View in EDS HTML Full Text PDF Full Text

A Decomposition of the Han-Minority Achievement Gap in Rural China

Saved in:
Bibliographic Details
Title: A Decomposition of the Han-Minority Achievement Gap in Rural China
Language: English
Authors: Xiyuan Jia, Xinwu Zhang (ORCID 0000-0002-4574-536X), Andrew Rule, Yue Ma
Source: Social Psychology of Education: An International Journal. 2025 28(1).
Availability: Springer. Available from: Springer Nature. One New York Plaza, Suite 4600, New York, NY 10004. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-460-1700; e-mail: customerservice@springernature.com; Web site: https://link.springer.com/
Peer Reviewed: Y
Publication Date: 2025
Document Type: Journal Articles
Reports - Research
Education Level: Elementary Education
Descriptors: Foreign Countries, Rural Schools, Ethnic Groups, Achievement Gap, Mathematics Achievement, Mathematics Anxiety, Student Attitudes, Elementary School Students, Minority Group Students, Correlation
Geographic Terms: China
DOI: 10.1007/s11218-025-10021-z
ISSN: 1381-2890
1573-1928
Abstract: Despite increased investment in rural education in recent years, the academic achievement of ethnic minority students in rural China continues to lag behind their Han Chinese peers. Few studies have used up-to-date data to measure the size of this achievement gap, and none have examined whether differential math attitudes and math anxiety play a role in perpetuating it. Using data from a survey of 16,747 students in 235 rural primary schools, we find a significant disparity in math performance between ethnic minority and Han students, with minority students scoring 0.25 standard deviations lower than their Han peers. Multivariate analysis reveals that differences in math affinity and math anxiety were significantly associated with math achievement among ethnic minority and Han students. Decomposition analysis further reveals that up to 37% of the achievement gap can be explained by variations in student, parent, and teacher characteristics. The difference in household asset was the largest explained contributor to the Han-minority achievement gap, followed by math affinity and math anxiety. This study underscores the importance of addressing students' math affinity and math anxiety in narrowing rural achievement gaps and promoting equitable educational outcomes in rural China and abroad.
Abstractor: As Provided
Entry Date: 2025
Accession Number: EJ1461163
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwEd03BRzOTUf9efI2201RSMAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDNaYrRM3FMRh0Qo8YQIBEICBmyJm6ukqHC9Pn5lSJs1ef2Pej05oErKc-_ULnkRrVY0xJh-tNQWXvIwPXPbhH4r5TBnpWlcEPffXAKLaXEhezStaMrf7ASwSLlmVyW40LOMMG8fV66RSs2UIK2Dbeb8ryIdhkmVttv7fIjHuOzRA92ujrEpUuTKftAg-V6cSL41rdLe66lvOI9WTVEavQ5OLWspQBsp-GAi4rCXA
Text:
  Availability: 1
  Value: <anid>AN0183282411;luo25feb.25;2025Feb28.03:44;v2.2.500</anid> <title id="AN0183282411-1">A decomposition of the Han-minority achievement gap in rural China </title> <p>Despite increased investment in rural education in recent years, the academic achievement of ethnic minority students in rural China continues to lag behind their Han Chinese peers. Few studies have used up-to-date data to measure the size of this achievement gap, and none have examined whether differential math attitudes and math anxiety play a role in perpetuating it. Using data from a survey of 16,747 students in 235 rural primary schools, we find a significant disparity in math performance between ethnic minority and Han students, with minority students scoring 0.25 standard deviations lower than their Han peers. Multivariate analysis reveals that differences in math affinity and math anxiety were significantly associated with math achievement among ethnic minority and Han students. Decomposition analysis further reveals that up to 37% of the achievement gap can be explained by variations in student, parent, and teacher characteristics. The difference in household asset was the largest explained contributor to the Han-minority achievement gap, followed by math affinity and math anxiety. This study underscores the importance of addressing students' math affinity and math anxiety in narrowing rural achievement gaps and promoting equitable educational outcomes in rural China and abroad.</p> <p>Keywords: Han-minority achievement gap; Rural China; Math attitudes; Math anxiety; Education Specialist Studies In Education</p> <p>Copyright comment Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.</p> <hd id="AN0183282411-2">Introduction</hd> <p>Increased investment in education in rural China has contributed to a significant rise in rural educational attainment in recent decades (Yang et al., [<reflink idref="bib70" id="ref1">70</reflink>]). Although access to education improved for the rural population as a whole, a large gap in education attainment persists between the Han Chinese[<reflink idref="bib1" id="ref2">1</reflink>] majority and its 55 officially recognized ethnic minority groups[<reflink idref="bib2" id="ref3">2</reflink>] (Wang et al., [<reflink idref="bib63" id="ref4">63</reflink>]). 