Revisiting Classroom Practices in East Asian Countries: Examination of Within-Country Variations and Effects of Classroom Instruction

by Yoonjeon Kim - 2018

Background/Context: East Asian schools receive much attention for the comparatively high achievement of their students. To account for this success, scholars and commentators advance broad claims about the rote character of instruction or the complexity of classroom practice, typically generalizing to an entire nation. Yet little is known about the variation in classroom practices within East Asian countries and how classroom organization affects student achievement.

Purpose/Objective/Research Question/Focus of Study: This study extends the previous literature on East Asian classrooms by considering the heterogeneity of classroom organization within societies. It focuses on four aspects of classroom instructional practice: complex instruction, procedural instruction, teacher-centered instruction, and student-centered instruction. This study asks the following research questions: (1). To what extent do classroom instructional practices in East Asian countries differ in terms of overall prevalence and within-country variation, compared with to practices found in other nations? (2). How are classroom instructional practices associated with student achievement within East Asian countries, controlling for student, classroom, and school background variables?

Research Design: Drawing from the Trends in International Mathematics and Science Study (TIMSS) 2007 data, I examine how the country means and within-country variation of the four aspects of classroom instructional organization in five East Asian countries—Chinese Taipei, Hong Kong, Japan, Korea, and Singapore—compare with those in the other 45 nations in the sample. Then, I focus on two particular East Asian countries that display vastly different school structures, Japan and Singapore, to examine how classroom practices covary with student achievement within these nations.

Findings/Results: East Asian classrooms do tend to be more intensely teacher- centered and display less complexity than in other nations on average. But classrooms with more complex and student-centered instruction within East Asian societies display higher achievement; an opposite association is found when comparing between-country relationships worldwide. At the same time, these positive effects observed in East Asia diminish when characteristics of schools and the social- class backgrounds of students are taken into account, similar to patterns long observed in the West.

Conclusions/Recommendations: While classroom practices prevalent in East Asian countries are often celebrated as predictive of stronger achievement—or criticized for their rigidity and not importable to the West—these findings reveal greater variability than previously understood and suggest that classroom practices interact with social- class backgrounds and student achievement in more complex ways. And East Asian nations face educational challenges similar to those observed in the United States and other developed countries. Once we acknowledge the commonality as well as the differences, cross-national research would allow us not only to better understand perennial educational problems and the assumptions we hold about classroom practices, but also inform valid implications for policy and practice.

INTRODUCTION

The persisting success of East Asian students in international assessments has triggered much curiosity about their schools and the social organization of their classrooms (Hiebert et al., 2003; Stevenson & Stigler, 1992; Stigler & Hiebert, 1999). Generalities about allegedly nationwide teaching practices and instructional organization abound. Secondary classrooms in East Asia, for example, are often characterized as having teachers who use more didactic approaches, placing students in passive roles, or emphasizing procedural knowledge and whole-class instruction (Huang & Leung 2004; Watkins & Biggs 1996). These characteristics contrast sharply with practices believed to be conducive to learning in the West, which emphasize complex and real-world problem solving, conceptual understanding, and active student participation and reflection. Often, the presumably teacher-centered practices of East Asian countries are interpreted as driving the high performance of their students, to be copied elsewhere, or criticized for not engaging students in more participatory and engaging ways. But using national averages of educational characteristics (e.g., national averages of teacher-centered instruction) without consideration of the variability within nations may lead to superficial interpretations of classrooms and teaching practices in other countries (Baker, 1997; Ercikan et al., 2015; Torney-Purta & Amadeo, 2013).

Cross-national qualitative studies have long offered a deeper understanding of the historical, social, and cultural value and meaning of the seemingly rote instructional practices prevalent in East Asian classrooms (An et al., 2004; Ma, 1999; Marton et al., 1996; Tobin et al., 2009; Wang & Paine, 2003). They elaborate on how teachers make sense of their instruction, how the instruction is received, and, in turn, how it may contribute to high performance. Still, this work falls short of considering variation in classroom instruction that may exist within a country or of testing the presumed effect of instructional practice on achievement.

This study addresses this gap by using the 2007 Trends in International Mathematics and Science Study (TIMSS) to delve into (1) the within-nation variability in the social organization of instruction among classrooms in East Asia, in addition to the country average features, and (2) the extent to which complexity or regimented simplicity accounts for higher student achievement within the East Asian countries, controlling for student, classroom, and school characteristics.

The study shows that when classrooms in East Asian countries are compared with classrooms in other countries, East Asian classrooms are less likely to incorporate complex tasks, such as solving problems with no obvious solution or relating learning to daily life, and more likely to have teacher-centered, lecture-style instruction on average. However, the deployment of these practices varies considerably across classrooms within each country. Within country, classrooms with more complex and student-centered instruction show higher achievement; this is in contrast with the associational pattern observed between countries. In addition, I find that within-country associations diminish or disappear when the socioeconomic status of students and broader characteristics of schools are taken into account. These findings yield a more textured understanding of how variability in instructional organization ties to achievement within East Asian nations and how these relationships are conditioned by features of classroom and school structure as well as social class stratification of families.

REVIEW OF THE LITERATURE

This section reviews three streams of literature on classroom organization and teaching practices in East Asia. Earlier studies, based on either observation or large-scale data, focused on comparing gross national patterns of classroom practices. Other qualitatively oriented studies were geared toward more complex and contextualized descriptions of schooling and the cultural meaning underlying surface differences. Finally, a smaller set of quantitative studies began exploring the link between classroom practices and student achievement in East Asian countries. These three lines of work inform us on East Asian classroom instruction at different levels but rarely intersect with each other. Informed by this body of work, this article attempts to extend the line of inquiry by investigating the variation and effect of classroom practices within East Asian countries.

 

GROSS NATIONAL COMPARISON OF CLASSROOM TEACHING

Many earlier studies of East Asian schools were prompted by the superior academic performance of East Asian students on international measures. Researchers, often led by Western scholars, have focused on distinct national patterns of classroom organization and pedagogy in East Asian classrooms (Hiebert et al., 2003; Schmidt et al., 1996; Stigler & Hiebert, 1999). The primary goal of these observational studies was to identify the weaknesses of schooling in the United States. Teaching patterns that are rarely observed in American classrooms but are present in East Asian classrooms were pointed out as potential areas of improvement for American schools. Harold Stevenson, James Stigler, and their associates cross-national studies, which include Japanese, Taiwanese, German, and American classrooms, take this gross national comparison approach (Stevenson & Stigler, 1992; Stigler & Hiebert, 1999). They highlight behavioral patterns in Japanese or Taiwanese classrooms, such as the prevalence of whole-class academic activities in elementary classrooms or devotion to recitation and memorization strategies, and contrast them with those in American classrooms. The observed cross-national differences are assumed to stem from distinct national cultures that are believed to be relatively stable across countries and homogenous within countries (Stevenson & Stigler, 1992; Stigler & Baranes, 1988).

Using a similar gross comparison approach, other researchers take the unit of analysis up to the cross-cultural level and compare schooling in Eastern and Western countries (Leung, 2001; Li, 2002; Tweed & Lehman, 2002; Watkins & Biggs, 1996). The distinctive schooling patterns and the high achievement in East Asian countries are attributed to their shared cultural and historical backgrounds, including Confucian cultural heritage (Leung, 1995) or the long history of examination systems (Wong, 2004; Zeng, 1999). Studies that employ the EastWest comparative framework assume that different cultural models of teaching and learning work in these different cultural systems. Dahlin and Watkins (2000) reported that East Asian students tend to think of understanding as a long process that requires mental effort on the part of the learner, whereas Western students tend to view learning as a process of sudden insight and attribute academic success to less controllable factors such as inherent ability. Other scholars posit that the national examination system, which developed from the civil servant screening system in ancient China and was later adopted by other East Asian nations, has a greater impact on the high achievement of East Asian students than Confucian culture (Li, 2003; Wong, 2004). It is also argued that overemphasis on memorization and rote learning in classrooms came as a result of preparing students for the highly competitive examinations (Huang & Leung, 2004; Zeng, 1999).

