, New Zealand, and Singapore (OECD, 2006). For example, Spain shifted from a sending to a receiving country during this time period, attracting immigrants from Northern Africa, Latin America, and other European Union countries. Migration patterns in England, Scotland, and Ireland shifted when they opened the job market up to workers from European Union countries, allowing many new workers from central and eastern Europe. Norway's immigration rate reached its highest levels ever in 2011 with newcomers from Poland, Lithuania and other nations. In many countries, changing migration patterns have meant an influx of students whose first language is not the language of instruction in the schools and/or whose cultural and experiential backgrounds differ significantly from those of the majority of the nation's students and teachers. In some nations, including the U.S. and New Zealand, new immigration patterns have added further complexity to the historical marginalization or colonization of racial groups. In the U.S., this includes formerly enslaved and/or indigenous minorities, whose rights have been foregrounded since the Civil Rights and Indian Rights movements of the 1960s and 1970s (Banks, 2009b). In N.Z., this includes renewed commitment to, and enactment of Maori rights, stipulated in the 1840 Treaty of Waitangi between the Maori and British colonizers, but foregrounded in response to protest and social pressure in the 1970s (Waitangi Tribunal, 2011).
Although global migration has decreased somewhat as a result of the recession, settled immigrants have been disproportionately affected by economic hardship in many of their host countries. For example, after an economic boom turned Ireland into a destination country by the early 2000s, the recession significantly reduced the total number of jobs available, leading to rising unemployment in immigrant communities in particular. According to a recent report, Ireland now has a 3% differential in unemployment rates between migrant and native populations. Along similar lines, Spain has the secondhighest unemployment differential rate between migrant and native populations in the world, at 5.4% (Kozer & Laczko, 2010).
Even in some countries that have long been considered homogeneous in language, ethnicity, culture, or religion, immigration patterns during the 1990s and 2000s changed dramatically (Banks, 2009a;Castles, 2009). In Japan, for example, although the number of immigrants is far smaller than the number of immigrants to the U.S., Canada, or to western and northern European countries, the direction of the current trend is the same as the one noted above, with many Japanese returnees as well as newcomers from African and South American countries coming into the country (Hirasawa, 2009). Norway has become a popular destination country for Arabic, Somali and Urdu refugees due, in part, to the country's protections of these people (Beckmann-Dierkes & Fuhrmann, 2011). Along different lines, in Ireland prior to the recession, there were large numbers of immigrants primarily from Eastern Europe and Africa who came to Ireland to take advantage of the dramatic gains in standard of living and employment opportunities produced by the "Celtic Tiger" (Corona, 2010). Globally, new patterns of migration have heightened awareness of the educational challenges posed by diversity and of the inequities in achievement and other school-related outcomes that persist between majority and minority groups in many nations (May, 2009).
During the last decade of the 20 th century and the first decade of the 21 st , these two trendsunprecedented emphasis on teachers as the key factor in educational quality and in the achievement of greater social expectations, on the one hand, and heightened attention to the increasing diversity of school populations and to the inequities that often exist in educational opportunities and outcomes between majority and minority populations, on the other hand-intensified and converged. The result is that in many nations, teachers are now expected to play a major role in meeting the challenges of a diverse globalized society by ensuring that all school students have both rich learning opportunities and equitable learning outcomes (OECD, 2010). Accordingly, teacher education programs and pathways are expected to produce teachers who are ready, willing and able to take on these tasks and who are informed about the needs and strengths of diverse school populations.

Teaching for Social Justice as a Teacher Education Outcome
Many initial teacher education programs have addressed the issues outlined above by emphasizing the roles of teachers individually and collectively in efforts to accomplish greater social equity and academic achievement for all school students. In some places, this has involved identifying "teaching for social justice/social equity" as an important and measurable goal of initial teacher education.

Defining Teaching for Social Justice
In an effort to theorize teacher education for social justice, Cochran-Smith (2010) notes that in contemporary political philosophy, there have been debates about whether social justice is a matter of distribution or recognition (see, for example, Fraser & Honneth, 2003;North, 2006). The distributive paradigm of justice focuses on equality of individuals, civic engagement, and a common political commitment to all citizens' autonomy to pursue their own ideas of the good life (Rawls, 1971). From this perspective, which was prevalent during the last half of the 20 th century, the remedy for injustice is redistribution of material and other goods, including opportunity, power and access with the goal of establishing a society based on fairness and equality (Fraser, 2003). Despite the deepening socioeconomic inequalities in contemporary society, however, many political philosophers argue that focusing solely on equality and distribution of goods is theoretically inadequate to the task of conceptualizing justice in today's diverse society. As Young (1990) points out, the social movements of the 1960s and 1970s made it clear that failure to recognize and respect social groups was a central dimension of injustice, and thus the goal of recognition had to be central to justice theories.
Most current conceptions of social justice acknowledge its plural dimensions (Gerwitz & Cribb, 2002), and the most important question now is not whether justice is a matter of distribution or recognition, but rather how to conceptualize the relationship between the notion of distributive justice that is central to modern liberal democracies, on one hand, and, on the other hand, contemporary struggles for the recognition of social groups in relation to the politics of identity and difference (Fraser & Honneth, 2003;North, 2006).
At a general level, perspectives from political philosophy are instructive for theorizing teacher education for social justice, even though none of them addresses teaching or teacher education directly. Cochran-Smith (2010) suggests that a theory of justice for teacher education must connect the key ideas of distributive justice, which locates equality and autonomy at the center of democratic societies (Howe, 1997(Howe, , 1998, with current political struggles for recognition, which challenge school and knowledge structures that reinforce the oppression of particular social groups and ignore the knowledge traditions of marginalized groups (King, 2006;Young, 1990).
The meaning of social justice in teacher education varies and has generally been under-theorized (MacDonald & Zeichner, 2009). However, there is a certain degree of consistency (Sleeter, 2009) about the idea, and we found that there were goals related to social justice as both redistribution and recognition reflected in the published literature on this topic (e.g., Adams, Bell & Griffin, 1997;Ayers, Hunt & Quinn, 1998;Cochran-Smith, 1999Darling-Hammond, French, Garcia-Lopez, 2002;Mac-Donald & Zeichner, 2009;Michelli & Keiser, 2005;Oakes & Lipton, 1999;Sleeter, 2009;Villegas & Lucas, 2001;Zeichner, 1993). Generally, proponents of teacher education for social justice argue that: (1) There are significant disparities in the distribution of educational and other resources and opportunities between minority and/or low-income students and their White, middle-class counterparts.
(2) In addition, long-standing policies, practices, and systemic structures-including traditional curricula and school norms (King, 2006(King, , 2008 as well as larger social policies related to health care, employment, and transportation (Anyon, 2003;Lipman, 2003)-privilege dominant groups and disadvantage others. (3) Inequities in opportunities and outcomes as well as lack of recognition of the knowledge traditions of minority groups run counter to the democratic ideal, which depends on widespread civic participation and deliberation about diverse perspectives (Howe, 1997;Ladson-Billings, 1995;Reich, 2002). (4) Thus, part of the job of teaching is enhancing students' learning and life chances by allying with others to challenge school and societal inequities (Cochran-Smith, 1995, building on the cultural and linguistic resources students bring to school to broaden the curriculum and build new knowledge (e.g., Brisk, 2007;Ladson-Billings, 1995;Lee & Ball, 2005;King, 2006), and understanding schools and schooling in terms of issues related to knowledge, politics and power (e.g., Gadsden, Davis & Artiles, 2009;Moll, 2009). Following this line of reasoning, part of the job of teacher education is preparing teachers who are committed to, and know how to, teach for social justice as well as assessing teacher candidates' progress in learning to teach in terms of their knowledge, beliefs, and skills in these areas.

