Cambodian teachers' responses to child-centered instructional policies: A mismatch between beliefs and practices

Teaching and Teacher Education 50 (2015) 36e45

Contents lists available at ScienceDirect

Teaching and Teacher Education journal homepage: www.elsevier.com/locate/tate

Cambodian teachers' responses to child-centered instructional policies: A mismatch between beliefs and practices Sopheak Song* Hiroshima University, Japan

h i g h l i g h t s
Teachers do not totally shed conventional beliefs to embrace child-centered pedagogy.
Teachers do not fundamentally change their conventional teaching practice.
Classroom realities hinder the adoption of child-centered instruction.
Teachers select and put into practices only some superficial aspects of reform.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 16 September 2014 Received in revised form 3 March 2015 Accepted 23 April 2015 Available online 14 May 2015

Cambodia and her international development partners have been promoting child-centered pedagogy for almost two decades. However, classroom instruction remains predominantly front-oriented and textbook-based. Drawing on questionnaire and interview surveys with primary school teachers in two districts of Cambodia, this study analyzes teachers' beliefs in and classroom implementation of childcentered pedagogy. It has found that teachers fail to adopt this new pedagogy in their classrooms despite their overwhelming support for it. Teachers' failure to act on their beliefs stems from two major reasons: constraining classroom realities and their superficial understanding of the principles underlying child-centered pedagogy. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Child-centered pedagogy Instructional reform Implementation research Classroom environment Teacher beliefs Cambodia

1. Introduction In 1996, Cambodia launched a large-scale educational reform. The Ministry of Education, Youth, and Sport (MoEYS) installed a new 6th grade into the existing 11-year (5þ3þ3) education model, developed a new curriculum, wrote textbooks and teacher manuals accordingly, and prescribed a modern teaching and learning approach broadly identified as “child-centered” pedagogy. The reform reflected the spread to developing countries of the childcentered pedagogy, which migrated from Western nations in the early 1990s. This widespread pedagogical migration has apparently stemmed from two interrelated global forces: globalization and human rights movement.

* Present Address: A205, 8510-1 Hara, Hachihonmatsu-cho, Higashihiroshimashi, 739-0151, Japan. Tel.: þ81 9064136005. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.tate.2015.04.004 0742-051X/© 2015 Elsevier Ltd. All rights reserved.

After the end of Cold War, many developing countries showed commitment to liberal democracy and market openness consistent with Western industrialized countries (Chisholm & Leyendecker, 2009). National education programs started to be significantly influenced by educational agendas promoted by major international organizations and donor agencies such as UNESCO (United Nations Educational, Scientific, and Cultural Organization) and USAID (United States Agency for International Aid) as well as regional and international financial institutions such as ADB (Asian Development Bank) and the World Bank. With their world views of how children should be educated, these organizations prescribe educational reforms to developing countries, the prescriptions that give policy makers of those countries little choice but to implement the reforms in exchange for access to needed fund (Anorve, 1999). Child-centered pedagogy has also been promoted through human rights ideals. Many developing countries have ratified a number of rights-based conventions on education, including Convention on the Rights of the Child (1989), the World Declaration

S. Song / Teaching and Teacher Education 50 (2015) 36e45

on Education for All (1990) and the UN Millennium Development Goals (2000). An exemplar of rights-based education is the worldwide movement of “child-friendly” schools (CFS), which have been vigorously promoted by UNICEF. Explicitly linked to childcentered pedagogy is the movement's principle of “childcenteredness,” which states, “[Classroom instruction] should be an interactive process in which children are active participants in observing, exploring, listening, reasoning, questioning and ‘coming to know’” (UNICEF, 2009, p.13). Meyer and Ramirez (2010) observe that human rights emphasis of education are associated with the globally rising child-centered pedagogy. To some scholars, these transnational forces have resulted in a worldwide convergence of patterns of educational organization, of school curricula, and of patterns of expansion taking place on all levels of educational systems (Meyer & Ramirez, 2000). However, studies which investigate the changes at local level consistently suggest that the global prescriptions neither replace already existing models, nor mean the same thing in various cultural contexts (Anderson-Levitt, 2003; Steiner-Khamsi & Stolpe, 2006). Even the globalization-minded researchers acknowledge that education appears to have changed little at classroom level in most countries (e.g., Carnoy & Rhoten, 2002). This paper aims to investigate primary school teachers' responsesdbeliefs and classroom practicesdto educational reform on child-centered pedagogy. Previous studies (e.g., Courtney & Gravelle, 2014; McCormick, 2012) show that global educational agendas have significantly influenced educational policies in Cambodia. However, little is known about how much these imported educational ideals have changed teaching at classroom level. To narrow this gap, this paper will provide some evidence of how primary school teachers implement child-centered instructional policies in their classrooms. Implementation of learner-centered education policies at classroom level has been a recurrent research theme starting with the work in 1990 by the California Study of Elementary Mathematics conducted by scholars at Michigan State University (See Cohen & Ball, 1990a, for a review of the project). However, the predominant mode of analyses in developing countries still focus on how the principles underlying this constructivist approach to education are embedded in national educational reform (Schweisfurth, 2011). Relatively few studies investigate how learner-centered education policy works out in classroomsdwhere the national policy is translated to practice. This study attempts to make an important contribution to the limited knowledge on educational policy implementation that is based on local voices and to inform the global debate on how to best promote learner-centered education by bringing evidence from a least-studied country: Cambodia. 2. Pedagogical reform and teacher guidance