International surveys such as the TIMSS and Progress in International Reading Literacy Study (PIRLS), which collect information on classrooms and teaching practices, have allowed scholars to make gross national comparisons of instructional practice for a broader range of countries (Desimone et al., 2005; Schmidt et al., 2001; Stevenson & Baker, 1991), often with the direct intention of understanding the high achievement of East Asian countries (Leung, 2002; Shen, 2005). Based on cross-national comparison of various features of national educational systems, William H. Schmidt and his colleagues found that instructional topics tend to be more scattered in American classrooms, leading to disconnected learning opportunities for students (Schmidt et al., 1999; Schmidt et al., 2001). By contrast, classroom practices in high-performing East Asian countries such as Japan and Korea were reported to be more coherent with the national curriculum (Schmidt et al., 1999). Some studies that intended to do an EastWest comparison using TIMSS data failed to find a consistent set of variables that explained the high achievement of East Asian countries and ended up resorting to cultural effect (Leung, 2002; Shen, 2005). Other studies found that country average classroom practices were quite similar in terms of time spent on various instructional activities (LeTendre et al., 2001) and the emphasis on conceptual or computational instruction (Desimone et al., 2005).

COMPLEX AND CONTEXTUALIZED PORTRAYALS OF EAST ASIAN CLASSROOMS

Whereas gross comparative studies provide a broad-brush description of classroom instruction in East Asian countries, a substantial body of literature presents a more contextualized and nuanced picture. The most prominent and well-documented research has historically taken place in Japan (Fukuzawa & LeTendre 2001; Lewis, 1995; Rohlen, 1983; Rohlen & LeTendre, 1996; Tobin et al. 2009). Refuting the monolithic image of child-centered or drill-oriented Japanese classrooms, field studies find that classroom structure and pedagogical approach change over the course of schooling as a result of the onset of high-stakes national tests (Baker & LeTendre, 2000). Scholars also explain that the educational system as a whole is built in accordance with Japanese assumptions about human maturation and that classroom organization at different levels of schooling reflects the developmental expectation for each stage (Rohlen & LeTendre, 1996). Studies further place the static portrait of Japanese classrooms into context. Scholars explain how school organizational structure interacts with or reinforces instruction in classrooms (LeTendre et al., 2001; Yang 1999) and how various nonacademic activities in schools function to balance out the intense, text-oriented classroom instruction (Fukuzawa, 1998).

Another line of scholarship has focused on the so-called East Asian classroom paradox, which refers to the seemingly incompatible phenomena of the high achievement of East Asian students and classroom features that are assumed to be less conducive to learning, such as teacher-centered instruction and rote learning (Leung, 2006; Mok, 2006; Watkins & Biggs, 1996; Zhou et al., 2012). Some suggest that the highly structured classroom activities and teachers with strong subject knowledge compensate for practices that are deleterious to learning and eventually lead to higher student achievement (An et al., 2004; Ma, 1999; Wang & Paine, 2003). Other studies report that carefully designed official curricula and textbooks and teacher collaboration in planning the lessons further assist in producing well-structured lessons (Fan & Kaeley, 2000; Lewis et al., 2006). Researchers of mathematics education and learning theory discovered a common instructional pattern in East Asian classrooms that involves different levels and types of variations and explains how students acquire understanding in teacher-dominated classrooms. They argue that the use of this instructional strategy enables students to discern critical dimensions of the learning topic (Fan et al., 2004; Fang & Gopinathan, 2009).

Ethnographic studies take an even deeper look into East Asian classrooms. Scholars point out that classroom practices hold different values and meanings within each social, historical, and cultural setting and have different educational implications. For example, in East Asia, memorization is viewed as complementing deeper understanding, as opposed to the general Western notion that memorization constitutes rote learning (Kember, 1996; Marton et al., 1996). Moreover, studies report that what translates as a content-oriented and teacher-dominated lesson from a Western perspective is perceived as a student-centered one from an East Asian perspective, in that the teacher develops a framed experience in consideration of how his or her students think and learn (Clarke & Xu, 2008; Mok, 2006). Other studies find that teachers may appear authoritarian or student centered depending on the setting and that such contingent shifts are natural in the Confucian cultural context. For example, Chinese teachers tend to keep absolute control over the situation in the classroom but interact with students on more friendly terms outside the classroom (Ho, 2001; Paine, 1990).

These rich observational studies begin to break away from the conventional assumption of within-country homogeneity prevalent in gross national comparison studies by exploring variability within countries. Joseph Tobin and his colleagues studies of preschools in China, Japan, and the United States have examined variation within cultures by showing videos of instructional practices in preschools to individuals within and outside of the preschool community and collecting the varied responses to the video cues (Tobin et al., 1989, 2009). Earlier Japanese studies also acknowledged the vastly different teaching approaches observed across grade levels (Rohlen & LeTendre, 1996) and school types (Kariya, 2011; Kariya & Rosenbaum, 1999). But little is known about whether classroom practices differ across classrooms in the same grade level and how the variation in classroom practices associates with varied student achievement.

WITHIN-COUNTRY RELATIONSHIP BETWEEN CLASSROOM PRACTICE AND ACHIEVEMENT

Finally, a number of studies have interrogated the relationship between classroom practices and student achievement using large-scale data (House & Telese, 2008; Paik, 2004; Schaub & Baker, 1991). Using a common cross-national quantitative analytic strategy that compares the effect of variables of interest on student achievement in different national settings, these studies have explored the association between learning environment and student achievement within East Asian countries. Studies report that instructional strategies that use real-world examples and independent learning activities are positively related to mathematics achievement in Japan and Taiwan (House, 2002, 2009). By contrast, a study on Japan finds a negative association between student individual or group work and mathematics achievement and a positive association between teachers time spent on whole-class lecturing and mathematics achievement (Schaub & Baker, 1991). Other studies report no statistically significant association between classroom instructional organization and student achievement (Mohammadpour, 2013; Paik, 2004). The mixed results may be partly attributable to the different data and items involved in the analyses. Setting aside this matter, because most of the studies have failed to control for student, classroom, or school background variables, it is likely that the estimated associations between classroom practice and achievement are spurious. Moreover, most studies used ordinary least squares regression analyses when multilevel modeling would have been appropriate given the nested structure of the data. Failing to recognize the hierarchical structure may have resulted in misestimating the standard errors of the instructional practice coefficients (Raudenbush & Bryk, 2002).

RECOGNIZING HETEROGENEITY AND INCREASING DIVERSITY WITHIN EAST ASIAN COUNTRIES

The literature discussed thus far has rarely taken into consideration the variation in classroom practices that may exist across classrooms within the East Asian countries. The assumption of within-country homogeneity has been more readily made about East Asian countries because of their relatively small and/or less diverse populations and their long-maintained standardized and tightly controlled educational systems (Park, 2013; Schmidt et al., 2001). But as continually argued by the research community, doing cross-national comparison using national averages of educational characteristics (e.g., national averages of teacher-centered instruction) without considering the variability within nations may lead to misleading educational inferences (Baker, 1997; Ercikan et al., 2015; Torney-Purta & Amadeo, 2013).

This study extends the previous literature on East Asian classrooms by considering the heterogeneity of schooling within East Asian countries. I focus on four aspects of classroom instructional practice that are commonly referred to in depictions of East Asian classrooms: complex instruction, procedural instruction, teacher-centered instruction, and student-centered instruction. Complex instruction involves mathematical tasks that require students complex thinking and reasoning, and procedural instruction involves tasks that require memorization or routine procedures. Teacher-centered instruction and student-centered instruction refer to the degree of teacher control or participation of students in classroom activities. Specific operationalization of the four instructional practices is described in the methods section that follows. I examine the classroom practices of East Asian countries with the following research questions:

1. To what extent do classroom instructional practices in East Asian countries differ from those in other countries?

A.

Are classroom instructional practices in East Asian countries more procedural and teacher centered and less complex and student centered compared with other countries?

B.

Are classroom instructional practices more homogeneous among classrooms within East Asian countries compared with other countries?

2. How are classroom instructional practices associated with student achievement within East Asian countries, controlling for student, classroom, and school background variables?

ANALYTIC STRATEGY

To address the research questions, I use the TIMSS 2007 data to define the four key aspects of instructional organization, the prevalent use of complex and procedural tasks, and student- and teacher-centeredness. First, I examine how the country means and within-country variation of the four aspects of classroom instructional organization in the five East Asian countries compare with those in the other 45 nations in the sample.