Measuring Learning to Teach for Social Justice
As part of a larger initiative to improve university-sponsored teacher education in the United States, 2 the Boston College (BC) Evidence Team developed a "Learning to Teach for Social Justice-Beliefs" scale (LTSJ-B), which was embedded in a suite of five surveys administered to teacher candidates at entry and exit from the initial teacher education program and then one, two and three years out of the program (Ludlow, Enterline & Cochran-Smith, 2008a;Ludlow, Pedulla, Enterline, Cochran-Smith, Loftus, Salomon-Fernandez & Mitescu, 2008b). The scale is based on a Rasch measurement framework wherein the basic premise is that differences in the degree to which people accept, believe in, and feel they are prepared to teach in ways consistent with social justice principles are measurable. Furthermore, these differences may be understood and represented as a one-dimensional continuum along which people may be located based on their degree of endorsement of the social justice principles.
The LTSJ-B scale. It is important to note that the LTSJ-B scale is not intended to account for the entire complex idea of learning to teach for social justice as a goal and outcome of teacher education. First, the scale represents only beliefs and perspectives and not classroom practice, relationships with students and families, content and pedagogical knowledge, advocacy and activism, or student learning outcomes, all of which are critical parts of the larger meaning of teaching for social justice. Thus the LTSJ-B scale tells only part of the story about learning to teach and should be understood in terms of its limited focus on beliefs. Second, there are many ways to conceptualize beliefs related to social justice, and this scale knowingly includes only a tiny portion of the possible range of items it might contain, an issue we return to in the concluding sections of this article.
The LTSJ-B items were developed to reflect the idea of teachers as classroom and societal advocates for change and to encompass a number of key ideas about justice as both distribution of learning opportunities and outcomes, on one hand, and recognition of the knowledge traditions, strengths and assets that all students bring to school, on the other. The key ideas behind the LTSJ-B scale include: high expectations and rich learning opportunities for all students; an asset-based perspective on the cultural, linguistic and experiential resources students and families bring to school; the importance of critical thinking in democratic societies; the role of teachers as advocates and agents for change; challenges to the notion of a meritocratic society; teaching as an activity that is related to teachers' deep underlying assumptions and beliefs about race, class, gender, disability and culture; and the idea that issues related to culture, equity, and race ought to be part of what is openly discussed and visible in all aspects of the school curriculum.
To reflect these ideas, the 12 social justice beliefs items listed in Table 1 were selected following an extensive series of pilot tests. For each item, respondents answer using a 5-point rating scale (1=Strongly Disagree, 2=Disagree, 3=Uncertain, 4=Agree, and 5=Strongly Agree). Some items are positively worded (e.g., "Good teaching incorporates diverse cultures and experiences"), which means that a response of "Strongly Agree" corresponds to the strongest degree of endorsement of social justicerelated beliefs. Other items are negatively worded (e.g., "It's reasonable for teachers to have lower expectations for students who don't speak English as their first language"), which means that a response of "Strongly Disagree" corresponds to the strongest degree of endorsement of social justice beliefs. Based on the above conceptualization of beliefs related to teaching for social justice, candidates with a stronger commitment to social justice would be expected to agree with (or "positively endorse") items 1, 2, 4, 7, and 8 on the scale and disagree with (or "negatively endorse") items 3R, 5R, 6R, 9R, 10R, 11R, and 12R. Items that respondents are expected to disagree with are reverse-scored ["R"] so that higher total scores correspond to stronger beliefs in principles related to teaching for social justice.
Previous analyses using the LTSJ-B scale. This article presents findings from the third in a sequence of studies that assume that teaching for social justice is a legitimate and measurable outcome of teacher education. The objective of this line of research over a span of seven years has been to develop a measure of social justice beliefs that is internally consistent, content-and construct-valid, sensitive to changes in beliefs, and independent of the specific characteristics of any single teacher preparation program. The purpose of this article is to present evidence about the extent to which teacher candidates in three different English-speaking international settings demonstrated growth (from program entry to exit) in their beliefs about teaching for social justice and also to identify and interpret the areas in which teacher candidates made greater and lesser gains. First we provide a brief review of the psychometric work to date.
The first study established the technical psychometric characteristics of the LTSJ-B scale (Ludlow, et al 2008a). As noted above, the fundamental working assumption was that even though people differ from one another in the degree to which they understand, accept, believe and are prepared to teach in ways consistent with social justice principles, these differences can, nonetheless, be measured and represented along a single ordered continuum of beliefs ranging from relatively weak to relatively strong levels of commitment to teaching for social justice. Furthermore, it was assumed that teacher candidates' beliefs and commitments to these principles develop and evolve over time.
In addition to standard exploratory factor analyses and reliability analyses, the first study included a Rasch rating scale analysis (Rasch, 1960;Wright & Masters, 1982). These complementary analyses confirmed that the 12 LTSJ-B items defined an ordered construct comprised of increasingly more controversial and debatable beliefs and ideas related to teaching for social justice (see Figure 1).

SJ1
An important part of learning to be a teacher is examining one's own attitudes and beliefs about race, class, gender, disabilities, and sexual orientation SJ2 Issues related to racism and inequity should be openly discussed in the classroom.

SJ3R
For the most part, covering multicultural topics is only relevant to certain subject areas, such as social studies and literature.

SJ4
Good teaching incorporates diverse cultures and experiences into classroom lessons and discussions.

SJ5R
The most important goal in working with immigrant children and English language learners is that they assimilate into American society.

SJ6R
It's reasonable for teachers to have lower classroom expectations for students who don't speak English as their first language.

SJ7
Part of the responsibilities of the teacher is to challenge school arrangements that maintain societal inequities.

SJ8
Teachers should teach students to think critically about government positions and actions.

SJ9R
Economically disadvantaged students have more to gain in schools because they bring less into the classroom. SJ10 R Although teachers have to appreciate diversity, it's not their job to change society.

SJ11
R Whether students succeed in school depends primarily on how hard they work.