37

UNESCO and UNDP financed a national education seminar in January 1994, attended by international consultants and ministry officials, to map out the national policy for the development of quality education and training in Cambodia (MoEYS, 1994). The Asian Development Bank intensively reviewed the education sector in 1994 (ADB, 1996), while UNICEF tried to enhance instructional quality by improving both pre-service and in-service teacher training, revising curriculum, and providing instructional materials such as textbooks and teacher manuals (MoEYS, 1994). The assistance and recommendations of these international organizations were incorporated into the MoEYS's Education Investment Plan 1995e2000 and served as the bases for the reform of Cambodian education in 1996. In 2007, the MoEYS issued the ChildFriendly Schools Policy, again based on the experiences of UNICEF and other development agencies such as SCN (Save the Children Norway), which had supported pilot projects to promote children's rights to education in several provinces since 2002 (MoEYS, 2007a). MoEYS aimed to make 94% of the country's primary and lower secondary schools “child-friendly” by 2011 (MoEYS, 2007b). As part of the 1996 reform package, MoEYS instructed the country's schools at all levels to use child-centered pedagogy as teaching and learning approach. Child-centered pedagogy is defined as the following: A teaching and learning approach which is based on pupils' activities. In child-centered pedagogy, pupils are expected to be active participants in the learning process while teachers are just facilitators or catalysts. Pupils can learn by themselves individually, in small groups, or whole groups in or outside the classroom. (MoEYS, 2001a, p.18) A former director of the Department of Teacher Training of the MoEYS wrote about the reform, “the new curriculum will inspire [pupils] to better understand their surrounding environment, apply methods of problem solving and enrich their spirit of innovations” (Bunroeun, 1999, p.16). From these policy statements emerges a new perspective on teaching and learning, which emphasizes problem solving skills and children's participation in the instructional process. The emphasis on the role of children as active problem solvers is also advocated in subsequent educational policies. According to Cambodia CFS policy issued in 2007, childcentered education is characterized by participation and cooperation; problem solving; and research, analysis, and critical thinking (MoEYS, 2007a). Another guideline for textbook development dictates, “textbooks should be written to promote pupils' higherorder thinking skills such as analysis, application, synthesis, and evaluation” (MoEYS, 2012, p.27). Policy statements of education in Cambodia consistently depict a new type of teaching and learning: an educational perspective that underscores pupils' active involvement in classroom activities and the development of higher-order thinking.

2.1. The reform and its ideals 2.2. Teacher training on child-centered pedagogy Like other developing countries, Cambodia is not immune to global forces. In the post-Cold War years, Cambodia, a former ally of the socialist Eastern Bloc, shifted to a nation governed by liberal democracy and market economy as enshrined both in the 1991 Paris Peace Agreements between the warring factions of the country and in the 1993 constitution (Royal Government of Cambodia, 1993, Articles 1 and 56). Bilateral and multilateral aid agencies such as UNDP (United Nations Development Programme), USAID, SIDA (Swedish International Development Cooperation Agency), and financial organizations such as Asian development Bank were in the line to help Cambodia, which had been tormented by wars for two decades (Ayres, 2000, p.157).

Teachers learn about child-centered pedagogy through various means: by attending teacher training centers/colleges, through workshops and seminars organized by MoEYS and NGOs, and through curriculum materials such as textbooks and teacher manuals they use for their classroom teaching. Fig. 1 shows the process through which instructional policy making and implementation in Cambodia take place. From the highest level where technical consultants prescribe reform ideas to policymakers down to classroom level where teachers translate those ideas into concrete activities, the process involves several transit points, especially with regard to in-service teacher training.

38

S. Song / Teaching and Teacher Education 50 (2015) 36e45

Fig. 1. Pedagogical reform and implementation process.

Global ideals on education are often injected into national policies by experts affiliated with international organizations. Officials at the Teacher Training Department with other related departments (Department of Curriculum Development, Department of Primary Education) of MoEYS are responsible for developing training materials and selecting national core trainers among the officials in those departments. These national core trainers train provincial trainers, usually teacher trainers at Provincial Teacher Training Centers (PTTCs) or provincial educational officers. The provincial trainers, in turn, train district staff in seminars or workshops organized at provincial level. Then, the district staff (also known as the District Training and Monitoring Team [DTMT]) train school principals and teachers in workshops organized at a cluster-school, which is usually made up of five or six nearby schools. Generally dubbed “technical meetings” by teachers, the cluster-school level workshops, held on the last Thursday of the month, are the most sustained mechanism for teachers' continuous professional development. When child-centered pedagogy was mandated in 1996, MoEYS instructed that pre-service teacher training programs should use new textbooks, teacher manuals and teacher orientation materials to train new teachers (MoEYS, 1997). UNICEF and SCN were the main supporters for the development and distribution of training materials for teacher training centers (Popov, 1998). An example of such materials is a 195-page textbook, published in 2001, detailing how to use the new curriculum and mathematics textbook for grade 6. The textbook devotes a portion (12 pages) to the introduction of child-centered pedagogy and spends the remainder on the orientation of the new curriculum, lesson planning and utilization of the new pupils' textbooks (MoEYS, 2001a). In-service teachers were also trained with the same materials. However, only a small number of representative teachers received short-term training directly at the Provincial Teacher Training Centers (PTTCs). The rest of the teachers were supposed to learn from those representatives through locally-organized workshops (at cluster-school level). New training materials on child-centered pedagogy were developed when the MoEYS issued the CFS in 2007 and were

incorporated into the teacher education curriculum of PTTCs and training workshops organized to orient in-service teachers to the CFS and child-centered pedagogy (MoEYS, 2007c, 2007d, 2008). Again, supported by development agencies, MoEYS often runs these training workshops along the cascade model described above. 3. Theoretical background A unique angle to understand policy implementation is to investigate how local government agents make sense and make use of national policies. Such investigation would allow us to compare the meaning of national reform as it is intended by policy makers and as it is received by the public. The government agents at the bottom level are called “street-level bureaucrats” (Lipsky, 2010). In most cases, street-level bureaucrats must confront severe shortage of personal and organizational resources in relations to their tasks to transform policies into practice and must find ways to fulfill the demands placed upon them by routinizing procedures, modifying goals, rationing services, asserting priorities, and limiting or controlling clients (Weatherley & Lipsky, 1977, p.172). Studies on educational policy and practice both in developed and developing countries have shown that street-level bureaucrats, i.e., school principals and classroom teachers, considerably alter instructional policies as they interpret and put them into practice. McLaughlin (1990), reviewing early research on implementation of educational reform in the United States, notes that it is an extremely difficult task to make a policy work across layers of government and institutions because policymakers cannot mandate what happens at local level. Instead, a policy is transformed as individuals at each level of government interpret and respond to it (McLaughlin, 1987). Cohen and Ball (1990b, p.335) find that “teachers do not simply assimilate new texts and curriculum guides, altering their practice in response to externally envisioned principles. Rather, they apprehend and enact new instructional policies in light of inherited knowledge, belief, and practice”. More recent studies show that principals' expertise and agendas shape how they present policy messages to teachers and that the ways