Second, I narrow my focus to two of the East Asian countries that have vastly different school structures, Japan and Singapore, to examine the within-country association of classroom practices and student achievement. These two countries were chosen because they stand at the extreme ends of the spectrum in their institutional arrangements, Japan having an integrated school system and Singapore having a differentiated system. In Japan, all students are exposed to a common curriculum, and there is no official policy on tracking or ability grouping in the elementary and lower secondary schools (Mullis et al., 2008). A majority of students are assigned to neighborhood junior high schools unless they choose to attend private schools (Fujita, 2010). In Singapore, on the contrary, students are placed on different tracks based on the exam that takes place at the end of elementary school and exposed to different curricula that match their ability. Through the high-stakes examination, students are sorted into different schools, creating a hierarchy of schools according to students scores, which then come to reflect the schools prestige and selectivity (Ng, 2007; Tan & Dimmock, 2014).¹ I pay attention to whether and how classroom instructional organization maintains effects on student achievement after taking into account characteristics of students, classrooms, and schools, and how the patterns differ in these two countries, with their vastly different systems of differentiation.

METHODS

DATA

The data set used is the Trends in International Mathematics and Science Studies 2007 (TIMSS 2007), conducted by the International Association for the Evaluation of Educational Achievement (IEA). TIMSS provides mathematics and science assessment scores of nationally representative fourth- and eighth-grade students and a wealth of information on schools, curricula, teachers, and students in over 50 countries. In each country, a two-stage stratified clustered sampling design is conducted, with schools sampled in the first stage and intact classrooms sampled in the second stage. Typically, countries sampled 150 schools and one or two intact classrooms. TIMSS 2007 eighth-grade mathematics data contain information on about 240,000 students and about 12,000 teachers from 50 countries and seven benchmark participants.

To analyze country average instructional practices and within-country variation of instructional practices, I used survey items that asked about the frequency with which teachers used different instructional approaches in their classrooms. The operational sample included 8,864 teachers in 50 countries who completed the teaching items. In estimating the relationship between instructional practice and student achievement, it was essential to link teachers to the class or group of students they taught. Although most classes in Japan and Singapore were exposed to one mathematics teacher, some classes were taught by different teachers. In cases where classrooms were administrative units and not necessarily learning units, I treated the multiple teachers within classrooms as separate units. In estimating the effect of instructional practice on student achievement in Japan and Singapore, I used an operational sample of 140 teachers and 2,831 students in Japan and 284 teachers and 2,950 students in Singapore. All statistics reported use the appropriate student, teacher, and school sampling weights, as recommended in the TIMSS 2007 User Guide (Foy & Olson, 2009); methods for sampling weights are detailed in the following section.

MEASURES

Appendix A provides detailed descriptions of all variables used in this study. Descriptive statistics of the variables are provided in Table 1.

Table 1. Descriptive Statistics

Key Explanatory Variables: Instructional Practices

I used prior research on classroom instruction to sort the teacher survey items into categories of instruction. Complex instruction denotes classroom practices that emphasize real-world problem solving, elicit student reflection, and involve working with open-ended problems with no obvious solution (also called higher order instruction, authentic pedagogy, or conceptual instruction) (e.g., Carpenter et al., 1989; Desimone et al., 2005; Newmann et al., 1996; Raudenbush et al., 1993). In mathematics, it is believed that complex instruction allows students to develop a deeper understanding of mathematical concepts and logic. A greater emphasis on complex instruction is advocated in the United States and is a core component of current state and national education policies (Common Core State Standards Initiative, 2011; National Council of Teachers of Mathematics, 2000). To determine teachers emphasis on complex instruction, I used four items that asked how often a teacher asks students to explain the reasoning behind an idea, relate the learning topic to their daily lives, work on problems with no immediately obvious method of solution, and decide on their own procedures for solving complex problems. Response options ranged from never (1) to always (4). An average of the four items formed the complex instruction variable. Cronbachs alpha is 0.66.

Procedural instruction refers to classroom practices that emphasize computational skills and involve memorizing facts and procedures (also called basic skills instruction or traditional instruction) (e.g., Hamilton and Martinez 2007). Too much reliance on procedural instructional practices is generally thought to hinder students from developing a deeper understanding of the mathematical concepts behind the procedures (Baroody & Benson, 2001; National Council of Teachers of Mathematics, 2000; Romberg, 1988). I used three items that asked how often teachers ask students to practice adding, subtracting, multiplying, and dividing without using a calculator; memorize formulas and procedures; and apply facts, concepts, and procedures to solve routine problems, which I categorize as procedural instruction. An average of the three items is used as the procedural instruction variable. Cronbachs alpha is 0.52 for the procedural instruction scale.

To determine teachers emphasis on teacher- and student-centered instruction, I used another set of teacher survey questions that ask teachers to write in the percentage of time they spend on a list of activities, which should add up to 100%. Two of the seven items were used to derive the teacher-centered instruction variable: (a) lecture-style presentation by teachers and (b) reteaching and clarification of content/procedures. The sum of the percentages of time divided by 10 was used as the teacher-centered instruction variable. For the student-centered instruction variable, the percentage of time spent on student independent practice divided by 10 was used.

Dependent Variable: Student Mathematics Achievement

For the student achievement measure, I used the student mathematics test scores, which consist of five plausible values. TIMSS conceptualizes student mathematics achievement as a latent construct measured with uncertainty and provides five separate imputed scores. Any analysis that involves student achievement should be replicated five times, and the results should be aggregated to obtain the final estimate as well as the imputation error (Foy & Olson, 2009). Thus, I use all five plausible values for the analysis in this article.

Classroom and School Characteristics

To take into account possible nonrandom assignment of students to classrooms that emphasize different types of instructional practices, I included school and classroom characteristics that may be simultaneously correlated with classroom instruction and class average achievement. Classroom aggregates of students socioeconomic background variables, number of books at home, and parent education capture the effect of class socioeconomic status on average class achievement over and above the effect of individual socioeconomic status on student achievement within a classroom. I also included items from the teachers survey that asked whether different academic abilities of students and uninterested students limited how they taught their class.² These variables reflect teachers perceptions of class characteristics that may have affected their choice of instructional strategies. At the school level, I included school size, a dummy variable that indicates whether 50% or more of students are from affluent families, a dummy variable for high availability of mathematics resource, and a dummy variable for high school climate. I also included school characteristics variables that were used for stratified sampling of schools in each country. Japan had school-type information (public schools in large cities, public schools in small cities, and private schools), and Singapore had school performance rankings.

Control Variables

At the individual student level, I include five control variables: number of books in students homes, parent education, students sex, whether parents are foreign born, and frequency of speaking the test language at home. The first two variables are proxies for the students socioeconomic background. Because class mean socioeconomic status variables are included in the model, these variables indicate the effects of individual socioeconomic status after controlling for the class mean, that is, effects of socioeconomic status within the classrooms. I included student sex because sex differences in math achievement are traditionally prevalent in many countries. At the teacher level, teachers years of teaching experience and educational degree were included as proxies for teachers content knowledge and pedagogical content knowledge. Teachers sex is also controlled.

ESTIMATION APPROACH

The analysis is divided into two parts. First, to address the first set of research questions, I compare the prevalence and within-country variation of instructional practices in East Asian countries with the other 45 countries. Specifically, I examine whether classrooms in East Asian countries have more procedural and teacher-centered instruction and less complex and student-centered instruction, and whether there is less within-country variation in classroom instructional practices in East Asian countries. Prevalence of an instructional practice is measured by the country-level mean, and within-country variation is measured by the within-country coefficient of variation (standard deviation divided by mean, multiplied by 100). The measures are derived by using the mathematics teacher weight variable (MATWGT). Using MATWGT allows for describing the instructional practices faced by nationally representative samples of students (Foy & Olson, 2009). Note that the sampling design of TIMSS, where the second-stage sampling units are classrooms and not teachers, places a constraint on the interpretation of teacher data analysis. Inferences should be made at the student level and not the teacher level.