70.22% of China's 125 million ethnic minorities live in western, relatively rural provinces, and 91.6% of ethnic autonomous counties are in western provinces (Wang et al., [<reflink idref="bib63" id="ref5">63</reflink>]; Yang et al., [<reflink idref="bib71" id="ref6">71</reflink>]). Due to historical, natural, and geographical contexts, ethnic autonomous regions and regions with high ethnic-minority concentrations have more limited education resources and fewer professional teachers (Su & He, [<reflink idref="bib58" id="ref7">58</reflink>]).</p> <p>To improve equity in education, the Chinese government adopted a series of preferential policies to improve access to education for ethnic minorities. The 1980 Law on Regional and Ethnic Autonomy recommended extra subsidization in education for areas with high minority concentration beyond standard educational funding (Cherng et al., [<reflink idref="bib14" id="ref8">14</reflink>]). In 1984, the interior ethnic boarding schooling policy started recruiting Tibetan primary school graduates to complete secondary school and university studies in developed eastern cities of China (Leibold, [<reflink idref="bib35" id="ref9">35</reflink>]). Since the late 2000s, policies increasingly prioritized vocational education for ethnic minorities and committed to providing bilingual education in ethnic minority areas. More recently, the Chinese government developed and implemented the Prevention and Controlling School Dropout Policy at the year of 2017, which waived miscellaneous school fees and provided free books, a full lunch or extra meals subsidies for students from rural areas with low income (Wu, [<reflink idref="bib64" id="ref10">64</reflink>]).</p> <p>Despite these efforts, rural ethnic minority students have persistently lagged behind their rural Han Chinese peers in academic achievement. One 2015 study, for instance, found a gap of 0.3 SD in math scores and 0.2 SD in Chinese language scores between ethnic-minority and Han students in rural primary schools in Shaanxi, Gansu, and Qinghai provinces (Yang et al., [<reflink idref="bib71" id="ref11">71</reflink>]). A later nationwide survey suggests that this Han-minority achievement gap is not unique to these three provinces, but likely exists among primary school students throughout rural China (Wang et al., [<reflink idref="bib62" id="ref12">62</reflink>]).</p> <p>This academic achievement gap may have significant negative consequences for the future academic and employment outcomes of ethnic-minority students. Lower academic performance in primary school has been linked to high dropout rates in China's rural junior high schools (Yi et al., [<reflink idref="bib72" id="ref13">72</reflink>]). As China transitions away from an agricultural economy, minority students who performed poorly in high school or did not graduate may have difficulty competing in the labor market (Yang et al., [<reflink idref="bib71" id="ref14">71</reflink>]), contributing to the Han-minority income gap (Hannum & Wang, [<reflink idref="bib28" id="ref15">28</reflink>]). Large gaps in grades and fail rates exist between Han and minority students in higher education as well (Wu et al., [<reflink idref="bib65" id="ref16">65</reflink>]), possibly stemming in part from differing educational experiences and outcomes earlier in primary and secondary school.</p> <p>A small body of literature has attempted to identify factors contributing to the Han-minority achievement gap in rural China, but these studies lacked up-to-date data and only focused on external factors, such as student and family background. For instance, Hannum ([<reflink idref="bib27" id="ref17">27</reflink>]), an important early study documenting the existence of the Han-minority achievement gap, used data collected in 1992 in China and attributed the gap mainly to geographic distribution and family background. Likewise, Yang et al., ([<reflink idref="bib71" id="ref18">71</reflink>]) decomposed achievement gap data from 300 schools sampled from across rural Shaanxi, Gansu, and Qinghai provinces during the 2011/2012 academic year and found that the largest contributors to the Han-minority achievement gap were student and family background.</p> <p>Besides external socioeconomic factors, the math achievement of rural Chinese primary school students has been found to be associated with math attitude and anxiety, subjective factors often overlooked in the existing research on the Han-minority achievement gap. Using data collected in 2017/2018 from rural primary schools in Shaanxi and Taiwan, Fan et al., ([<reflink idref="bib21" id="ref19">21</reflink>]) found a significant negative association between math anxiety and math achievement. A study of 1,840 primary school students in 95 schools in rural Taiwan found that the degree to which a student enjoys math was positively associated with math achievement (Ma et al., [<reflink idref="bib38" id="ref20">38</reflink>]). Similarly, Cheo ([<reflink idref="bib13" id="ref21">13</reflink>]) found that math motivation played a healthy role in influencing student math performance among primary school students in rural China.