SJ12
R Realistically, the job of a teacher is to prepare students for the lives they are likely to lead. a Likert response categories: Strongly Disagree=1, Disagree=2, Uncertain=3, Agree=4, Strongly Agree=5 b R: denotes the categories were reverse scored.
Note: Saint Patrick's College (SPC) and the University of Auckland (UA) made wording changes to the scale to reflect their unique circumstances. Item SJ5R "assimilate into American society", for example, was changed to "assimilate into Irish society" in SPC and "assimilate into New Zealand society" in UA. In addition, SPC included a definition of "assimilation" which may have reduced confusion about its intended meaning and thus made this item easier to endorse in the Irish context. The graphic portrayal, or variable/construct map, of the LTSJ-B scale in Figure 1 represents the structure of the learning to teach for social justice-beliefs construct. 3 That is, the variable map indicates: (a) the range and spread of LTSJ-B items from easiest-to-hardest endorse (from the bottom to the top of the scale on the right of the vertical line), (b) the range and spread of the candidate's scores (indicated to the left of the vertical line), and (c) what a candidate's score on the continuum means in terms of endorsing and not endorsing specific items (the items located below and above a candidate's location on the scale, respectively). The extent to which the LTSJ-B scale was sensitive enough to measure changes in beliefs over time, and maintained this specific ordering of item locations along the continuum for students in different teacher preparation programs (i.e. "construct invariance") became the focus of the second and third studies.
The second study included three separate analyses examining the extent to which social justicerelated beliefs and perspectives differed among teacher candidates at entry to, exit from, and one year out of BC's teacher preparation program, which has an explicit social justice agenda (Enterline, Cochran-Smith, Ludlow & Mitescu, 2008). The first analysis of the second study established the scale's construct invariance within multiple cohorts of entering and exiting teacher candidates, establishing that the scale's structure and meaning remained the same for different candidates at different time points. The second analysis found that the exiting candidates' scores exceeded the scores of the entering candidates, establishing the sensitivity and usefulness of the instrument as a way to measure changes in beliefs over time. The third analysis showed that after one year of teaching, teachers maintained their higher scores on the LTSJ-B scale, establishing the strength of the teachers' belief systems and indicating the durability of teachers' changed beliefs beyond the short term of the preparation period itself. The present study, the third in this line of research, builds on these two earlier investigations.

Learning to Teach for Social Justice in Three Countries: Longitudinal Design
In the remainder of this article we focus on the institutional context and findings from the administration of the LTSJ-B scale at entry to and exit from initial teacher education programs at universities in three countries: (1) undergraduate level initial teacher education programs at Boston College, located in Chestnut Hill, Massachusetts, which is part of the greater Boston area in the northeastern United States and is the tenth largest metropolitan area in the country; (2) graduate level initial teacher education programs at the University of Auckland, in Auckland, New Zealand, the largest city in New Zealand with a population of 1.3 million (nearly a third of the country's population); and, (3) the three-year B.Ed. diploma in elementary school teaching at St. Patrick's College, Dublin City University, Ireland, the country's capital and largest city, with a population of almost two million (40% of the country's total population).
In the following sections, we describe for each country national trends related to student diversity and to teacher quality/teacher preparation. Then we discuss the goals of the initial teacher education programs at each site in terms of social justice/social equity. Next we compare the results of multiple administrations of the LTSJ-B scale at the points of program entry and exit.

Research Site 1: Initial Teacher Education Programs at Boston College, U.S.A.
The first research site consists of the undergraduate four-year initial teacher education program at BC. This university has a long commitment to social justice at the institutional and program level.
Challenges of diversity in the U.S. In the U.S., scholars often use the phrase, the "demographic divide" (Gay & Howard, 2000), to describe the educational context and the challenges of diversity. The racial and ethnic characteristics of the school population in the U.S. have changed dramatically from 78% students White (that is, European American) and 22 % students of color (that is, African American, Hispanic, Asian, or indigenous Native American) in 1972, to 55% White students and 45% students of color in 2008 (National Center for Education Statistics, 2003Statistics, , 2010a. In addition, the number of multilingual learners increased from 3.8 million in 1979 to 10.9 million almost 20 years later (National Center for Education Statistics, 2010b). Further, the number of students with disabilities who receive special education services, often in regular education classrooms, increased from just over 4 million in 1981 to more than 6.5 million in 2008, which is more than 13% of the total school population (National Center for Education Statistics, 2009).

Note:
There are 110 graduating seniors, 12 items, and 5 Likert scoring categories. The intended directionality of the social justice responses is indicated by "SA-strongly agree" or "SD-strongly disagree". The item content is presented in abbreviated form. Each '#' represents 2 seniors.

Figure 1. 2005 Boston College Exit Variable Map
Coupled with the increasing diversity of the student population, in the U.S., there is also a widelyrecognized and enduring "achievement gap" among student groups that differ from one another racially, culturally, linguistically, socioeconomically and geographically (Villegas & Davis, 2008). White and Asian students consistently score higher than their African American and Hispanic counterparts on standardized achievement tests. Likewise White and Asian students have lower rates of dropping out of high school and higher rates of high school graduation than their African American and Hispanic counterparts.
In the U.S., although the student population has become increasingly diverse, the teacher population continues to be primarily White, European American, middle class, and monolingual (Villegas & Lucas, 2004). In addition, there is evidence that unless they are specifically prepared to do otherwise, many White teachers have difficulty functioning as role models for students of color (Goodwin, 2000), helping students bridge home-school differences (Gay, 2000), and constructing curriculum, instruction, and assessments that are culturally responsive (Ladson-Billings, 1999). Perhaps the most serious problem, many White middle-class teachers understand diversity as a deficit to be overcome and tend to have lower expectations for students of color (Irvine, 1990;Villegas & Lucas, 2004).
The Boston College Initial Teacher Education Program. Boston College, a Jesuit and Catholic university, is committed to nurturing its students' intellectual, personal, ethical, and religious formation, and to merging high academic achievement with service to others. The Lynch School of Education's (LSOE) mission is to pursue excellence and ethics in teaching, research, and service.
Initial teacher education programs at BC specifically aim to prepare teachers who teach for social justice by enhancing students' learning and their life chances in all schools and for all students, but with particular emphasis on meeting the needs of marginalized students in urban schools. Undergraduate teacher candidates are required to complete two majors-one in education and one in the Arts and Sciences or in an interdisciplinary area; all undergraduate students take the university core curriculum, which includes courses related to diversity and theology. In education courses, teacher candidates are exposed to a variety of theoretical and practical perspectives on teaching, learning and schooling within social, political, cultural and academic contexts as well as to a variety of teaching methods. Program requirements also include multiple fieldwork experiences prior to student teaching (at least two of which must be in urban or other contexts with high levels of student diversity) and seminars that help students critically reflect on and engage in inquiry about their own schools and classrooms.
Each of the program's multiple assessments includes evaluations of candidates' progress in various aspects of teaching for social justice, supporting the needs of diverse learners, and engaging in selfreflexive inquiry. In addition to completing the program's formal requirements, many teacher candidates also engage in a variety of BC service projects across the state, nation, and world. These allow students to immerse themselves in cultures different from their own and to serve populations other than those in the Boston community.

Research Site 2: Initial Teacher Education Programs at The University of Auckland, New Zealand
The second research site is the graduate initial teacher education program for elementary teaching at the University of Auckland, New Zealand. The program is committed to preparing teachers who can improve outcomes for all learners. This commitment comes from both within and beyond the institution itself.
Challenges of diversity in New Zealand. New Zealand is a bicultural nation, founded by the Treaty of Waitangi in 1840. This treaty, between indigenous Maori and British colonizers, was imperfect and led to many injustices and misunderstandings (Orange, 1987). However, since the 1970s, processes have been put into place that attempt to address these issues and to honor the treaty. These processes have significant implications for education.
Auckland is the largest Polynesian city in the world, with many people from different Pacific Island groups, such as Samoa and the Cook Islands, living there. In addition, in recent decades, immigration from Asia has further increased the diversity of the school population. One in five current New Zealand residents was not born in New Zealand (OECD, 2010). Among all first grade students across the country in 2010, 53% were New Zealand European, 25% were Maori, 11% were Pacific Island students, 9% were Asian, and 2% were students of other ethnicities (New Zealand Ministry of Education, 2010). Cultural and linguistic diversity is now a major feature of many New Zealand schools, particularly in Auckland.
In a similar pattern to that noted in our discussion of the U.S., there is a notable achievement gap in New Zealand as indicated by international survey results. While N.Z. students appear to do well overall, aggregate results mask one of the largest gaps between high and low achieving students among all OECD countries (New Zealand Ministry of Education, 2011). Maori and Pacific Island students are overrepresented in the low achieving group, while New Zealand European and Asian students are overrepresented in the high achieving group. The Ministry of Education's strategy for Maori Education, known as "Ka Hikitia" (New Zealand Ministry of Education, 2007), uses indicators such as participation in early childhood education, participation and retention rates across the schooling sectors, the attainment of basic school leaving qualifications and obtaining university entrance qualifications to monitor progress for Maori students. While steady progress is being made, neither Maori nor Pacific Island students share the success experienced by other student groups (Ministry of Education, 2007;Ministry of Education, 2009).
There are approximately 52,000 teachers in New Zealand. Seventy-five percent of them are New Zealand European, with 9% Maori and 3% Pasifika teachers. This leads to a mismatch similar to that described for the U.S., with mostly female, monolingual, middle class White teachers working with an increasingly diverse student population. The importance of cultural competence for New Zealand teachers has been highlighted by the "Te Kotahitanga" project, a large-scale professional development project, which has experienced success in improving Maori achievement by focusing on eliminating deficit thinking about students and developing teachers' sense of agency (Bishop, 2008).