Second, I focus on two East Asian countries, Japan and Singapore, to examine the effects of instructional practices on student achievement. To take into account the data structure (students nested within classes nested within schools), I employ multilevel models to gauge the effects of instructional practice on student achievement. Singapore has more than one classroom sampled within each sampled school, which makes it possible to specify three-level models, whereas Japan sampled one classroom per school, which allows for two-level models. A three-level model for Singapore is specified as follows:

TIMSS uses a two-stage stratified clustered sampling design, which requires an appropriate use of sampling weights when running multilevel models. The multilevel models are estimated with HLM 6.0 software using weights that are manually calculated for each level (Rabe-Hesketh & Skrondal, 2006). For three-level models, in the case of Singapore, the total student weight is decomposed into the inverse of the probability of selection for a school, for a class within a school given that the associated school was selected, and for a student given that the students school and class were selected. For two-level models, in the case of Japan, class weight is used for the second level and student weight is used for the first level. The HLM software allows for considering all five plausible values of mathematics achievement for the analyses (Rutkowski et al., 2010).  

FINDINGS

CROSS-NATIONAL COMPARISON OF THE PREVALENCE AND WITHIN-COUNTRY VARIATION OF CLASSROOM INSTRUCTIONAL PRACTICES IN EAST ASIAN COUNTRIES

The first set of research questions centers on the issue of whether the average levels and within-country variations of classroom instructional practices in East Asian countries differ from those in other countries. Is classroom instruction in East Asian countries more procedural and teacher-centered compared with that in other countries? Are classrooms in East Asian countries more homogeneous within each country in terms of instructional practices compared with those in other countries? Figure 1 addresses these questions by plotting the country average and coefficient of variation of instructional practices in the 50 countries (see Appendix B for the values of each country).

Figure 1. The relationship between country average and coefficient of variation of instructional practices

With respect to complex instruction, four of the East Asian countries (the exception being Korea) showed the lowest scores among the 50 countries. Less distinctively, East Asian countries scored lower in student-centered instruction. This result is consistent with the traditional notion that East Asian classrooms are less student centered and less complex. But more importantly, classrooms within East Asian countries vary considerably, as indicated by the large coefficients of variation in complex, procedural, and student-centered instruction, countering the general assumption of homogeneity within East Asian countries. On the other hand, classrooms in East Asian countries did consistently employ teacher-centered instruction. Four of the East Asian countries (the exception being Singapore) not only ranked highest on the teacher-centeredness scale but also had small within-country variations compared with other countries in the sample.

 

WITHIN-COUNTRY RELATIONSHIP OF CLASSROOM INSTRUCTION AND STUDENT ACHIEVEMENT

Before moving on to the multivariate models for Japan and Singapore, we examine the correlation of the key variables in each country (Tables 2 and 3). In Japan, class average math scores tend to be higher in classrooms with more complex instruction (r = 0.17, p < 0.05), and classrooms with a higher proportion of parents with college degrees are more likely to be taught with complex instruction (r = 0.23; p < 0.01). Public schools in large cities have more student-centered classrooms, whereas public schools in small cities tend to be lower in student-centeredness. Private schools tend to enroll students from higher socioeconomic backgrounds and have better school climates compared with public schools.

Table 2. Correlations of Variables, for Japanese Classes (n=140)

^a Average of the five class math scores based on the five plausible values was used for simplicity.

*p < 0.05. **p < 0.01. ***p < 0.001

Table 3. Correlation of Variable, for Singaporean Classes (n=284)

^a Average of the five class math scores based on the five plausible values was used for simplicity.

*p < 0.05. **p < 0.01. ***p < 0.001

In Singapore, class average math scores are significantly correlated with complex, student-centered, and teacher-centered instruction (r = 0.23, r = 0.21, and r = 0.14, respectively). Classrooms with higher socioeconomic background students are more likely to have complex and student-centered instruction. Schools of higher performance ranking tend to have more complex (r = 0.13; p < 0.05) and teacher-centered instruction (r = 0.21; p < 0.001). Higher ranking schools also tend to enroll students from higher socioeconomic backgrounds.

Now we turn to findings of the multivariate models pertaining to the second research question (Tables 4 and 5). Instructional practice, classroom characteristics, and school characteristics are sequentially added to the models to examine how classroom and school characteristics confound the relationship between instructional practice and achievement. Table 4 shows that in Japan, classrooms with more complex instruction tend to have higher class average mathematics scores, after controlling for student and teacher characteristics (Model 1). But the effect of complex instruction is no longer significant when class average student background variables and classroom and school characteristics are introduced to the model (Models 2 and 3). As shown in the correlational analysis (Table 2), this may be due to the fact that classrooms with a greater proportion of highly educated parents are more likely to be exposed to complex instruction. Other classroom instructional practices are not significantly associated with student achievement. The final models accounted for about 77% of achievement variance at the classroom level and 8% of variance at the student level.

Table 4. Relationship between instructional practices and student achievement in Japan

           

Table 5. Relationship between instructional practices and student achievement in Singapore

^a Significant control variables are foreign born parent, language at home, and male teacher.

p < 0.1. ^*p < 0.05. ^**p < 0.01. ^***p < 0.001.

As shown in Table 5, in Singapore, the effects of complex, student-centered, and teacher-centered instruction on student achievement in models with student and teacher background controls are significantly positive (Models 1, 7, and 10). When the socioeconomic makeup of classrooms is included in the models, the effects of complex and teacher-centered instruction are attenuated (Models 2 and 11). As illustrated in the correlational analysis (see Table 3), this may be because classrooms with higher average student socioeconomic background have more complex and teacher-centered instruction. The full models show that although the effects of complex instruction and student-centered instruction decrease slightly, they remain significant after including all school characteristics variables (Models 3 and 9). The effect of teacher-centered instruction loses its significance when school rank is added into the model. The decrease in size and significance of the classroom instructional practice effects signals that higher-ranking schools may be employing a higher degree of complex instruction as well as student- and teacher-centered approaches in classrooms. The final models explain over 55% of achievement variance at the classroom level and over 90% of variance at the school level. However, less than 1% of the student-level variance in achievement was accounted for by the final models. This is not surprising given the tiered school system in Singapore, where there is very little achievement variation within the classrooms to begin with.

In both Japan and Singapore, the results demonstrate that the effect of classroom instruction on student achievement is confounded by class average student background. In other words, wealthier students may be selecting into classrooms that adopt certain types of instructional approaches, or teachers may be differentiating their instruction based on the average student class background. Students differential exposure to different types of classroom practice is further assisted by educational institutions in Singapore, where student sorting is arranged through the school ranking system.

Private schools in Japan and higher ranking schools in Singapore tend to score higher in mathematics on average. As is common in most countries, students family socioeconomic background is highly associated with mathematics achievement in both countries (coefficients for all other control variables are available from the author).

DISCUSSION

To gain a deeper holistic understanding of classroom life and instructional practices in East Asian countries, this article examined both the prevalence and distribution of complex, procedural, student-centered, and teacher-centered instruction, along with the estimated effects of such practices on student learning within nations. The analytic strategy focused on variation in classroom practices within East Asian countries, as opposed to past approaches that stress between-country variability in mean levels of certain classroom practices or how teachers organize instruction. By taking a fresh analytic approach, I have reassessed long-held stereotypes about classroom practices in East Asian countries, offering more complete evidence on the efficacy of complex or more regimented instruction in cultural context.

Three major findings emerged from this analysis. First, although classrooms in East Asian countries are indeed among the most teacher centered and least complex on average, the extent to which classrooms within these countries employ certain practices varies considerably. The levels of within-country variation match or exceed the heterogeneity of classroom practice observed in the other 45 participating countries. Again, this suggests the need for caution before generalizing about East Asian classrooms based on national averages, or assuming causal links between average features of East Asian classrooms and high student performance.

This finding also contrasts with prior studies reporting that countries with central government control over curricula, including most East Asian countries, have less variation across classrooms in terms of instructional content (Schmidt & Prawat, 2006; Stevenson & Baker, 1991). The central control of curriculum and the consistency of instructional topics across classrooms, stemming from central ministry regulation, have been identified as drivers of high student performance in East Asia. But my findings show that although teachers may show similarity in what they teach, how they teach various topics differs greatly among classrooms within nations. Literature on Japanese elementary schools reports that teachers are trained to design classroom activities that allow students to inventively discover the contents listed in the national curriculum (Lewis & Tsuchida, 1997). Whether the same explanation applies in secondary schools should be explored in future work. Moreover, the mechanisms by which variability in classroom organization in East Asian countries explains high achievement could be further unpacked in future studies.