S. Song / Teaching and Teacher Education 50 (2015) 36e45

teachers interact with one another in schools influence how they understand policy messages and ultimately what they bring into their classrooms (Coburn, 2001, 2006; Spillane, Reiser, & Reimer, 2002; Stein & Brown, 1997). Teachers' understandings of policy sometimes differ by a great distance from what is originally intended (Hill, 2001). An important explanation for the contortion of policy is conceptualized under the “zones of enactment” (Spillane, 1999). A zone of enactment refers to the space where reform initiatives are translated into practice, delineating the zone in which teachers notice, construe, construct and operationalize the instructional ideas advocated by reformers (Spillane, 1999, p.144). Based on cognitive theories, this explanation claims that teachers' sensemaking of policy messages is central to implementation and is shaped by their prior knowledge, experience and beliefs (Spillane et al., 2002). However, people are conservative in their sensemaking process: they tend to misinterpret the new information as old ideas, to focus on concrete surface-level similarities, and to ignore the deeper structural and conceptual changes (Gentner, Rattermann, & Forbus, 1993; Spillane, 2004). Accordingly, reform that requires teachers to fundamentally change their repertoire of existing knowledge often poses a serious intellectual challenge for them. While obstacles to educational reform are a common theme in implementation research worldwide, they are more prominent in developing countries, which are facing aggravated challenges and constraints such as teacher resistance to external ideas, lack of resources, and expertise. In developing countries, modern educational ideas, which are often imported from Western developed worlds, have to be transported across and transformed by several stations at national, provincial, and sub-provincial levels (Brook Napier, 2003) before arriving at the classrooms, only to be modified further by teachers when those ideas are passed to the pupils. Implementation of the reform ideas ultimately rests with the teachers (O'Sullivan, 2002). Teachers have considerable discretion to thwart policy objectives, either as conscious reaction against reform, or more subtly as manifestation of their identities, priorities, and perceived limitations (Schweisfurth, 2011). In an analysis of teachers' use of child-centered teaching, Brodie, Lelliott, and Davis (2002) found that most teachers took up the forms or strategies such as resources, tasks, questions, and group work but they failed to elicit and engage with learners' ideas and interests in order to develop new knowledge. Researchers on educational reform implementation in developing countries have proposed a theory to explain why teachers modify or do not comply with policy expectations. This theory is known as the “zone of feasible innovation”, i.e., the extent to which the new teaching and learning practices are realistic and achievable, taking into account school capacity such as physical and human resources, learners' attributes, and school environment (Rogan, 2007). It proposes that innovation is most likely to take place when it proceeds just ahead of existing practice (Rogan & Grayson, 2003). Extending the “zones of enactment” explanation, this theory takes into account variables like educational context and school resources in addition to teachers' expertise. The contextual and resource challenges might be so big that, although teachers highly support child-centered curriculum goals, they barely find child-centered teaching strategies useful for their classrooms (Isikoglu, Basturk, & Karaca, 2009). Van den Berg (2002) suggests, “teachers do not just passively implement what they are told, but their behaviors are strongly influenced by the context in which they work.” Rogan (2007) pointed out that teachers often mined new textbooks for activities that they felt could be successfully achieved, taking into account the resources they had available, the cognitive level of their learners, and their own level of

39

comfort with the subject matter. Two important points should be clear from the literature reviewed so far. First, practice does influence policy: agents at the lowest level of government, where policy intersects practice, shift policy in significant ways, taking into account availability of expertise and resources and other contextual constraints. Second, implementation of child-centered pedagogies is an issue of both developed and developing countries. However, while the issue in the former tends to be determined by teachers' level of expertise in translating pedagogical principles to classroom practice, cases in the latter are replete with constraints in resources and challenging contexts. The current author argues that the adoption of modern instructional ideas in developing countries is hindered by not only teachers' limited understandings of the principles underlying those ideas but also by an array of classroom constraints, most of which are beyond teachers' control. 4. Methods 4.1. Design, sampling, and participants This paper reports part of the author's doctoral study on the quality of teaching and learning at primary education level in Cambodia. The study was “mixed methods research” corresponding to the “parallel mixed design” in Teddlie and Tashakkori (2009) typology. The researcher concurrently collected and independently analyzed data using both quantitative and qualitative approaches. At the end of the project, the results from each strand were compared and contrasted to form overall inferences of educational quality at primary schools. The data used here were derived from interview and questionnaire surveys from two fieldworks in two districts in January 2012 and February 2013. The researcher visited primary schools in the districts, distributed questionnaires to 379 teachers (about 70% of all primary school teachers in the two districts), and interviewed 30 randomly-selected teachers. The two districts are distant and have differential economic development levels. While one district, by virtue of its proximity to Phnom Penh city, experiences a rapid urbanization and industrialization, the other remains predominantly agricultural. However, the latter has received particular attention from both the government and development agencies in the last two decades. Consequentially, there is no significant difference between them in terms of teacher training. Together, the two districts can provide insightful evidence of how the efforts of both Cambodian government and international development agencies to promote child-centered education have changed classroom instruction. About 40% of the participant teachers aged 30 years or less and those under 40 years old accounted for 62% of all teachers. A small majority of the teachers (54%) had completed 12 years of school education, the educational level currently required for entrance to preservice teacher training colleges while 40% of them possessed only lower secondary education. A few teachers (4%) had a bachelor degree, the highest educational level detected among the participants. About half of the teachers had been trained in the pre-service training curriculum advocating child-centered learning and had been into their teaching for 15 years or less while the other half who had entered their service before the reform learned about child-centered pedagogy in in-service training programs organized by MoEYS. Cambodian teachers were poorly remunerated and worked in resource-strained classrooms. The total amount of an initial base salary and other allowances was about US$80 per month. After 16 years of experience, they would have their salaries increased by 20%. Because they could barely survive on their low income, most of (60%) Cambodian teachers held a second paid job (Benveniste,