University of Auckland Graduate Initial Teacher Education Program. The University of
Auckland's (UA) preparation program for graduate-level students is a one-year intensive course of study. Teacher candidates entering this program have undergraduate degrees, which can be in any discipline. They undertake a year's study of elementary teaching, which consists of six courses in elementary-school curriculum areas (for example, mathematics education, science education) and three courses in education and learning more generally. Despite being a compressed course, the teacher candidates undertake up to twelve weeks of practicum in three different schools. At least one of these placements must be in a diverse urban school. In addition they spend a day a week for the full year in a fourth school which gives them the opportunity to form deeper relationships and understandings.
While all the courses contain elements of teaching for social justice, one course in particular critically examines the context of education in New Zealand and provides an overview of the political, social and historical background for education and how this results in inequities. A second course looks specifically at the aspirations and cultural needs of Maori students and considers why learning Maori language and worldview is important for all learners. All graduates from teacher preparation programs in New Zealand must meet the Graduating Teacher Standards (New Zealand Teachers Council, 2007), which have been set by the regulating body for teachers in New Zealand. These standards emphasize the importance of "te reo me nga tikanga Maori" (Maori language and culture), cultural competence, the significance of relationships with learners, understanding the influence of the broader social milieu on teaching and learning, and the use of evidence of learning to drive planning and teaching. These emphases are reflected in the one-year preparation program, with teacher candidates compiling an electronic portfolio of their work with students, demonstrating their achievement of the Graduating Teacher Standards.

Research Site 3: Initial Teacher Education Programs at Saint Patrick's College, Dublin, Ireland
Saint Patrick's College is a college of Dublin City University. It is one of five state-funded colleges in the Republic of Ireland that provides initial teacher education programs for elementary school teachers.
Challenges of diversity in Ireland. Due to the economic success of the so-called "Celtic Tiger," society in Ireland has undergone a period of rapid change over the past decade. An important feature of that change has been the arrival of significant numbers of foreign nationals. Ireland has had a long history of emigration, but immigration is a relatively new phenomenon in the country and, despite the recent recession, diversity is likely to remain a feature of Irish society for the foreseeable future. It is now estimated that as many as 10% of students in Irish schools represent diverse countries, languages and cultures (Taguma, Moonhee, Wurzburg & Kelly, 2009). This is a significant change for Ireland, a country wherein for many years, "Travellers"-an indigenous minority ethnic group that comprises just 0.5% of the population-constituted the only ethnic minority group of any size in Ireland (Central Statistics Office, nd). New population patterns pose new challenges for schools and other institutions with little prior experience dealing with cultural and linguistic diversity on a larger scale. It is also the case that while the student population in Ireland is becoming increasingly diverse, the teaching population is almost 100% White and middle-class. As the decades pass and the children of first-generation immigrants come to maturity, attention will need to turn to ensuring that they are represented within the ranks of the teaching profession.
A number of recent studies conducted by the OECD (Perkins, Moran, Cosgrove & Shiel, 2010;Taguma et al., 2009) and the Economic and Social Research Institute (Smyth, Darmody, McGinnity & Byrne, 2009) have provided a snapshot of some of the key characteristics of Ireland's "newcomer" (the preferred term) school-going population at this time. It has been noted in these studies that the vast majority of Irish high schools have newcomer students while just under half of all elementary schools have newcomer children, with some newer schools having very high concentrations of this group. As is the case in most OECD countries, newcomer students in Ireland, on average, achieve education outcomes well below their local-born peers.
These studies also suggest that, if not addressed, the language needs of newcomer students will hinder their academic development and social integration. For example, most newcomers do not have English as a first language, and there is a gap in achievement between those who speak English at home and those who do not. There is also anecdotal evidence that in some schools the parent body is made up of as many as 40 nationalities, and dozens of languages are spoken in the school community.
Unfortunately, few teachers with newcomer students in their classes have received teacher preparation or professional development concerning teaching English as an additional language. And few schools have access to translation services in cases where newcomer parents do not speak English, which makes it difficult to involve them in their children's education. Although teachers generally view newcomer students as motivated and hard-working and believe that their families place a high value on education, elementary and high school principals note that Irish curricula and textbooks do not take adequate account of the diversity evident in classrooms (Smyth et al., 2009). Given rapid changes in immigration patterns, relations between newcomer and Irish students are generally positive but there has been some identification of "segregation in friendship patterns" and incidences of bullying (Smyth et al, 2009, p. 102). It has been argued that in a context of an ever-increasing national debt with the allied budgetary restrictions being implemented, avoiding a trade-off between the needs of newcomer and Irish students is paramount (Smyth et al., 2009).
A case in point is the Minister for Education's recent decision to reverse an earlier cut in support for disadvantaged primary schools (where most newcomer students are) but instead, to recoup the lost budgetary savings by decreasing all school capitation funding, which is used to cover the costs of utilities such as heating and electricity. The minister's actions were described by the director of the Irish Primary Principals' Network as "an attempt to rob Peter to pay Paul" (Caroll & Flynn, 2012). Saint Patrick's College Initial Teacher Education Program. Students entering St. Patrick's College (SPC) undergraduate initial teacher education program are among the highest achievers (top 20%) of the cohort leaving high schools in any given year. These students complete a three year Bachelors of Education program, which is designed to prepare them to teach the full range of areas specified in the state curriculum for Irish primary schools. The first year of the program comprises three majors-education and two humanities subjects (e.g. History, English). In the final two years, students study education and one subject from the humanities. 4 SPC's mission and institutional strategy statements reflect its commitment to social justice in an increasingly multi-cultural Ireland. For example, college documents state that the college will "instill in teachers an awareness of their role in conveying cultural, civic and social values and ensuring that they respond flexibly to the challenges of a changing Irish society" and that "justice and equity inform its research themes, its access and international strategies, and its foregrounding of special needs and disadvantage education" (St Patrick's College, 2009, pp. 5-6). This suggests that students are exposed to issues related to human rights, diversity, educational disadvantage and developmental education throughout their three years in the program. This is done informally within the many program modules and more formally in courses on educational disadvantage or elective subjects in diversity and global justice in the classroom.
Students also engage with various college organizations focused on providing support for disadvantaged members of the wider college community. The recent establishment of the Centre for Human Rights and Citizenship Education in Ireland within the college serves as a focal point for the dissemination of information and resources pertinent to social justice issues (Waldron & Ruane, 2011).