Learner-centered pedagogy may be becoming a global norm in terms of governments official proclamations (Bromley et al., 2011), including East Asian education ministries that promote student-centered and active learning approaches in classrooms. But, as my findings show, East Asian classrooms continue to employ more teacher-centered practices and fewer student-centered practices. And even under greater centralization, we observe wide heterogeneity in classroom practices within country. How teachers buffer compliance expected by government or tailor their own pedagogical approaches deserves more research in East Asia. This within-country variability may reflect the multiple models of good teaching present in the system, not unlike debates in the West. Some schools may be faster at incorporating new models, while others adhere to more didactic conceptions of teaching. How teachers adopt, negotiate, or resist novel models of instructional practice becomes a pivotal question once we appreciate the heterogeneity in how they organize instruction.

Second, by examining the association between classroom instructional organization and student achievement in two East Asian countries, Japan and Singapore, we see how complex or student-centered instruction is positively associated with pupil performance in both nations. The direction of these associations within the countries contrasts with that at the between-country level, where classrooms in East Asian countries remain high in teacher-centeredness and low in complexity, while ranking high on pupil assessments. The positive association between complex or student-centered instruction and classroom achievement within Japan and Singapore is consistent with learning theories positing that complex or student-centered instruction is conducive to student learning. Thus, it turns out that the East Asian classroom paradox does not apply when we look within a country.

These seemingly contradictory patterns at the between- versus within-country levels may indicate that differing causal mechanisms are in play at each level. To investigate the relationship between instructional practice and achievement at the country level, country-level mechanisms, and not the learning theories that operate on the micro level, should be considered (Robinson, 2009). For example, it may be that country-level attributes such as standardization of curriculum or test-based controls affect the country average instructional practice and achievement. This warrants further investigation.

To understand the classroom-level associations between instructional practice and achievement, we should attend to classroom-level mechanisms within each country. Taking this approach leads us to the third finding. In both Japan and Singapore, the association between classroom practices and student achievement decreased in size or lost statistical significance after adjusting for the social class status of pupils. This could result from sorting of diverse students into a stratified system of schools. And the correspondence between socioeconomic status of students and instructional practice is reinforced by institutional structures. In the United States, for example, different types of instructional practice are differentially provided through institutional arrangements, such as curriculum tracks or ability grouping, based on the social class composition of the school (Barr et al., 1983; Oakes, 1985; Raudenbush et al., 1993; Zohar et al., 2001).

Evidence of institutional stratification is especially pronounced in Singapore, given that classrooms in higher ranking schools tend to employ more complex tasks and both teacher-centered and student-centered approaches compared with those in lower-ranking schools. Institutional differentiation is less sharply defined in Japan, perhaps because of its egalitarian and comprehensive lower secondary system. But even when formally structured curricular differentiation is weaker, I found that Japanese students do sort into differing types of schools and geographical locations that covary with social class origins. As shown in the correlational analyses, private schools tend to enroll students from wealthier families, then expose them to more teacher-centered and less student-centered instruction compared with peers selecting into public schools. Teachers in public schools, especially those situated in urban areas, tend to employ more student-centered instruction. This correspondence pattern departs from findings of U.S.-based studies, where pupils from poorer families experience more procedural and didactic instruction, compared with students from more affluent families (Bodovski & Farkas, 2007; Desimone & Long, 2010; Lucas & Gamoran, 2002; Maaz et al., 2008; Oakes, 1985). My findings are consistent with previous Japanese studies finding that wealthier parentsreportedly dissatisfied with the governments Western-style curriculum reform to pull back on regimented didacticsopt for private schools that employ faster-paced, lecture-oriented instruction, often believed to effectively lift students performance in high-stakes exams (Fujita, 2010; Kariya & Rosenbaum, 1999).

These findings accent the importance of digging deeper into East Asian classrooms, drawing on nuanced, culturally situated constructs and mechanisms, both quantitative and qualitative in nature, to lay a foundation for future research. We must learn more about how students or parents in East Asia may intentionally select into differing schools or classrooms based on their educational philosophies or pedagogical practices. In addition, we know little about how teachers may adjust their instructional practices based on the capacities or family backgrounds of their students.

Overall, we should pursue a variety of fresh questions within the cultural context of East Asian societies: To what extent does classroom organization correspond with the social class attributes of students in countries like Japan, where the system is comprehensive and no institutional sorting is in place? Do the perceptions of students prior learning or capacities, held by teachers, affect classroom practices? How do teachers differentiate their teaching by student ability in countries like Singapore, where schools are ranked quite publicly by achievement levels? How do culturally specific beliefs about pedagogy and learning affect how teachers organize classroom instruction? These research studies can also promote greater dialogue and integration between U.S.-based learning theoriesfor example, adaptive teaching (Corno, 2008; Corno & Snow, 1986), culturally relevant pedagogy (Ladson-Billings, 1995; Ladson-Billings, 2014)and streams of research on East Asian education.

National quantitative assessments with stronger longitudinal design are particularly suitable in testing causal relationships between classroom practices and student achievement and possible sorting via parental choice or teachers selective adjustment (Porter & Gamoran, 2002). And we sorely need stronger measures of how teachers understand and enact their classroom practices. Educational surveys, for example, could ask teachers to compare normative pedagogical practices with their own educational beliefs and how they organize their own classrooms. Qualitative studies might uncover what instructional approaches are sought by parents in East Asia and whether certain practices are deemed desirable for particular students, perhaps varying along lines of social class (LeTendre, 2002). This would inform what is driving variability in classroom practices, as well as reveal how teachers are affected by cultural and institutional context of their society (e.g., norms about teacher authority and student participation, values attached to student performance in high-stakes exams and their implication for classroom practice).

Often, classroom practices prevalent in East Asian countries are celebrated as best practices to emulate, criticized for their rigidity, or simply regarded as not importable for the West because of the peculiarity of East Asian culture. But as this study shows, there is more variability than previously expected, and student achievement within the countries is a product of instructional practices interacting with social class backgrounds of students similar to patterns observed in the United States and other developed countries. Once we acknowledge the commonality as well as the differences, educational research on East Asian countries might not only allow us to better understand perennial educational problems and the assumptions we hold about classroom practices but also inform valid implications for policy and practice.

Notes

1. Based on a preliminary analysis that decomposed the variance in student math scores into student, classroom, and school-level variances in the five East Asian countries, the total variation in math scores that stem from between classrooms or school levels was smallest in Japan (19%) and largest in Singapore (77%) (Appendix C). That is, 81% of the variance in math scores is attributable to between-student sources within classrooms in Japan, and 23% in Singapore. Different patterns of apportioning reflect the different educational systems of the two countries, namely, the comprehensive system in Japan and the stratified system in Singapore.

2. The terms different academic abilities and uninterested students are direct quotes from the TIMSS teacher survey and not the authors.

References

An, S., Kulm, G., & Wu, Z. (2004). The Pedagogical content knowledge of middle school, mathematics teachers in China and the U.S. Journal of Mathematics Teacher Education, 7, 14572.

Baker, D. P. (1997). Surviving TIMSS: Or, everything you blissfully forgot about international comparisons. Phi Delta Kappan, 78(4), 295300.

Baker, D. P., & LeTendre, G.K. (2000). Comparative sociology of classroom processes, school organization, and achievement. In M.T. Hallinan (Ed.), Handbook of the Sociology of Education (pp. 34564). New York, NY: Kluwer Academic/Plenum Publishers.

Baroody, A. J., & Benson, A. (2001). Early number instruction. Teaching Children Mathematics, 8(3), 15458.

Barr, R., Dreeben, R., & Nonglak, W. (1983). How schools work. Chicago, IL: University of Chicago Press.

Bodovski, K., & Farkas, G. (2007). Do instructional practices contribute to inequality in achievement? The case of mathematics instruction in kindergarten. Journal of Early Childhood Research, 5(3), 30122.

Bromley, P., Meyer, J.W., & Ramirez, F.O. (2011). Student-centeredness in social science textbooks, 1970-2008: A cross-national study. Social Forces, 90(2), 54770.

Carpenter, T., Fennema, E., Peterson, P., Chiang, C., & Loef, M. (1989). Using knowledge of childrens mathematics thinking in classroom teaching: An experimental study. American Educational

Research Journal, 26(4), 499531.