40

S. Song / Teaching and Teacher Education 50 (2015) 36e45

Marshall, & Araujo, 2008). Instructional resources in the schools were not abundant. Teachers often had to make do with basic materials such as blackboard and textbooks while other teaching aids were virtually inexistent. Schools even lacked basic utilities like water supply and electricity, let alone the usage of computers and the internet to help with instruction. 4.2. Measures and analytic procedures The questionnaires consisted of two scales: one for teachers' beliefs about instructional pedagogies and the other for classroom practices in mathematics lessons. In each scale, teachers were presented with a mixture of reform and conventional statements about teaching and learning to see how they rated the new ideas against conventional ones. Past studies have shown that teachers may embrace reform ideas at the expense of more conventional ones, or they may adopt the reform ideas and, at the same time, maintain a great deal of the conventional principles (Cohen & Hill, 2001). Van den Akker (1994, p.1492) observes, “implementation implies a process of learning new roles (and often unlearning old ones) for teachers.” These belief and practice statements were constructed based on Cohen and Hill (2001) questionnaire and the researcher's review of relevant educational policies, curriculum guidelines and teacher training documents authorized by the MoEYS (e.g., MoEYS, 2001a, 2001b, 2007a, 2007b, 2008). Teachers' beliefs were measured on a 4-point Likert scale (1 ¼ strongly disagree; 2 ¼ disagree; 3 ¼ agree; 4 ¼ strongly agree) and consisted of 15 items, each of which was either consistent with child-centered pedagogy (10 statements) or conventional didactic approach (5 statements). Some of the items did not necessarily reflect the principles of child-centered education as advocated by constructivist. For example, ‘Pupils learn best when the instruction is conducted in 5 steps’ was treated as an element of child-centered learning. Even though child-centered educators might not agree with this statement, Cambodian instructional policies have emphasized the 5-step lesson structure (e.g., MoEYS, 2001a, 2008). Through an exploratory factor analysis, a subset of 8 items were determined to represent child-centered beliefs (Cronbach's a ¼ .805) and another subset of 4 items to represent conventional beliefs (Cronbach's a ¼ .638) (see Appendix A, Table A1). Teachers' instructional practices in mathematics lessons were measured by a 6-point scale (1 ¼ Never; 2 ¼ less than once a month; 3 ¼ 1e3 times per month; 4 ¼ 1e2 times per week; 5 ¼ 3e4 times per week; 6 ¼ everyday) and consisted of 15 classroom activities. Again, the 15 items ranged from activities commonly observed in conventional classrooms (6 items) to those that were more in line with child-centered pedagogy principles (9 items). Exploratory factor analyses produced a child-centered practice scale consisting of 8 items (Cronbach's a ¼ .847) and a conventional practice scale consisting of 5 items (Cronbach's a ¼ .694) (see Appendix A, Table A2). The interviews were conducted with 30 teachers randomly selected from the two districts. The purpose of the interviews, which lasted about 30e40 min, was to understand what the teachers' daily teaching was like and what teaching methods they thought were best for pupils' learning. Some of the questions that the researcher asked the teachers included the following: what is your teaching like? what kind of teaching methods do you think produce the most learning in pupils? what do you think of childcentered pedagogy? what do you think about its effectiveness relative to teacher-centered instruction? As for analysis, the researcher transcribed the interviews verbatim and read the transcripts several times before coding and analyzing for emerging themes by means of “qualitative content analysis” (Berg, 2001; Graneheim & Lundman, 2004). The researcher started with open coding by assigning each sentence of

the interview texts with a description of two or three words. Then, these descriptions were compared across the interviews and grouped together to form categories using the constant comparative method (Strauss & Corbin, 1990). Finally, the researcher moved to progressively higher level of abstract themes such as textbookbased instruction and classroom constraints. This study contains a major methodological limitation. It measures classroom instruction through teachers' responses on questionnaire and in interviews. Such self-reported practice might be vulnerable to self-defensive representation of what really happens in the classroom. Although the current researcher was highly cautious to develop good rapport with the teachers during the survey, it is advisable that similar studies should supplement their data with classroom observation or videotaping of instruction so as to better understand classroom life. 5. Findings 5.1. Teachers' beliefs about teaching and learning The questionnaire data show that teachers expressed a strong support for belief statements associated with child-centered pedagogy (reform beliefs). As shown in Fig. 2, none of the average ratings for child-centered pedagogy statements fell below 3 (Agree). Of particular note, Statement 2, “Pupils learn best when instruction is conducted in 5 steps,” received the highest approval, indicating that the 5-step lesson structure had gained legitimacy in Cambodian primary school classrooms. “Higher-order questions”, a core element of child-centered learning, obtained the least support from the teachers as compared to other reform ideas. Although average scores across items in each of the two scales reveal a significant difference between belief statements denoting child-centered pedagogy and those denoting conventional didactic pedagogy (t ¼ 17.91, p < .001), with teacher support leaning toward the former, Fig. 2 shows that reported average rating of each item for the latter was also above the middle value, 2.50, of the 1 to 4 Likert-type scale. This finding suggests that teachers still held dear to conventional teaching principles. In addition, “memorization” was still highly valued as an effective means to teaching and learning, a belief that contrasted reformers' intention that was explicitly expressed in the curriculum guidelines and teacher training materials. In summary, the questionnaire data on teachers' beliefs show that teachers' overwhelming support for child-centered pedagogy

Fig. 2. Teachers' beliefs about teaching and learning (N ¼ 379).

S. Song / Teaching and Teacher Education 50 (2015) 36e45

did not necessarily lead them to shed their conventional beliefs. Rather, they slightly suppressed their support for but remained essentially faithful to those conventional ideas. Unlike policymakers, teachers saw child-centered pedagogy not as a change but as a supplement to their existing approach. Teachers' preferences to combine the principles of conventional and reform pedagogies were well represented in the interviews. However, it appeared that teachers tended to cite different reasons for blending the two approaches. Some said that they used the conventional approach if the topic was new to the pupils and reserved child-centered pedagogy for familiar topics: [The two approaches] are useful in their respective ways based on the circumstances. In some lessons, pupils do not know anything. So, we use teacher-centered method. Then, pupils can learn from us. But in the lessons that the pupils have some basics from pervious lessons, we just lead and give them little explanation and let them work on their own. (Interview #109, p.3) Other teachers said they used child-centered pedagogy with difficult tasks and the conventional approach with easy tasks, as expressed in the following: I use child-centered approach for tasks that require pupils to discuss such as when they write a paragraph or solve a difficult problem … I used teacher-centered method with short word problems or computational exercises. They are quick and easy activities. (Interview #202, p.2) Teachers' beliefs about which teaching approaches worked best for pupils' learning also varied with subject matters, although the teachers seemed to be inconsistent in their choicesdwhich pedagogy should go with which subject. Teachers saw either childcentered pedagogy or conventional approach beneficial to them in different ways so, rather than discarding one, they embraced both. This belief again tended to conflict with the official guidelines, which instructed that child-centered pedagogy worked with pupils of all ability levels and subject matters. 5.2. Classroom practices The results on classroom practices show a contradicting picture to those on teachers' beliefs. Overall, teachers reported that their pupils were engaged in conventional classroom activities more frequently than in child-centered activities. Fig. 3 shows that the most common activity in mathematics classrooms was to have

Fig. 3. Teachers' self-report of classroom practices in mathematics lessons (N ¼ 379).