Learning to Teach for Social Justice in Three Countries: Analysis of Results
As was the case in the previous two studies using the LTSJ-B scale, the Rasch rating scale model was employed for the analysis of the LTSJ-B data from the three research sites. This specific Rasch model is appropriate when Likert response categories are intended to have a fixed meaning for each of the items in a scale. The model generates an estimate of each candidate's "commitment to teaching" for social justice, an estimate of each item's "difficulty-of-endorsement," and a set of response category estimates corresponding to the "transition difficulty" of choosing successively higher level Likert responses.

Data Analysis
The U.S. data consist of four cohorts of 585 entering undergraduate students and six cohorts of 738 exiting undergraduate students. The New Zealand data consist of two cohorts of 569 entering graduate-level students and one cohort of 398 exiting graduate students. Ireland's data consist of one cohort of 283 entering B.Ed. students and two cohorts of 533 exiting B.Ed. students.
It is important to note that although the exit data do not represent exactly the same students as those who provided the entry data, 5 these data come from comparable groups of students within each institution. Specifically, prior to conducting the entry-to-exit change analyses, the similarity and comparability of entering and exit cohorts were tested within each institution using procedures presented in Enterline, et al (2008). These tests established that entering cohorts within each of the respective institutions were similar in terms of demographic characteristics and level of commitment to teaching for social justice as measured by the LTSJ-B scale, as was expected due to relatively stable admissions policies within each institution. The tests also established that graduating cohorts within the respective institutions were similar-also as expected due to relatively stable programs of study and educational experiences of the students. These results suggest that any differences subsequently found from entry to exit are more likely related to candidates' program experiences rather than idiosyncrasies found within different cohorts.
Each institution's entry-to-exit change analysis was then independently conducted utilizing the Winsteps Rasch software package (Wright & Linacre, 1998). This strategy ensured that the LTSJ-B scale at entry and exit maintained the same contextual, cultural and linguistic meaning to candidates within a given institution. These initial analyses indicated that the LTSJ-B scale structures were similar from entry to exit within each separate institution (Ludlow, Enterline, O'Leary, Ell, Bonilla, & Cochran-Smith, 2010). These findings provide sufficient evidence that, within a given institution, the scores may be interpreted the same way at entry and exit, which is necessary for conducting the change analyses reported in the following sections.
Those same analyses, however, revealed that while the scale structures for all three institutions were similar, the SPC scale structure was slightly different from those of the BC and UA. For example, Irish teacher candidates were more likely to agree with two items-SJ11R (school success depends on hard work) and SJ5R (the goal for working with ELLs is assimilation)-than were their U.S. or N.Z. peers. These minor differences among the ways that the LTSJ-B scale items were understood by candidates at the three research sites meant that the three sets of entry-to-exit data could not be merged to form a single scale. Hence we conducted the analyses of change from program entry to exit within each respective institution, although, as we discuss below, there were important patterns that cut across all three sites.
Finally, even though previous analyses had established the invariance property of the item estimates at entry and exit, a single set of estimates had to be selected as the common metric by which change could be measured. The procedure and rationale established in Enterline, et al (2008) was followed here whereby the item and category estimates derived from an institution's exit solution were used as "anchor estimates" for the entry data. This means that the LTSJ-B scores of entering students were placed on the scale's continuum in such a way that they could be interpreted in terms of distance from the level of exiting students. This anchoring process was performed for each separate institution.
We turn now to the findings of the entry-to-exit change analyses for teacher candidates at each research site, BC, UA, and SPC, respectively. For each, we present a variable map, representing entry and exit responses to the items on the LTSJ-B scale. On each variable map, the hardest to endorse items are to the right of the vertical line at the top of the scale, the easiest to endorse items are at the bottom. The highest scoring candidates are to the left of the vertical line at the top of the scale, the lowest scoring are at the bottom. The average scale score is designated with "M." The entry and exit item estimates are in identical locations because they are anchored on the exit results.

Research Site 1: Boston College Results
We begin with BC, Research Site 1. Figure 2 contains two variable maps, comparing BC entry and exit data. This figure indicates that the teacher candidate responses at exit from the program were considerably higher, on average, than those at entry. The entry mean logit estimate is .42, which translates into an average total score of 40.6 (S.D. = 4.1) out of a total possible of 60. The person separation and person reliability statistics were 1.2 and .57, respectively (Wright & Masters, 1982). These two statistics indicate very little differentiation in the range of responses at entry. This finding of relatively slight variation in responses is typical when respondents have little training or experience with the items on a scale prior to their exposure to some form of intervention-in this case, the teacher education program. Hence, once exposure to an intervention such as the teacher education program occurs, the mean of the responses tends to shift upward, and the responses tend to exhibit greater variation-particularly in the upper range.
The exit mean logit estimate was 1.36, which corresponds to a mean score of 48.8 (S.D.=4.4). The separation and reliability statistics were 1.8 and .76, respectively. These statistics tell us that at exit, the average level of the BC teacher candidates' LTSJ-B commitment increased, the distribution of their scale scores shifted upward on the continuum, and the spread and variation in their scores, particularly in the upper range, increased. This pattern of results is ideal when any positive intervention, such as a teacher education program, changes attitudes, content knowledge, behaviors, or practices. These statistical differences in entry and exit levels of commitment to teach for social justice may be more fully understood in terms of the responses expected, under the probabilistic Rasch model, at the average locations on the scale. In the present context, the term "expected" is a statistical prediction of how a candidate should have responded-given the candidate's estimated level of commitment to teaching for social justice and an item's level of difficulty of endorsement. Recall from Figure 1 that depending on the wording of the item, the strongest endorsement of an item is either Strongly Agree or Strongly Disagree. Furthermore, based on a candidate's level of endorsement across all items, their expected response to any given item may range from Strongly Agree to Strongly Disagree. This means that every candidate location along the continuum represented in Figure 2 has a different pattern of expected responses to the 12 items. These expected responses, computed and reported in the software output, may be used to explain what it means to score at the average scale score level ("M" in Figure 2).  At entry, the BC teacher candidate average scale score location corresponded to Strongly Agree or Agree with the relatively less controversial, positively worded items, such as the importance of examining one's own attitudes and beliefs (SJ1), the need to incorporate diversity into lessons (SJ4), the importance of openly discussing racism and inequity in the classroom (SJ2), the need to challenge school inequities (SJ7), and the value of teaching students to think critically (SJ8). The candidates also "appropriately" Disagreed with a number of the more controversial negatively worded items, such as the reasonableness of teachers having lower expectations for ELL students (SJ6R) and believing that economically disadvantaged students bring less to the classroom (SJ9R). However, teacher candidates were also Uncertain about a number of these issues, such as whether multicultural topics are relevant to all subjects (SJ3R), whether social change is part of the teacher's job (SJ10), and whether or not assimilation is the main purpose of teaching for immigrant students (SJ5R). Moreover, they tended to "inappropriately" Agree with the two items on the scale that are the hardest to endorse-the belief that school success depends primarily on individual students' hard work (SJ11R) and that the teacher's main job is preparing students for their "likely" roles in life (SJ12R).