Clarke, D., & Xu, L.H. (2008). Distinguishing between mathematics classrooms in Australia, China, Japan, Korea and the USA through the lens of the distribution of responsibility for knowledge generation: Public oral interactivity and mathematical orality. ZDM Mathematics Education, 40(6), 96372.

Common Core State Standards Initiative. (2011). Common Core State Standards for Mathematics. Washington, DC: National Governors Association Center for Best Practices and the Council of Chief State

School Officers.

Corno, L., & Snow, R.E. (1986). Adapting teaching to individual differences among learners. In M.C. Wittrock, (Ed.), Handbook of Research on Teaching, (3rd ed.). New York, NY: Macmillan.

Corno, L. (2008). On teaching adaptively. Educational Psychologist, 43(3), 16173.

Dahlin, B., & Watkins, D.A. (2000). The role of repetition in the process of memorising and understanding: A comparison of the views of German and Chinese secondary school students in Hong Kong. British Journal of Educational Psychology, 70(1), 6584.

Desimone, L. M., Smith, T., Baker, D., & Ueno, K. (2005). Assessing Barriers to the Reform of U.S. Mathematics Instruction from an International Perspective. American Educational Research Journal, 42(3), 51035.

Desimone, L. M., & Long, D. (2010). Teacher effects and the achievement gap: Do teacher and teaching quality influence the achievement gap between Black and white and high- and low-SES students in the early grades? Teachers College Record, 112(12), 302473.

Ercikan, K., Wolff, M. R., & Asil, M. (2015). Cautions about inferences from international assessments: The case of PISA 2009. Teachers College Record, 117(1), 128.

Fan, L., & Kaeley, G.S. (2000). The influence of textbooks on teaching strategies: An empirical study. Mid-Western Educational Researcher, 13(4), 29.

Fan, L., Wong, N., Cai, J., & Li, S. (2004). How Chinese learn mathematics: Perspectives from the insiders, series on mathematics education. Singapore: World Scientific.

Fang, Y., & Gopinathan, S. (2009). Teachers and teaching in Eastern and Western schools: A critical review of cross-cultural comparative studies. In L.J. Saha & A.G. Dworkin (Eds.), International Handbook of Research on Teachers and Teaching (pp. 55772). New York, NY: Springer.

Ference, M., DallAlba, G., & Kun Tse, L. (1996). Memorizing and understanding: The keys to the paradox? In D.A. Watkins & J.B. Biggs (Eds.), The Chinese learner: Cultural, psychological and contextual influences (pp. 6983). Hong Kong: Comparative Education Research Centre, University of Hong Kong.

Foy, P., and J. F. Olson. (2009). TIMSS 2007 User Guide for the International Database. Chestnut Hill, MA: TIMSS & PIRLS International Study Center, Boston College.

Fujita, H. (2010). Whither Japanese schooling? Educational reforms and their impact on ability formation and educational opportunity. In J.A. Gordon, H. Fujita, T. Kariya, & LeTendre, G.K. LeTendre (Eds.), Challenges to Japanese education: Economics, reform, and human rights (pp. 1753). New York, NY: Teachers College Press.

Fukuzawa, R. I. (1998). The Path to Adulthood According to Japanese Middle Schools. In Rohlen, T.P. & LeTendre, G.K. (Eds.), Teaching and learning in Japan (pp. 295-321). Cambridge, UK: Cambridge University Press.

Fukuzawa, R. E., & LeTendre, G.K. (2001). Intense years: How Japanese adolescents balance school, family, and friends. New York, NY: Routledge Falmer.

Hamilton, L. S., & Martinez, J. F. (2007). What can TIMSS surveys tell us about mathematics reforms in the United States during the 1990s? In T. Loveless (Ed.), Lessons learned: What international assessments Tell Us About Mathematics Achievement (pp. 12774). Washington, DC: Brookings Institutions.

Hiebert, J., Gallimore, R., Garnier, H., Bogard Givvin, K., Hollingsworth, H., Jacobs, J. K., Miu-Ying Chui, A., et al. (2003). Teaching mathematics in seven countries: Results from the TIMSS 1999 Video Study. Washington, DC: National Center for Education Statistics.

Ho, I. T. (2001). Are Chinese Teachers authoritarian? In D. A. Watkins & J. B. Biggs (Eds.), Teaching the Chinese learner: Psychological and pedagogical perspectives (pp. 99114). Hong Kong: Comparative Education Research Centre.

House, J. D. (2002). Instructional practices and mathematics achievement of adolescent students in Chinese Taipei: Results from the TIMSS 1999 assessment. Child Study Journal, 32(3), 15778.

House, J. D. (2009). Elementary-school mathematics instruction and achievement of fourth-grade students in Japan: Findings from the TIMSS 2007 assessment. Education, 130(2), 3017.

House, J. D., & Telese, J.A. (2008). Relationships between student and instructional factors and algebra achievement of students in the United States and Japan: An analysis of TIMSS 2003 Data. Educational Research and Evaluation: An International Journal on Theory and Practice, 14(1), 10112.

Huang, R., & Leung, F.K.S. (2004). Cracking the paradox of Chinese learners: Looking into the mathematics classrooms in Hong Kong and Shanghai. In L. Fan, N.Y. Wong, J. Cai, & S. Li (Eds.), How Chinese learn mathematics: Perspectives from insiders (pp. 348–81). Singapore: World Scientific Publishing.

Kai, M., Trautwein, U., Lüdtke, O. & Baumert, J. (2008). Educational transitions and differential learning environments: How explicit between‐school tracking contributes to social inequality in educational outcomes. Child Development Perspectives, 2(2), 99106.

Kariya, T, & Rosenbaum, J.E. (1999). Bright flight: Unintended consequences of detracking policy in Japan. American Journal of Education, 107(3), 21030.

Kariya, T. (2011). Japanese solutions to the equity and efficiency dilemma? Secondary schools, inequity and the arrival of universal higher education. Oxford Review of Education, 37(2), 24166.

Kember, D. (1996). The Intention to both memorise and understand: Another approach to learning? Higher Education, 31(3), 34154.

Ladson-Billings, G. (1995). Toward a theory of culturally relevant pedagogy. American Educational Research Journal, 32(3), 46591.

Ladson-Billings, G. (2014). Culturally relevant pedagogy 2.0: a.k.a. the remix. Harvard Educational Review, 84(1), 7484.

LeTendre, G. K. (2002). Advances in conceptualizing and analyzing cultural effects in cross-national studies of educational achievement. In A.C. Porter & A. Gamoran (Eds.), Methodological advances in cross-national surveys of educational achievement. Washington, DC: National Academy Press.

LeTendre, G. K., Baker, D. P., Akiba, M., Goesling, B., & Wiseman, A. (2001). Teachers Work: Institutional isomorphism and cultural variation in the U.S., Germany, and Japan. Educational Researcher, 30(6), 315.

Leung, F. K. S. (1995). The mathematics classroom in Beijing, Hong Kong and London. Educational Studies in Mathematics, 29(4), 297325.

Leung, F. K. S. (2001). In search of an East Asian identity in mathematics education. Educational Studies in Mathematics, 47, 3551.

Leung, F. K. S. (2002). Behind the high achievement of East Asian students. Educational Research and Evaluation, 8(1), 87108.

Leung, F. K. S. (2006). Mathematics education in East Asia and the West: Does culture matter? In F.K.S., Leung, K.D. Graf, & F.J. Lopez-Real (Eds.), Mathematics education in different cultural traditions Comparative study of East Asia and the West. The 13th ICMI study (pp. 127). New York, NY: Springer.

Lewis, C. (1995). Educating hearts and minds: Reflections on Japanese preschool and elementary education. Cambridge, UK: Cambridge University Press.

Lewis, C., & Tsuchida, I. (1997). Planned educational change in Japan: The case of elementary science instruction. Journal of Education Policy, 12(5), 31331.

Lewis, C., Perry, R., & Murata, A. (2006). How should research contribute to instructional improvement? The case of lesson study. Educational Researcher, 35(3), 314.

Li, J. (2002). A cultural model of learning: Chinese heart and mind for wanting to learn. Journal of Cross-Cultural Psychology, 33(3), 24869.