41

pupils to solve computational exercises on the blackboard, which was done on a daily basis. “Getting pupils to memorize formula and rules” had the lowest probability of occurrence compared to other conventional activities with an average rating of 4.90 on the frequency scale of 1e6. Even this least common conventional activity, however, tended to occur almost every day and more frequently than any one of the child-centered activities, the highest average rating of which is 4.65 for “working in small groups”. Lower in order of frequency were child-centered activities like “solving word problem” (M ¼ 4.59), “discussing different ways to solve problems” (M ¼ 4.57), “using materials” (M ¼ 4.53), “explaining other pupils how to solve a problem” (M ¼ 4.44), and “making conjectures” (M ¼ 4.32). “Working on problems with multiple solutions” and “posing problems for other pupils to solve” were the least common activities to be utilized in classrooms, particularly in comparison with conventional activities. The difference between childcentered and conventional practices, as measured by average scores across items in each scale, was highly significant (t ¼ 20.11, p < .001). This evidence shows clearly that the conventional didactic approach predominated over the child-centered pedagogy in Cambodian primary school classrooms. Teachers' description of their actual instruction suggested that their teaching was heavily dependent on the blackboard and textbooks. Teachers often put some problems on the blackboard and solved one of them as an illustration, before calling a few pupils one at a time, starting with the most able, to come to the blackboard. Sometimes, the teachers left some time for pupils to work at their seats before calling on them. The following teacher described how he arranged his classroom tasks in a mathematics lesson: “I write some problems on the blackboard and ask the pupils to solve them in their notebooks. Then, I ask the one who does it correctly to write the answer on the blackboard” (Interview #105, p.4). Textbooks served as the main source for both class work and homework, which composed mainly of computational exercises. Another teacher noted, “if there are too many exercises in the textbooks, I assign some for homework after they do some here [in the classroom]” (Interview #212, p.2). In the past, when textbooks were less available, teachers would write the lesson on the blackboard and get pupils to repeat, read, and copy it down onto their notebooks. Reading, memorizing, and reciting were the common practices. Now, with the textbooks in their hands, pupils were free from taking extensive notes and had more time for practice. Also with the textbooks, it was easier for pupils to read by themselves and work in groups with little control from the teachers. Teachers tended to perceive the use of textbooks by pupils as an improvement to rote learning: “now, it is good that the pupils have the textbooks for all subjects. They might have studied the lesson beforehand at home. So, when we put them in groups, the clever pupils can use the textbooks to help the weaker ones” (Interview #108, P.4). There is nothing wrong with the use of textbooks, but when classroom life is solely, or virtually so, constituted of what is in the textbooks, pupils' creativity might be curtailed. In fact, teachers tended to equate the term “pupils' activities” or “pupils' participation”, a core principle of child-centered pedagogy, with questions or exercises from textbooks; to them, encouraging pupils' activities simply meant making pupils work with textbooks and seemed to have nothing to do with problem solving, as one of the teachers put it: Every lesson has exercises for practices … These are their activities … If we just copy the lesson onto the blackboard and ask pupils to copy onto their notebooks, they don't know what the lesson is all about. So we have to get them to do activities like

42

S. Song / Teaching and Teacher Education 50 (2015) 36e45

answering questions or doing exercises. (Interview #216, pp.2e3) In a way, Cambodian teachers believed that their instructional practices had moved away from the teacher-centered tradition and had developed into a novel form of instruction which they interpreted to be child-centered. This locally-molded trajectory of childcentered pedagogy instruction in Cambodian classrooms was characterized by “pupils' working in groups (or individually) on problems from textbooks.” 5.3. Constraints on instructional reform implementation The results presented so far testified that teachers expressed strong beliefs in the effectiveness of child-centered pedagogy. Yet, their reported classroom instruction was largely occupied by activities more closely resembling those of the conventional teaching approach. Why did teachers' support for child-centered pedagogy fail to materialize in classrooms? Analysis of teachers' interviews unearthed four major constraints at classroom level that helped answer this question. The most often cited barrier to implementing child-centered pedagogy was the difference in pupils' ability. Teachers reported that the gap in pupils' ability especially at the higher grades of primary schools was quite astonishing, “Some of my pupils are too weak. They are in sixth grade now but they have not memorized the vowels and consonants yet” (Interview #201, p.4). This ability gap also undermined the effectiveness of collaborative work: “When we give them work to do in groups, only the smart pupils will do and the weak will not help” (Interview #203, p.3). Another teacher echoed, “in a group of five pupils, three will think while the other two will just play. They don't help each other. So, instead, I ask them to work individually. I do not use group work” (Interview #214, p.2). The second constraint was over-crowdedness. Large classes made it hard for the teachers to control pupils' behaviors during group work. One of the teachers said, “I have more than 40 pupils, so when I use child-centered pedagogy the pupils will be very noisy and out of control.” With many pupils under their charge, teachers also had difficulty monitoring pupils' learning because the output of group work submitted to the teachers did not necessarily mean that every member had contributed to the task. This was evident in this teacher's comment: We cannot oversee them all. There are too many pupils. In group work, only some will do the task but all will get the same marks. So I get them to work individually and I spend a lot of time checking their work. (Interview #211, p.6e7) The third constrain was the scarcity of teaching resources. Both teachers and policymakers considered teaching materials a core component of child-centered pedagogy. But teachers were provided with only textbooks and teacher manuals and were supposed to develop other instructional materials using local resources. However, teachers barely had any time for producing materials because they had to do a second job to supplement their small salary. “If they [the teachers] don't have enough to eat, they don't have time for developing materials”(Interview #213, p.3), commented a senior teacher who reported that she was not engaged in other jobs but had to take care of her two grandchildren, whose parents were both working. Adding to the quite common cases of involvement in side jobs, the case presented here shows how low income forced some teachers to be caretakers of their own children or grandchildren alongside their teaching job, as evidenced in the case where teachers brought their young to their classrooms (3 out