Candidates--MAP--Items
From an ideal theoretical perspective and assuming change over time toward greater endorsement of beliefs related to social justice teaching, teacher candidates' appropriate Agree responses at program entry would become Strongly Agree at exit. In other words, for example, teacher candidates who Agreed at entry that classroom lessons and discussion should incorporate diverse examples and perspectives (SJ4), would Strongly Agree with this idea at exit. Likewise, appropriate Disagree responses at entry would become Strongly Disagree at exit. Uncertain responses would shift, either to Agree or Disagree or to Strongly Agree or Strongly Disagree depending on the positive or negative direction of the item.
Looking at actual results, we found that at exit, the average BC teacher candidate's scale score location "M" corresponds to appropriate Strongly Agree responses to items about the importance of teachers attending to diversity (SJ4), examining their own beliefs (SJ1), discussing inequity (SJ2), challenging inequities (SJ7), and thinking critically (SJ8). Candidates also appropriately Strongly Disagreed with the idea that teachers could reasonably have lower expectations for ELL students (SJ6R) and for economically disadvantaged students (SJ9R) and Disagreed with the belief that multicultural topics are only relevant to certain subjects (SJ3R), the idea that it is not the teacher's job to change society (SJ10R), and the assumption that the goal for ELL students is assimilation (SJ5R). All of these items have to do with low teacher expectations and goals for students and/or for teachers themselves. At exit, teacher candidates also tended either to be Uncertain about or Disagree with the two hardest to endorse items--SJ11R (students' success depends on their effort) and SJ12R (teachers should prepare students for their "likely" lives). Appropriate disagreement with these two items involves rejection of the idea that schools and society are meritocratic systems and that success depends on individual effort and action, rather than unearned advantages accrued from an hegemonic system. According to the definition of social justice beliefs we have used here, these shifts in teacher candidates' response patterns from entry to exit reflect positive changes in their beliefs. Interpretation of these changes and discussion of their significance along with comparison of the BC results with the results from the other two sites are addressed in the final section of this article. Figure 3 contains the variable maps for the UA entry and exit data. The structure of the LTSJ-B scale is nearly identical to the BC results, i.e., the UA and BC item order and location estimates are remarkably similar. Similar to the BC results, the UA entry mean logit estimate is .3 (mean score = 39.3, S.D.=3.8), and the person separation and person reliability statistics are 1.0 and .52, respectively. These reflect the previously seen pattern of relatively low scores along with slight score variation in the entering teacher candidate responses. In addition, the UA teacher candidate responses at exit are higher, on average, than their responses at entry. 186 The UA exit mean logit estimate is .72 (mean score = 44.7, S.D.=4.5). The separation and reliability statistics are 1.7 and .75, respectively. These UA results at exit are consistent with the BC results-the average level of LTSJ-B commitment has increased, the distribution of scores has shifted upward on the continuum, and the spread and variation in scores has increased.

Figure 3. University of Auckland Entry and Exit Variable Maps
Based on their average scale score location ("M"), the UA teacher candidates at entry tended to Strongly Agree or Agree with the relatively easier to endorse, less controversial items (SJ4, SJ1, SJ2, SJ8, SJ7), as described above. They were Uncertain, however, about some of the more controversial items (SJ6R, SJ9R, SJ3R), which have to do with the acceptance of lower expectations for some students and belief that multicultural topics are only relevant to certain subject areas. UA teacher candidates also tended to "inappropriately" Agree with four of the hardest to endorse items (SJ5R, SJ10R, SJ11R, SJ12R), which, like the BC results, have to do with assimilation, meritocracy, hegemony, and teachers' roles in social change.
At exit, the average UA teacher candidate scale score location corresponded to Strongly Agree with SJ4, SJ1 SJ2, SJ8, SJ7, as above, and Disagree with SJ6R, SJ9R, and SJ3R, as above, a pattern that is again very similar to the BC results. At exit, most candidates had shifted to being Uncertain about, rather than inappropriately agreeing with SJ5R, SJ10R, SJ11R (as above) but still tended to "inappropriately" Agree with SJ12R, the hardest to endorse item about the teacher's role in terms of preparing students for the lives they are "likely" to lead. These results, similar to BC, show positive development in terms of our definition of learning to teach for social justice. Figure 4 contains the variable maps for the SPC entry and exit data. The structure of the scale is similar to the BC and UA results, yet unique in the location of a few items. For example, as mentioned earlier, items SJ5R (assimilation should be the goal of teaching immigrant students) and SJ11R (school success depends primarily on students' hard work) were relatively more likely to be agreed to by SPC teacher candidates while SJ10R (it is not the teacher's job to try to change the world) was less likely to be disagreed with in comparison to the responses of BC and UA teacher candidates. Nonetheless, since the SPC entry and exit estimates are anchored on SPC-only data, it is evident that the teacher candidate responses at exit were higher, on average, than those at entry.

Research Site 3: Saint Patrick's College Results
The entry mean logit estimate for the SPC candidates is .5 (mean score = 40.2, S.D.=3.1). The person separation and person reliability statistics are .7 and .35, respectively. The exit mean logit estimate is .95 (mean score = 45.5, S.D.=3.4). The separation and reliability statistics are 1.5 and .68, respectively. Consistent with the BC and UA results, the SPC average level of LTSJ-B commitment increased at exit, the distribution of scores shifted upward, and the spread and variation in the scores increased.
Based on their average scale score location at entry ("M"), the SPC teacher candidates tended to Strongly Agree with SJ4 (incorporating diversity into lessons) and Agree with SJ1 (examining one's own beliefs) and SJ2 (the importance of discussing inequity). They were Uncertain about SJ7 (challenging school inequities) and SJ8 (thinking critically about government positions). They tended to "appropriately" Disagree with a number of the harder to endorse, more controversial items SJ6R (acceptance of lower expectations for ELLs), SJ5R (assimilation as the goal for ELLs), and SJ9R (low expectations for socioeconomically disadvantaged students), which have to do with assimilation and lowered expectations for some students, but were Uncertain about SJ3R (relevance of multicultural topics) and SJ11R (school success depends primarily on hard work). Like the BC and UA teacher candidates, at entry SPC teacher candidates tended to "inappropriately" Agree with the hardest to endorse items SJ10R (it's not the teacher's job to change society), and SJ12R (the teacher's job is to prepare students for their "likely" lives), which relate to teachers' roles as agents of change and society equity.
At exit, the average SPC response shifted from Agree to Strongly Agree with SJ4 (good teaching incorporates diversity), SJ1 (teachers should examine their own beliefs), and SJ2 (inequity should be openly discussed in the classroom). On average, teacher candidates appropriately Agreed with SJ7 (teachers should challenge school inequities) and SJ8 (teachers should teach students to think critically about government positions) and appropriately Disagreed with SJ6R (teachers may have lower expectations for ELL students), SJ5R (assimilation is the appropriate goal for ELL students), SJ9R (economically disadvantaged students bring less to school), SJ3R (multicultural issues are only relevant to certain subjects), and SJ11R (school success depends on how hard students work). They moved from In this final section, we discuss our interpretation of the results of the administration of the LTSJ-B scale across the three research sites, and we also consider the notion of learning to teach for social justice as a cross-cultural construct. We make three major points along these lines.