Li, J. (2003). The core of Confucian learning. American Psychologist, 58(2), 146147.

Lucas, S. R., & Gamoran, A. (2002). Tracking and the achievement gap. In J. E. Chubb & T. Loveless (Eds.), Bridging the achievement gap (pp. 171-198). Washington, DC: Brookings Institution Press.

Ma, L. (1999). Knowing and teaching elementary mathematics: Teachers understanding of  fundamental mathematics in China and the United States. Mahwah, NJ: Lawrence Erlbaum Associates.

Mohammadpour, E. (2013). A three-level multilevel analysis of Singaporean eighth-graders science achievement. Learning and individual differences, 26, 21220.

Mok, I.A.C. (2006). Shedding light on the East Asian learner paradox: Reconstructing student  centredness in a Shanghai classroom. Asia Pacific Journal of Education, 26(2), 131142.

Mullis, I.V.S., Martin, M.O., Olson, J.F., Berger, D., Milne, D., & Stanco, G.M. (2008). TIMSS 2007 Encyclopedia: A guide to mathematics and science education around the world. Chestnut Hill, MA: TIMSS & PIRLS International Study Center, Boston College.

National Council of Teachers of Mathematics. (2000). Principles and standards for school mathematics, (vol. 1). Reston, VA: National Council of Teachers of Mathematics.

Newmann, F. M., Marks, H.M., & Gamoran, A. (1996). Authentic pedagogy and student performance. American Journal of Education, 104(4), 280312.

Ng, P. T. (2007). Quality assurance in the Singapore education system in an era of diversity and innovation. Educational Research for Policy and Practice, 6(3), 23547.

Oakes, J. (1985). Keeping track: How schools structure inequality. New Haven, CT: Yale University Press.

Paik, S. J. (2004). Korean and U.S. families, schools, and learning. International Journal of Educational Research, 41, 7190.

Paine, L. W. (1990). The teacher as virtuoso: A Chinese model for teaching. Teachers College Record, 92(1), 4981.

Park, H. (2013). Re-evaluating education in Japan and Korea: Demystifying stereotypes. New York, NY: Routledge.

Porter, A.C., & Gamoran, A. (2002). Methodological advances in cross-national surveys of educational achievement. Washington, DC: National Academies Press.

Rabe-Hesketh, S., & Anders, S. (2006). Multilevel modeling of complex survey data. Journal of the Royal Statistical Society, 169(4), 80527.

Raudenbush, S. W., & Bryk, A.S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage Publications.

Raudenbush, S. W., Rowan, B., & Yuk, F. C. (1993). Higher order instructional goals in secondary schools: Class, teachers, and school influences. American Educational Research Journal, 30(3): 523553.

Robinson, W. S. (2009). Ecological correlations and the behavior of individuals. International Journal of Epidemiology, 38(2), 33741.

Rohlen, T. P. (1983). Japans high schools. Berkeley, CA: University of California Press.

Rohlen, T. P., & LeTendre, G.K. (1996). Teaching and Learning in Japan. Cambridge, UK: Cambridge University Press.

Romberg, T. A. (1988). One Point of View: NCTMs curriculum and evaluation standards: What they are and why they are needed. Arithmetic Teacher, 35(9), 23.

Rutkowski, Leslie, Gonzalez, E., Joncas, M. & von Davier, M. (2010). International large scale assessment data: Issues in secondary analysis and reporting. Educational Researcher, 39(2): 14251.

Schaub, M., & Baker, D.P. (1991). Solving the math problem: Exploring mathematics achievement in Japanese and American middle grades. American Journal of Education, 99(4), 62342.

Schmidt, W. H., Jorde, D., Cogan, L.S., Barrier, E., Gonzalo, I., Moser, U., Katsuhiko, S. et al. (1996). Characterizing pedagogical flow: An investigation of mathematics and science teaching in six countries. Dordrecht, Netherlands: Kluwer Academic.

Schmidt, W. H., McKnight, C. C., Cogan, L.S., Jakwerth, P.M., & Houang, R.T. (1999). Facing the consequences: Using TIMSS for a Closer Look at U.S. mathematics and science education. Dordrecht,

Netherlands: Kluwer Academic Publishers.

Schmidt, W. H., McKnight, C. C., Houang, R. T., Wang, H., Wiley, D., Cogan, L. S., & Wolfe, R. G. (2001).  Why schools matter: A cross-national comparison of curriculum and learning. San Francisco, CA: Jossey-Bass Education Series.

Schmidt, W. H., & Prawat, R.S. (2006). Curriculum coherence and national control of education: Issue or non-issue. Journal of Curriculum Studies, 38(6), 64158.

Shen, C. (2005). How American middle schools differ from schools of five Asian countries: Based on cross-national data from TIMSS 1999. Educational Research and Evaluation, 11(2), 17999.

Stevenson, D. L., & Baker, D. P. (1991). State control of the curriculum and classroom instruction. Sociology of Education, 64(1), 110.

Stevenson, H. W., & Stigler, J. W. (1992). The learning gap: Why our schools are failing and what we can learn from Japanese and Chinese education. New York, NY: Summit Books.

Stigler, J. W., & Baranes, R. (1988). Culture and mathematics learning. Review of Research in Education, 15, 253306.

Stigler, J. W., & Hiebert, J. (1999). The teaching gap: Best ideas from the worlds teachers for improving education in the classroom. New York, NY: The Free Press.

Tan, C. Y., & Dimmock, C. (2014). How a top-performing Asian school system formulates and implements policy: The case of Singapore. Educational Management Administration & Leadership, 42(5), 743763.

Tobin, J. J., Wu, D., & Davidson, D. H. (1989). Preschool in three cultures: Japan, China, and the United States. New Haven, CT: Yale University Press.

Tobin, J. J., Yeh, H., & Karasawa, M. (2009). Preschool in three cultures revisited: China, Japan and the United States. Chicago, IL: University of Chicago Press.

Torney-Purta, J., & Amadeo, J. (2013). International large-scale assessments: Challenges in reporting and potentials for secondary analysis. Research in Comparative and International Education, 8(3), 248258.

Tweed, R. G., & Lehman, D. R. (2002). Learning considered within a cultural context: Confucian and Socratic approaches. American Psychologist, 57(2), 8999.

Wang, J., & Paine, L.W. (2003). Learning to teach with mandated curriculum and public examination of teaching as contexts. Teaching and Teacher Education, 19, 7594.

Watkins, D. A., & Burville Biggs, J. (1996). The Chinese learner: Cultural, psychological and contextual influences. Hong Kong: Comparative Education Research Centre

White, M. (1987). The Japanese educational challenge: A commitment to children. New York, NY: Free Press.

Wong, N. Y. (2004). The CHC Learner's Phenomenon: Its Implications on mathematics education. In L. Fan, N. Y. Wong, & S. Li (Eds.), How Chinese learn mathematics: Perspectives from the insiders, Singapore: World Scientific.

Yang, H. (1999). Work roles and norms for teachers in Japan and the United States. In G. K. Letendre  (Ed.), Competitor or ally? Japans role in American educational debates. New York, NY: Falmer Press.

Zeng, K. (1999). Dragon Gate: Competitive examinations and their consequences. New York, NY: Cassell & Continuum.

Zhou, N., Shui-Fong, L., & Kam, C. C. (2012). The Chinese classroom paradox: A cross cultural comparison of teacher controlling behaviors. Journal of Educational Psychology, 104(4), 6274.

Zohar, A., Degani, A., & Vaaknin, E. (2001). Teachers beliefs about low-achieving students and higher order thinking. Teaching and Teacher Education, 17(4), 46985.