of the 30 teachers interviewed). This would translate to a reduction in time for both preparationdproducing teaching materialsdand teaching. The fourth factor that discouraged teachers from exploiting child-centered pedagogy was the fact that there were too many contents to be covered in the curriculum. Teachers had to align their teaching pace with the schedule designated by MoEYS to achieve uniform progression through the curriculum contents. Teachers complaint that child-centered pedagogy activities usually took longer time to complete and would put them behind the schedule: “child-centered activities take too much time. I will not be able to follow the syllabus. I will be behind the schedule.” The requirement for uniform progression posed a big challenge for teachers who had to teach low ability classes, which was quite common, particularly in rural areas. 6. Discussion 6.1. Inconsistency between teachers' beliefs and practices This study found that Cambodian primary school teachers enthusiastically embraced the principles of child-centered pedagogy. Yet, their strong support for this reform pedagogy was not accompanied by fundamental changes in their classroom instruction. This finding is inconsistent with prior research (Leatham, 2006; Stipek, Givvin, Salmon, & MacGyvers, 2001; Wilkins, 2008). Note that these previous studies investigate how variation in teachers' beliefs about child-centered pedagogy relates to variation in their classroom practices and do not include teachers' beliefs about and practices of conventional teaching approach in their analyses. In other words, they investigate only one side of the story and do not compare and contrast the conventional and reform beliefs or the conventional and reform practices. Therefore, the documented conclusion that teachers' beliefs about and practice of child-centered pedagogy are significantly correlated does not necessarily mean that the magnitude of child-centered classroom practices is higher than that of traditional practices. The evidence of this study shows that the reverse is more likely to be the case: despite teachers' overwhelming support, only some selected childcentered practices have started to take root while activities associated with conventional teaching still dominate the classrooms. The inconsistency of beliefs and practices, however, corroborates a number of other previous studies (Kennedy, 2005; Raymond, 1997; Schuh, 2004). Schuh (2004) points out that teachers' practices are observed to align strongly with teachercentered teaching even though the teachers themselves are faithful to child-centered principles in their beliefs. This finding loudly echoes Argyris and Schon (1974) claim that an individual's “espoused theory” is often incompatible with his/her “theory-inuse”. The remainder of this section will discuss why this phenomenon happens. 6.2. Mismatch between teachers' interpretation and policy intention Embedded in the results shown in this study is the mismatch between teachers' interpretation of child-centered instruction and reformers' intentions. Two examples would suffice to understand this phenomenon, one related to group work and the other related to pupils' activities. For group work, reformers expect that working together in small groups will help build classroom community and allow pupils to help each other so that ultimately everybody will learn. However, teachers perceive group work to be a barrier to learning, at least for some pupils, because they cannot ensure that every member of the groups is engaged due to the crowded

S. Song / Teaching and Teacher Education 50 (2015) 36e45

classrooms. Moreover, their experiences inform them that only the more able pupils benefit from group work. With regard to pupils' activities, child-centered approach policies dictate that instruction should emphasize activities that require higher-order thinking skills such as problem solving and critical thinking while teachers take pupils' activities/participation to mean solving computational exercises from textbooks, which usually require little more than application of rules and procedure. Teachers claim that they are adopting child-centered pedagogy, but their reported classroom practices consist mainly of getting pupils to work on exercises based on textbooks, nothing corresponding to reformers' proposals for problem solving and critical thinking. This evidence shows that specialized terms that policymakers use to express deep learning can become something superficial when teachers translate them into classroom practice. This finding is particularly plausible for the case of Cambodia, where these specialized terms are sifted through a cascade model of teacher training. Through this model, instructional policies lose their potentials before reaching classrooms. Similar finding is also reported by other researchers on relationship between educational policy and practice in Cambodia (Courtney, 2008; Ogisu, 2013). Ogisu (2013), in an analysis on Cambodian teachers' understanding of policy documents on child-centered approach, suggests that teachers misinterpret policy messages because of the contradiction inherent in the reform itself, which “tries to make teaching and learning more bottom-up and relevant through a highly top-down implementation process.” The superficial take-ups of reform ideas reported here also provide credence to research on educational reform implementation that adopts sense-making perspective, especially the “zone of enactment” theory (e.g, Coburn, 2001; Hill, 2001; Spillane, 1999). The pedagogical ideals advocated by the reform might be so much different from the existing teaching models in Cambodia that, coupled with the lack of professional development, teachers cannot understand them critically (Spillane, 2004). As a result, they select and put into practices only some superficial aspects of the reform that are compatible with their experiences, knowledge, and beliefs. 6.3. Prevailing influence of classroom realities While teachers agree with the philosophy of child-centered pedagogy, they tend to question the applicability of this teaching and learning approach. They believe that they cannot apply childcentered pedagogy because their classrooms are constrained by a number of factors including over-crowdedness, differences in pupils' ability, scarcity of teaching resources, and over-loaded contents. Teachers' concerns over these immediate classroom conditions tend to overweigh their expressed beliefs about childcentered instruction and prevent them from translating those beliefs into classroom practices. This finding falls neatly with Rogan and Grayson (2003) “zone of feasible innovation” theory which posits that school resources and environment significantly influence teacher's implementation of educational innovation. Previous studies in other developing countries (O'Sullivan, 2002, 2004, for Namibia; Sriprakash, 2010, for India; Wang, 2011, for China) have also reported similar obstructing classroom environments. Policies to reform pedagogy are often made to change teachers' teaching skills and, particularly in developing countries, often not accompanied by measures to improve their teaching environment that hinders implementation of new pedagogies. Kennedy (2005) finds that teachers must consider a wide variety of immediate instructional issues such as time constraints, maintenance of classroom order, and pupils' willingness to participate, none of which is addressed in reformers' agendas, which focus more on administrative reforms. Policymakers announce policies, publish guidelines,

43

send them to schools, and expect teachers to implement them. Teachers, however, are not thinking along the same lines. They are consumed with how to supplement their meager income and how to deal with the poor classroom conditions, leaving them little time to study the ministry guidelines. Reform ideals often have to compete against local realities for predominance in classrooms; in developing countries, where teachers lack support of various kinds, distant ideals are often outweighed by immediate realities. 7. Conclusion Reform ideas often come attached to aid and loans. This study traces the development of instructional reform in Cambodia in the last two decades and shows that international development agencies, particularly UNICEF and ADB, have substantially influenced the child-centered instructional policiesdthe curriculum reform in 1996 and Child-Friendly Schools Policy in 2007dand implementation process of those policies. The influences are reflected in the joint efforts of MoEYS and these development partners to design instructional policies, develop new curriculum materials, and train teachers both at teacher training centers and in-service training workshops. These joint efforts have been able to convince teachers of the importance of child-centered pedagogy to pupils' learning. What has remained significantly unaffected, however, is the efforts' ultimate goal: classroom instructiondclassroom instruction that is characterized by a combination of conventional and new practices but heavily leans towards the former. Textbooks continue to be the main source of classroom activities, consisting barely anything more than computational exercises and on-theline questions, and the blackboard continues to be the dominant material with which pupils work, while problem solving and the use of other learning resources are yet to be part of daily instruction. Cambodian primary school teachers do not translate their strong support for child-centered pedagogy into classroom practices. This failure to act on their beliefs stems from two reasons: on the one hand, teachers take up the reform only at face values without critically examining its meanings and, on the other hand, their teaching is heavily constrained by unfavorable classroom environments such as crowdedness, pupils' uneven ability, and scarcity of resources. While the findings presented in this study speak best for Cambodian case, they are by no means unique to Cambodia. Many countries are introducing modern pedagogies and, especially developing countries as late adopters, are struggling to make those pedagogies work in their classrooms. This study provides evidence that it is not always the case that policy influences teachers' instruction; conversely, teachers exert considerable influence on instructional policies. They adapt, combine, shift, and adopt policy to fit with their workplace constraints. Many educational innovations fail because they ignore this important fact. Pronouncing policies is not enough. A new instructional policy is just another pressure to the teachers, who are working in an increasingly changing environment and are already coping with various kinds of reforms. Improving teachers' expertise through professional development programs produces less than satisfactory results since teachers find nowhere they can use the new skills. These efforts will not be effective unless they are accompanied by measures to remove constraints at teachers' workplace. Decreasing the number of students in each class and hiring assistant teachers to help share classroom burden are some promising solutions to help teachers adopt the new pedagogy more effectively. Improving teachers' incentives will also help teachers to concentrate their time and energy on teaching.