Dramatic Differences in Contexts, Yet Striking Similarities in Results
Based on our analysis of data in the three sites described above, we conclude that the LTSJ-B scale appears to work the same way across cultures and countries. Learning to teach for social justice thus appears to represent a cross-cultural concept that is significant to initial teacher education.
As is clear in our discussion above, the three countries in which we administered the LTSJ-B scale to groups of entering and exiting prospective teachers have in common a teaching population that is primarily (in the U.S. and N.Z.) or overwhelmingly (in Ireland) White and western European in ethnic background, along with a student population that is increasingly diverse. Thus the initial teacher education programs we studied in all three sites are struggling with the challenge of preparing teachers to work successfully with students who are different from them in ethnic, racial, cultural, geographic and/or linguistic background.
Despite the shared challenge of preparing monocultural teachers for multicultural student populations, however, the three sites are remarkably different. They are located in vastly different parts of the world with strikingly dissimilar cultural, social, demographic, and geopolitical histories. In the U.S., new immigration patterns coupled with a long history of institutional racism have brought issues of diversity and inequality to the forefront of educational policy and practice issues. At the same time, however, a "new civil rights" movement has emerged (Kumashiro, 2010) that actually works against the historical goals of mid 1900s Civil Rights movements by undermining public education and reinforcing stratification and inequitable opportunities and outcomes (Zeichner, 2010). This is reflected in the backlash against multicultural and social justice approaches to teaching and teacher education (Cochran-Smith, Barnatt, Lahann, Shakman, & Terrell, 2009), and there is increasing pressure to focus on only those aspects of teacher preparation that can be directly linked to student achievement on high stakes tests (e.g., U.S. Department of Education, 2011).
In contrast, as we noted above, New Zealand is officially a bicultural nation, legally comprised of indigenous Maori people and former British colonizers, now New Zealand Europeans. However, as the statistics indicate, increasingly New Zealand is demographically multicultural. Given the history of the oppression of the indigenous Maori by European colonizers, however, and current enforcement and respect for the Treaty of Waitangi that established the nation in the 1800s, it is clear that New Zealand will continue to be bicultural legally and politically. Further, the marked and enduring inequalities between New Zealand's high and low achievers coupled with the over-representation in the lowachieving group of Maori and Pacific Island students have exacerbated highly volatile issues regarding appropriate educational curriculum, instruction, and assessment for Maori, as well as more recently identified minority groups.
In contrast to the U.S. and New Zealand, Ireland is a nation with a traditionally highly homogeneous White, Catholic and native-born Irish population with a very short history of diversity among students. However, recent new immigration patterns, noted above, have brought many "newcomer" students to Ireland, 70-75% of whom do not speak English as a first language, which is almost certain to have an impact on the nation's levels of achievement. In fact, recent international comparisons and national reports already indicate that students from immigrant groups, socially disadvantaged backgrounds, and/or from the Traveller community, are more likely to fail in school (Ireland Department of Education and Skills, 2010).
In addition to these geopolitical and historical differences in the three research sites, the teacher candidates we surveyed were also quite different from one another in terms of level and duration of initial teacher education-the U.S. site was a four-year undergraduate program leading to a bachelor's degree in education with certification at the elementary or secondary levels; the N.Z. site was a postgraduate level program leading to certification in elementary teaching; and, the Irish site was a threeyear undergraduate program leading to a bachelor's degree and certification at the elementary level. This means that not only were the teacher candidates we surveyed learning to be teachers in dramatically different social and cultural milieus, but also the candidates differed from one another in age, educational background, life stage, educational interest, and subject matter knowledge.
Given these stark differences among the three sites, the striking similarity of patterns in teacher candidates' responses to the LTSJ-B survey items and the patterns in changes in teacher candidates' responses from program entry to program exit is remarkable. And it is important to emphasize that these are not simply general similarities, but, with the slight exceptions noted above, similarities in responses to specific items and virtually the same patterns of change in responses over time. In fact, the most important finding of this study is that, despite the fact that teacher candidates' beliefs are very difficult to change (Wideen, Mayer-Smith & Moon, 1998), there was a similar pattern of responses and a very similar pattern of change from entry to exit in the LTSJ-B scale across dramatically different contexts for initial teacher preparation and across cultures, age differences, and duration of program.
What our study clearly shows is that despite differences in cultures and contexts, at program entry, teacher candidates in all three sites were most likely to believe in and agree with those items that had to do with what they perceived as their work as individual teachers, their actions within their own classrooms, and their students as individuals. In contrast, the items teacher candidates were least likely to believe in or agree with had to do with recognition of the impact on students' educational experiences and outcomes of larger societal structures, such as the ways institutions, including schools, systematically and historically structure advantage and disadvantage for particular groups, including the lack of opportunity to participate in the establishment of educational purposes and goals in the first place. We elaborate below.

A Modest Shift from Individual to Structural Perspectives
At entry candidates tended to agree with survey items related to their individual practices as teachers and their own work inside classrooms. For example, at entry, most candidates endorsed the idea that teachers should incorporate diverse cultures and experiences into their classroom lessons and discussions; they also generally agreed that teachers should critically examine their own attitudes and beliefs about aspects of diversity. However, also at the point of entry to the program, candidates tended to agree with or, at best, be uncertain about, the assertion that the job of the teacher is to prepare students for the lives they are "likely" to lead. Agreement with this latter statement runs counter to the idea of teaching for social justice, as we have defined it here, in two ways-it ignores the negative impact of larger societal structures and systems on the educational opportunities and outcomes of student groups traditionally marginalized by the system, and it undermines the role of the teacher as advocate and activist. From a social justice perspective, teachers are conceptualized as agents who challenge stereotyped expectations about students, such as preconceived ideas about what is "likely" to become of students from various socioeconomic, racial, cultural, and gender groups. From this perspective, teaching is understood as a political activity, and teachers are expected to work with others against the grain of limited and limiting ideas about students' capacities and prospects based on their backgrounds, experiences, and demographic characteristics. Similarly, at entry teacher candidates at all three sites tended to agree with the idea that how students do in school depends primarily on how hard they work, an idea that focuses on individual student effort and motivation, rather than on the experiences of individuals as part of historically marginalized or advantaged groups for whom educators and educational systems have long held diminished expectations. Rather, the viewpoint of most teacher candidates at program entry assumes a meritocratic society and a meritocratic system of schooling and, again, ignores the fact that deeply embedded structures and systems have routinely limited students' educatio-nal opportunities and their life chances on the basis of socioeconomic status, race, culture, and language background.
Initial teacher education is a programmatic intervention intended to have an impact on prospective teachers' knowledge, skills, and beliefs about teaching. Thus teacher candidates' scores on the LTSJ-B scale upon program exit are especially important. As we have detailed above, we found that at program exit, the responses of teacher candidates in all three sites tended to shift from a more individualistic perspective about students and about themselves as teachers, toward more systemic understandings of how societal structures shape opportunities and outcomes and also determine educational goals and objectives in the first place. Teacher candidates' views also shifted away from the implicit assumption that the teacher is engaged in a politically neutral enterprise and toward the assumption that part of a teacher's role is to work with others to challenge inequities and enact social change. Specifically, at program exit, candidates were more likely to endorse survey items that reflected the idea that teachers' work was not simply about what happens inside classrooms between the individual teachers and their students, but is also about challenging the ways educational institutions and societies have historically perpetuated advantages and disadvantages for various groups, based on ethnicity, culture, language, race, and socioeconomic status. At entry, candidates tended to disagree with or be uncertain about items that implied an advocacy or activist role for the teacher; at exit they tended to shift from disagreement to uncertainty or from uncertainty to agreement. In other words, the pattern of change from entry to exit in all three sites was a move away from a focus primarily on the individual teacher working in his or her classroom to try to meet the needs of individual students, and a shift toward the idea of the individual teacher as part of a larger collective of educators and others working to challenge school and societal structures that perpetuate inequities for groups of students based on race, class, culture, language background and other demographic characteristics.
The modest, but very clear cross-site shift in beliefs that we documented among teacher candidates in three different programs and countries is consistent with some other work related to teacher education for social justice, and more broadly, work related to conceptions of critical citizenship and participation in democratic societies. In the context of this work, it is not surprising that many teacher candidates begin their initial teacher preparation work with a focus on individualism and a commitment to meeting the needs of individual students coupled with rejection of or uncertainty about the teacher's role in challenging larger structural issues or working to challenge the arrangements of schooling that perpetuate inequities. It requires a major (and difficult to accomplish) transformation in thinking for many teacher candidates to understand the structural and historical aspects of schooling and develop analyses and critiques at the macro-level (Cochran-Smith, Shakman, Jong, Terrell, Barnatt, & McQuillan, 2009). In fact, although there are some exceptions documented in the literature (e.g., Achinstein and Ogawa 2006;Cochran-Smith 1991), most research suggests that it often takes several years for new teachers simply to get a handle on the day-to-day work of teaching, let alone begin to address larger issues beyond the classroom.
It is also not uncommon for teachers to have a view of "good teaching" that focuses primarily on individualism and being personally responsible. Many teachers enter the profession because they want to change the world "one child at a time." The emphasis in the prevalent educational discourse on the individual student is consistent with Westheimer and Kahne's (2004) versions of citizenship education. They suggest that there are three common notions of "citizenship" in citizenship education programs, the first being the most common: the first is implicit in civic education programs that seek to promote personally responsible, hard-working, and law-abiding citizens, while the second is related to participatory citizen programs, which promote active participation in civic organizations and social life at all levels. However, it is only at Westheimer and Kahne's third version of citizenship education, which they call "justice-oriented" citizenship programs, that explicit attention is given to matters of injustice and to the importance of pursuing social justice goals at the structural and societal levels. 192 Westheimer and Kahne point out that few citizenship education programs take this approach. Likewise Zeichner (2006) criticizes the frequent use of social justice as a term describing teacher education since most programs emphasize individual efforts rather than structural changes in teaching and teacher education.