Appendix A

Description of Variables

Variables	Description
Student math achievement	Five plausible values of the math IRT scores (bsmmat01-05)
Instructional practice
Complex instruction	Average of four items: teachers reported frequency of asking students to explain the reasoning behind an idea (bt4masea), relate the learning topic to their daily lives (bt4masdl), work on problems with no immediately obvious method of solution (bt4masws), and decide on their own procedures for solving complex problems (bt4mascp). Recoded the original item from 1 (never) to 4 (every or almost every lesson).
Procedural instruction	Average of three items: teachers reported frequency of asking students to practice adding, subtracting, multiplying, and dividing without using a calculator (bt4maspc), memorize formulas and procedures (bt4masmf), and apply facts, concepts, and procedures to solve routine problems (bt4masac). Recoded the original item from 1 (never) to 4 (every or almost every lesson).
Student-centered instruction	Teachers report on the percentage of time students spend on working on own (bt4mptoo). Percentage is divided by 10.
Teacher-centered instruction	Teachers report on the percentage of time students spend listening to lecture-style presentation (bt4mptls) and listening to the teacher reteach and clarify content/procedures (bt4mptrt). Percentage is divided by 10.
Student Background
Male	Coded 1 if male and 0 if female (bs4gsex)
Foreign-born parent	Coded 1 if at least one parent is foreign born and 0 otherwise (bsdgborn)
Language at home	Coded 1 if student speaks language of test at home and 0 otherwise (bs4golan)
Books at home	Used bs4gbook. 5 = 1 to 10, 18 = 11 to 25, 64 = 26 to 100 , 151 = 101 to 200, 300 = over 200
Parent education	Coded 1 if at least one parent completed postsecondary degree and 0 otherwise (bsdgedup)
Teacher Background
Male teacher	Coded 1 if male and 0 if female (bt4gsex)
Teachers years of experience	Teachers report on their years of experience (bt4gtaut)
Teachers degree in math or math education	Coded 1 if majored in math or math education and 0 otherwise (bt4mpsma and bt4mpsem)
Class Characteristics
Class average number of books at home	Class average of students books at home
Class proportion of parents with college degree	Class average of parent education
Class size	Teachers report on class size (bt4mstud)
Different ability of students as limiting instruction	Teachers report on the extent to which different academic ability of students limits how they teach the class. Coded 1 if a lot and 0 otherwise (bt4mli01)
Uninterested students as limiting instruction	Teachers report on the extent to which uninterested students limit how they teach the class. Coded 1 if a lot and 0 otherwise (bt4mli04)
School Characteristics
School size	Principals report on school enrollment (bc4gtenr)
Student from affluent family > 50%	Principals report on students economic backgrounds. Coded 1 if more than 50% of students are from economically affluent homes and 0 otherwise (bc4gsbea)
School resource	Categorical variable derived by TIMSS using principal survey items. Coded 1 if school resource for math instruction is high and 0 if low or medium (bcdsrmi).
School climate	Categorical variable derived by TIMSS using principal survey items. Coded 1 if principals perception of school climate is high and 0 if low or medium (bcdgppsc)
School type (Japan)	School types are categorized to public schools in large or very large cities, public schools in small cities or noncity area (reference category), and private schools (idstrate)
School rank (Singapore)	School rank based on students performance (idstrati)

Note. All continuous variables are standardized to the country mean (for Japan and Singapore) when included as predictors in the multilevel models.

Appendix B

Means and Coefficient of Variations of Instructional Practices by Country

Country	Complex Instruction		Procedural Instruction		Student-centered Instruction		Teacher-centered Instruction
	Mean	Coef. of Var.	Mean	Coef. of Var.	Mean	Coef. of Var.	Mean	Coef. of Var.
ARM	2.51	21.38	2.58	36.75	1.58	49.52	3.44	32.50
AUS	2.44	24.47	2.56	22.71	2.40	55.74	2.65	38.35
BGR	2.88	19.02	3.37	15.32	1.69	52.18	2.86	44.50
BHR	2.79	23.74	3.05	18.83	1.20	50.27	3.72	32.42
BIH	2.84	19.02	2.87	21.02	1.50	54.25	3.96	36.27
BWA	2.78	21.05	3.01	20.98	2.10	54.62	2.32	39.14
COL	3.10	15.56	2.98	18.32	1.65	49.53	2.93	38.40
CYP	3.05	16.58	3.11	16.85	1.21	50.38	2.68	40.41
CZE	2.95	17.06	2.69	18.84	2.09	35.71	2.93	29.43
DZA	2.82	19.46	2.90	20.42	1.53	66.56	3.30	38.95
EGY	2.83	19.54	2.83	22.17	1.43	46.59	3.66	34.91
ENG	2.59	19.64	2.60	20.50	2.32	62.43	2.83	37.04
GEO	2.82	17.36	3.18	19.93	1.48	46.15	3.26	30.73
GHA	2.71	21.93	3.36	18.05	1.46	47.53	2.74	36.37
HKG	2.28	18.60	2.33	22.33	1.30	62.56	4.42	30.54
HUN	3.01	14.87	2.98	18.23	2.23	45.23	2.07	47.08
IDN	2.54	22.70	2.86	20.99	1.48	51.78	3.04	33.30
IRN	2.87	20.03	2.91	19.71	1.38	59.35	3.14	35.65
ISR	2.71	21.82	2.75	25.18	1.87	58.26	2.77	44.11
ITA	2.91	20.55	2.83	23.58	1.23	53.52	3.53	27.35
JOR	2.88	17.62	3.29	18.16	1.50	36.70	3.25	29.97
JPN	2.34	22.00	2.81	24.21	1.21	90.26	4.41	23.71
KOR	2.72	17.76	2.93	20.70	1.70	46.98	4.37	30.20
KWT	2.71	21.61	2.93	24.32	1.37	55.05	3.73	37.61
LBN	2.89	17.08	2.91	21.18	1.03	80.72	2.73	34.06
LTU	2.78	16.64	3.21	18.79	2.54	40.34	2.07	36.80
MAR	2.80	20.55	2.88	20.66	1.49	64.63	3.21	41.48
MLT	2.58	20.57	2.74	19.49	1.52	45.83	3.05	33.45
MNG	2.70	22.37	2.84	22.69	2.34	49.50	2.38	46.22
MYS	2.63	20.56	3.08	21.93	1.29	56.89	3.42	37.09
NOR	2.48	20.93	2.17	18.55	2.47	56.65	3.25	32.34
OMN	2.91	18.25	3.19	19.08	1.44	60.41	3.31	39.78
PSE	2.81	19.41	3.18	18.86	1.43	59.06	3.29	30.62
QAT	2.69	21.80	3.04	21.33	1.33	56.22	3.47	40.09
ROU	2.92	19.02	3.26	16.96	1.35	49.10	2.76	31.73
RUS	2.55	16.00	3.42	15.78	1.97	39.85	2.72	29.83
SAU	2.72	21.89	3.06	21.72	1.09	49.37	3.68	33.32
SCT	2.54	21.01	2.84	20.69	2.44	63.58	2.95	29.58
SGP	2.29	21.15	2.58	20.00	1.35	61.08	3.53	39.07
SLV	2.83	21.59	3.06	20.36	1.99	47.91	2.69	38.02
SRB	2.77	21.20	3.09	19.04	1.99	56.06	3.39	37.20
SVN	2.79	20.70	2.95	19.85	2.11	51.32	3.22	35.03
SWE	2.65	21.12	2.38	21.78	2.78	85.62	2.40	41.80
SYR	2.76	23.00	3.34	18.83	1.00	49.34	3.86	29.39
THA	2.85	21.28	3.07	21.00	1.20	55.43	3.59	35.17
TUN	2.67	21.23	3.05	22.10	1.61	74.11	3.03	46.07
TUR	2.89	17.85	3.06	20.03	1.28	63.68	3.43	36.78
TWN	2.40	19.42	2.75	19.67	0.74	64.32	5.07	24.50
UKR	2.73	16.82	3.39	15.11	1.81	38.75	3.12	30.03
USA	2.70	21.32	2.89	20.18	1.68	57.94	3.10	34.92
Total	2.73	19.92	2.94	20.47	1.64	55.18	3.21	35.47

Appendix C

Decomposition of Math Achievement Variance in the Five East Asian Countries

Variance	SGP	JPN	KOR	TWN	HKG
Between schools	1853.06	1360.97	2294.06	2588.31	5997.75
Between classrooms	4860.03	-	-	-	-
Within classrooms	2020.89	5747.24	5899.83	8088.63	3190.69
Variance at school level	21%	19%	28%	24%	65%
Variance at class level	56%	-	-	-	-
Variance at student level	23%	81%	72%	76%	35%

Cite This Article as: Teachers College Record Volume 120 Number 7, 2018, p. 1-42 https://www.tcrecord.org ID Number: 22388, Date Accessed: 10/22/2021 10:16:33 PM Purchase Reprint Rights for this article or review