44

S. Song / Teaching and Teacher Education 50 (2015) 36e45

Appendix A

Table A1 Items measuring teachers' beliefs about teaching and learning. Pupils learn best … A. Beliefs associated with child-centered pedagogy (reform beliefs) (Cronbach's a ¼ .805) 1. When teachers ask a lot of higher-order questions. 2. When they work in small groups (4e6 people). 3. When learning contents are suitable with pupils' ability level. 4. When the teacher continuously assess pupils' progress. 5. When the contents of the lesson is connected to everyday life. 6. In classrooms with a lot of study games. 7. In classrooms where their active participation in learning is encouraged. 8. When instruction is conducted in 5-steps. B. Conventional beliefs (Cronbach's a ¼ .638) 1. When the lessons are primarily based on lectures. 2. In classroom with pupils sitting orderly and no mobility. 3. When the MoEYS-designated syllabus is strictly followed. 4. When they memorize a lot.

Table A2 Items measuring frequency of occurrence of classroom activities (classroom practices). In this academic year, how often did the pupils in your mathematics classes do the following? A. Activities associated with child-centered pedagogy (reform activities) (Cronbach's a ¼ .847) 1. Create exercises/word problems for other pupils to solve. 2. Do problems that have more than one correct solution. 3. Make conjectures. 4. Explain other pupils about how to solve a problem. 5. Use manipulative materials to solve problems. 6. Discuss different ways that they solve particular problem. 7. Solve word problems. 8. Work in small groups on math problems. B. Conventional activities (Cronbach's a ¼ .694) 1. Memorize formula or rules to solve problems. 2. Work individually on problems from textbook. 3. Solve exercises on slates. 4. Recite the multiplication table. 5. Solve exercises on blackboard.

References Anderson-Levitt, K. (Ed.). (2003). Local meanings, global schooling: Anthropology and world culture theory. N.Y: Palgrave. Anorve, R. F. (1999). Reframing comparative education: the dialectic of the global and local. In R. F. Anorve, & C. A. Torres (Eds.), Comparative Education: The dialectic of the global and the local (pp. 1e23). Lanham, Maryland: Rowman & Littlefield. Argyris, C., & Schon, D. (1974). Theory in practice: Increasing professional effectiveness. San Francisco: Jossey-Bass. Asian Development Bank. (1996). Cambodia: Education sector strategy study. Manila: ADB. Ayres, D. (2000). Anatomy of a crisis: Education, development, and the state in Cambodia, 1953e1998. Honolulu: University of Hawaii Press. Benveniste, L., Marshall, J. H., & Araujo, M. C. (2008). Teaching in Cambodia. Phnom Penh, Cambodia: The World Bank. Berg, B. L. (2001). Qualitative research methods for the social science (4th ed.). MA: Allyn & Bacon. Brodie, K., Lelliott, A., & Davis, H. (2002). Forms and substance in learner-centred teaching: teachers' take-up from an in-service programme in South Africa. Teaching and Teacher Education, 18, 541e559. Brook Napier, D. (2003). Transformations in South Africa: policies and practices from ministry to classroom. In K. Anderson-Levitt (Ed.), Local meanings, global schooling: Anthropology and world culture theory (pp. 51e74). N.Y: Palgrave. Bunroeun, N. (1999). Discussion Paper. Teacher training systems in Cambodia (Vol. 78, pp. 1e52). Nagoya University. Carnoy, M., & Rhoten, D. (2002). What does globalization mean for educational change? A comparative approach. Comparative Education Review, 46(1), 1e9. Chisholm, L., & Leyendecker, R. (2009). Curriculum reform in sub-saharan Africa: when local meets global. In R. Cowen, & A. M. Kazamias (Eds.), International