LTSJ-B Scale as a Cross-Cultural Assessment Tool: Broad but Blunt
As we have made clear, the structure of the LTSJ-B scale, that is, item order and location estimates, was remarkably similar for all three institutions. The relatively minor differences in the SPC versus BC and UA results are attributable to administration and language differences implemented to suit local circumstances. This means that the LTSJ-B scale is an instrument that captures-across cultures, countries, ages of teacher candidates, and differences in program duration-the general range and variation in teacher candidates' beliefs about teaching for social justice. The fact that this instrument is able to capture aspects of teacher candidates' beliefs despite very different contexts is worth noting.
In addition, teacher candidates at each of the three institutions demonstrated modest changes in their beliefs about teaching for social justice. Again, even though the shift in beliefs is modest, the change is striking, particularly given that programs differed in their approaches and years of preparation. At the time of entry, the average teacher candidate was committed to fairness and equal opportunity for students within her/his individual classrooms, but was uncertain about teachers' roles as advocates for societal change, and generally did not reject stereotypes that limit students' learning opportunities and life chances. At exit, whether from a one-three-or four-year initial teacher preparation program, the beliefs of teacher candidates, on average, shifted in a positive direction in all these areas, although the easier to endorse items continued to be easier to endorse, and the more difficult to endorse items continued to be more difficult. We also know from previous analyses of data based on the BC program (Enterline, Cochran-Smith, Ludlow, & Mitescu, 2008) that graduates maintained their higher scores on the social justice beliefs at exit even after one year of teaching. This suggests that these changes in beliefs were not simply the short-term result of program emphasis.
The capacity of the LTSJ-B scale to measure status of social justice beliefs and change over time has generated interest across the United States and internationally. Graduate students from a variety of programs have requested permission to use the scale for thesis and dissertation research, and university professors have requested permission to use the scale for their individual research. In addition, the scale has been requested by Teacher Education programs and Schools of Education that seek to model the assessment system developed at Boston College for measuring programmatic impact and change from entry to exit. A brief listing of institutions currently using the LTSJ-B scale in the U.S. includes the University of Texas-Austin, University of Louisville, University of Washington, New York University, George Mason University, University of Colorado-Denver, and St. Joseph's University. Outside the U.S., in addition to St. Patrick's College in Ireland and the University of Auckland in New Zealand, the University of Puerto Rico and University of Chile have used a Spanish version of the scale, and the University of Minho (Portugal) has used a version of the scale translated into Portuguese. In addition non-profit organizations, including Wide Angle Youth Media, the Indiana Civil Rights Commission, and Kent and Medway Training have requested permission to use the scale.
It seems clear that the LTSJ-B scale has wide applicability and has been used in a variety of contexts, countries, cultures, and institutional settings, even though the issues related to diversity, equity, access, and outcomes in these settings vary considerably. In fact, it may be that the opposite side of the coin of the broad applicability of the scale may be its blunt treatment of the issues, which ultimately does not get us far enough in our understanding of how teacher candidates learn to teach for social justice. We need to understand much more about the nuances of teacher candidates' ideas and understandings and about the apparent inconsistencies of certain ideas and how these shift over time.
For example, what does it mean that teacher candidates across contexts and countries easily embrace the idea that teachers should examine their own attitudes and beliefs about race, class, gender, disabilities, and sexual orientation, yet at the same time also accept the assertion that school success depends primarily on students' individual efforts? Holding the former belief reflects acknowledgement that educators' assumptions about race, class and gender may be part of an ideology that perpetuates inequities in students' learning opportunities, while holding the latter belief reflects the assumption that students live in meritocratic societies and participate in meritocratic systems of schooling and school advancement. Although these two sets of ideas are not completely parallel, they are definitely related, and holding both of these beliefs at the same time reflects inconsistencies. Unfortunately the LTSJ-B scale does not capture the subtleties and nuances that would help to explain this. Now that we have a reliable and valid instrument for getting at the broad range and variation in teacher candidates' beliefs related to teaching for social justice, we need to develop additional instruments that are more incisive and are designed to capture nuances and local variations.
In addition, we know little about which aspects of the three programs described here actually influenced teacher candidates' beliefs. Although the three programs have similarities, as we noted, they differ in duration and student population, and the inequities in each country are the result of different geopolitical histories and national agendas. We now need further research to drill down into the ways teacher candidates experience the social, intellectual and organizational contexts of their programs that are intended to address issues related to justice and equity in order to understand whether there are basic differences and similarities in how teacher candidates' core ideas change over time.
In conclusion, it is very promising that teacher candidates in three very different institutional and cultural settings shifted their beliefs regarding teaching for social justice over the course of their initial teacher education experience. As noted above, we now need additional instruments that can get at some of the nuances of these beliefs and their more subtle and intricate meanings. As has long been known, however, beliefs and actions are not the same and are often not consistent. Thus we also need to know more about how the beliefs of teachers prepared in programs that specifically address social justice issues compare to their classroom and school practices, including activities that might be termed advocacy and activism, and how all of this is related ultimately to students' learning opportunities, outcomes, and life chances.