handbook of comparative education (pp. 685e702). N.Y: Springer Science þ Business Media. Coburn, C. E. (2001). Collective sensemaking about reading: how teachers mediate reading policy in their professional communities. Educational Evaluation and Policy Analysis, 23(2), 145e170. Coburn, C. E. (2006). Framing the problem of reading instruction: using frame analysis to uncover the microprocesses of policy implementation. American Educational Research Journal, 43(3), 343e349. Cohen, D. K., & Ball, D. L. (1990a). Policy and practice: an overview. Educational Evaluation and Policy Analysis, 12(3), 233e239. Cohen, D. K., & Ball, D. L. (1990b). Relations between policy and practice: a commentary. Educational Evaluation and Policy Analysis, 12(3), 331e338. Cohen, D. K., & Hill, H. C. (2001). Learning Policy: When state education reform work. New Haven: Yale University Press. Courtney, J. (2008). Do monitoring and evaluation tools, designed to measure the improvement in the quality of primary education, constrain or enhance educational development? International Journal of Educational Development, 28, 546e559. Courtney, J., & Gravelle, M. (2014). What makes the difference? An analysis of a reading intervention programme implemented in rural schools in Cambodia. Compare: A Journal of Comparative and International Education, 44(3), 416e434. Gentner, D., Rattermann, M., & Forbus, K. (1993). The roles of similarity in transfer: separating retrievability from inferential soundness. Cognitive Psychology, 25(4), 524e575. Graneheim, U. H., & Lundman, B. (2004). Qualitative content analysis in nursing research: concepts, procedures and measures to achieve trustworthiness. Nurse Education Today, 24, 105e112. Hill, H. C. (2001). Policy is not enough: language and the interpretation of state standards. American Educational Research Journal, 38(2), 289e318. Isikoglu, N., Basturk, R., & Karaca, F. (2009). Assessing in-service teachers' instructional beliefs about student-centered education: a Turkish perspective. Teaching and Teacher Education, 25(2), 350e356. Kennedy, M. (2005). Inside teaching: How classroom life undermines reforms. Cambridge, MA and London, England: Harvard University Press. Leatham, K. R. (2006). Viewing mathematics teachers' beliefs as sensible systems. Journal of Mathematics Teacher Education, 9(1), 91e102. Lipsky, M. (2010). Street-level bureaucracy: Dilemmas of the individuals in public services. NY: Russell Sage Foundation. McCormick, A. (2012). Whose education policies in aid-receiving countries? A critical discourse analysis of quality and normative transfer through Cambodia and Laos. Comparative Education Review, 56(1), 18e47. McLaughlin, M. W. (1990). The rand change agent study revisited: macro perspectives and micro realities. Educational Researchers, 19(9), 11e16. McLaughlin, M. W. (1987). Learning experience: lesson from policy implementation. Educational Evaluation and Policy Analysis, 9(2), 171e178. Meyer, J. W., & Ramirez, F. O. (2000). The world institutionalization of educationdorigins and implications. In J. Schriewer (Ed.), Discourse formation in comparative education (pp. 111e132). Frankfurt/M, Germany: Lang. Meyer, J. W., & Ramirez, F. O. (2010). Human rights in social science textbooks: cross-national analyses, 1970e2008. Sociology of Education, 83(2), 111e134. MoEYS. (1994, January). Rebuilding quality education and training in Cambodia. Phnom Penh: Author. MoEYS. (1997). Textbooks and teacher guides utilization master plan, 1997e2001. Phnom Penh: Author. MoEYS. (2001a). eksar somrap vaek damrong tis: kar preu pras kamvithi ning sievphov siksa thmey thnak ti 6 [Document for teacher orientation training: Utilization of the new curriculum and textbook for Grade 6]. Phnom Penh: Author. MoEYS. (2007a). Child-friendly school policy. Phnom Penh: Author. MoEYS. (2001b). sievphov kruu: Kaenitvityia thnak ti 6 [Teacher manual: Mathematics for grade 6]. Phnom Penh: Author. MoEYS. (2007b). Child friendly schools (CFS) Master plan 2007e2011. Phnom Penh: Author. MoEYS. (2007c). eksar kamvithi sala komar metrey [Documents for Child-Friendly School Program]. Phnom Penh: MoEYS, Teacher Training Department. MoEYS. (2007d). eksar kaomtroo kar anuvaot kamvithi sala komar metrey: prosithipheap nei kar bangrien ning rien [Support documents for Child-Friendly School Programme: Effective teaching and learning]. Phnom Penh: Author. MoEYS. (2008). Korl vithi sis machhaekmondol [Student-centered approach]. Phnom Penh: Author. MoEYS. (2012). korl noyorbay sdeipi sievpov siksa ning maekutes somrap kar aphivaot sievphov siksa korl [Policy and guidelines for core textbooks development]. Phnom Penh: MoEYS, Department of Curriculum Development. Ogisu, T. (2013). Cambodian instructional reform based on teachers' views of student-centered approaches. Comparative Education: Bulletin of the Japan Comparative Education Society, 47, 79e99. O'Sullivan, M. C. (2002). Reform implementation and the realities within which teachers work: a Namibian case study. Compare: A Journal of Comparative and International Education, 32(2), 219e237. O'Sullivan, M. C. (2004). The reconceptualization of learner-centered approaches: a Namibian case study. International Journal of Educational Development, 24, 585e602. Popov, O. (1998, May). Curriculum development project in Cambodia (Draft Report). Phnom Penh: Sida. Raymond, A. M. (1997). Inconsistency between a beginning elementary school

S. Song / Teaching and Teacher Education 50 (2015) 36e45 teacher's beliefs and teaching practice. Journal of Research in Mathematics Education, 28(5), 550e576. Rogan, J. M. (2007). An uncertain harvest: a case study of implementation of innovation. Journal of Curriculum Study, 39(1), 97e121. Rogan, J. M., & Grayson, D. J. (2003). Towards a theory of curriculum implementation with particular reference to science education in developing countries. International Journal of Science Education, 25(10), 1171e1204. Royal Government of Cambodia. (1993). The constitution of the Kingdom of Cambodia. Phnom Penh: Royal Government of Cambodia. Schuh, K. L. (2004). Learner-centered principles in teacher-centered practices? Teaching and Teacher Education, 20(8), 833e846. Schweisfurth, M. (2011). Learner-centred education in developing country contexts: from solution to problem? International Journal of Educational Development, 31, 425e432. Spillane, J. P. (1999). External reform initiatives and teachers' efforts to reconstruct their practice: the mediating role of teachers' zones of enactment. Journal of Curriculum Studies, 31(2), 143e175. Spillane, J. P. (2004). Standard deviations: How schools misunderstand education policy. Cambridge: Harvard University Press. Spillane, J. P., Reiser, B. J., & Reimer, T. (2002). Policy implementation and cognition: reframing and refocusing implementation research. Review of Educational Research, 72(3), 387e431. Sriprakash, A. (2010). Child-centred education and the promise of democratic learning: pedagogic messages in rural Indian primary schools. International Journal of Educational Development, 30, 297e304. Stein, M. K., & Brown, C. (1997). Teacher learning in a social context: Integrating collaborative and institutional processes with the study of teacher change. In

45

E. Fennema, & B. S. Nelson (Eds.), Mathematics teachers in transition. Mahwah, NJ: Lawrence Erlbaum. Steiner-Khamsi, G., & Stolpe, I. (2006). Educational import: Local encounters with global forces in Mongolia. N.Y.: Palgrave. Stipek, D. J., Givvin, K. B., Salmon, J. M., & MacGyvers, V. L. (2001). Teachers' beliefs and practices related to mathematics instruction. Teaching and Teacher Education, 17, 213e226. Strauss, A., & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and techniques. Newbury Park, CA: Sage. Teddlie, C., & Tashakkori, A. (2009). The foundations of mixed methods research: Integrating quantitative and qualitative techniques in the social and behavioural sciences. Thousand Oaks, CA: Sage Publications. UNICEF. (2009). Manual: Child friendly schools. New York: UNICEF. Van den Akker, J. (1994). Designing innovations from an implementation perspective. In T. Husen, & T. N. Postlewaite (Eds.), International encyclopedia of education (pp. 1491e1494). Oxford, England: Elsevier Science. Van den Berg, R. (2002). Teachers' meanings regarding educational practice. Review of Educational Research, 72(4), 577e625. Wang, D. (2011). The dilemma of time: student-centered teaching in rural classroom in China. Teaching and Teacher Education, 27, 157e164. Weatherley, R., & Lipsky, M. (1977). Street level bureaucrats and institutional innovation: Implementing special education reform. Harvard Educational Review, 47, 171e197. Wilkins, J. M. (2008). The relationship among elementary teachers' content knowledge, attitudes, beliefs, and practices. Journal of Mathematics Teacher Education, 11, 139e164.