</p> <hd id="AN0183282411-3">Theoretical framework</hd> <p>The stereotype threat hypothesis theorizes that minority students may experience a decline in performance when they are subtly reminded of their ethnic group and the associated prejudicial beliefs before completing a task (Steele & Aronson, [<reflink idref="bib56" id="ref22">56</reflink>]). Further exploration of the mechanism of stereotype threat and performance outcomes shows that minority students may experience a significant decline in attitude (Fogliati & Bussey, [<reflink idref="bib22" id="ref23">22</reflink>]) and heightened anxiety (Brodish & Devine, [<reflink idref="bib10" id="ref24">10</reflink>]; Seo & Lee, [<reflink idref="bib54" id="ref25">54</reflink>]), driven by the fear of confirming these stereotypes, which can lead to a self-fulfilling prophecy, wherein the stereotype threat diminishes their mathematical ability and contributes to persistent academic disparities.</p> <hd id="AN0183282411-4">Math attitudes and math achievement</hd> <p>Math attitudes refer to a cluster of patterns of beliefs and emotions about math (Daskalogianni & Simpson, [<reflink idref="bib17" id="ref26">17</reflink>]), including math affinity and math intrinsic motivation (Lin & Huang, [<reflink idref="bib36" id="ref27">36</reflink>]).</p> <p>The Reciprocal Theory indicated that math affinity and math achievement may be related in a vicious cycle (Carey et al., [<reflink idref="bib11" id="ref28">11</reflink>]). For instance, Pekrun et al. ([<reflink idref="bib45" id="ref29">45</reflink>]) found math affinity and math achievement to positively predict each other among 3,425 German fifth to ninth graders, indicating that math affinity predicted higher math achievement. Likewise, Putwain et al. ([<reflink idref="bib48" id="ref30">48</reflink>]) found that math affinity was positively and significantly associated with higher math achievement among 1,298 English primary school students.</p> <p>The self-determination theory defines intrinsic motivation as being engaged in an activity because of one's inherent interest in and pleasure from this activity rather than due to external contingencies (Ryan & Deci, [<reflink idref="bib52" id="ref31">52</reflink>]). In this study, we focus on intrinsic motivation for math. Various studies have shown that math intrinsic motivation positively predicts academic achievement in math (Kriegbaum et al., [<reflink idref="bib34" id="ref32">34</reflink>]; Stevens et al., [<reflink idref="bib57" id="ref33">57</reflink>]). Since intrinsic motivation does not depend on expectations of success, it could be key to encouraging math learning for groups of students who lag behind on math achievement.</p> <hd id="AN0183282411-5">Math anxiety and math achievement</hd> <p>Math attitudes are based on motivational and cognitive factors, whereas math anxiety is a specific emotional factor (Dowker et al., [<reflink idref="bib18" id="ref34">18</reflink>]). Math anxiety is defined as a feeling of fear and tension towards math, which can result to an individual's inability to do math (Mutegi et al., [<reflink idref="bib40" id="ref35">40</reflink>]). Numerous studies of primary school students worldwide, including in China, have found math anxiety to be associated with lower math achievement (Barroso et al., [<reflink idref="bib6" id="ref36">6</reflink>]; Ching et al., [<reflink idref="bib15" id="ref37">15</reflink>]; Erturan & Jansen, [<reflink idref="bib20" id="ref38">20</reflink>]; Fan et al., [<reflink idref="bib21" id="ref39">21</reflink>]; Pantoja et al., [<reflink idref="bib44" id="ref40">44</reflink>]; Ramirez et al., [<reflink idref="bib50" id="ref41">50</reflink>], [<reflink idref="bib49" id="ref42">49</reflink>]; Vukovic et al., [<reflink idref="bib61" id="ref43">61</reflink>]; Wu et al., [<reflink idref="bib66" id="ref44">66</reflink>], [<reflink idref="bib67" id="ref45">67</reflink>]; Zhang et al., [<reflink idref="bib74" id="ref46">74</reflink>]). For instance, math anxiety in 1st grade predicts math achievement through at least 3rd grade among American primary school students (Pantoja et al., [<reflink idref="bib44" id="ref47">44</reflink>]).</p> <p>To provide up-to-date empirical evidence for these theories, this paper uses data from a comparative sample of rural students from central and southwestern China to evaluate the Han-minority achievement gap and its contributing factors, including math attitude and math anxiety. We pursue three specific objectives in this paper. First, we describe and compare the household and individual characteristics of sampled Han and ethnic minority students in two regions of rural China. To the best of our knowledge, this study is the first to offer insight into the educational performance of understudied regional populations by comparing rural communities in China's central and southwestern regions. Second, we describe the school performance of the sampled students and measure the achievement gap between Han and ethnic minority students, and also estimate the current size of the Han-minority gap in student achievement while controlling for demographic characteristics. Finally, we use a decomposition strategy to analyze the factors contributing to this achievement gap.</p> <hd id="AN0183282411-6">Materials and methods</hd> <p></p> <hd id="AN0183282411-7">Ethical approval</hd> <p>Ethical approval for this study was granted by the Stanford University Institutional Review Board (IRB No. 35635). All subjects gave written informed consent in accordance with the Declaration of Helsinki. All subjects were free to choose whether they were willing to participate in the research.</p> <hd id="AN0183282411-8">Sampling selection</hd> <p>Data for this paper were collected from two rural sample populations, one in southwestern China (City A) and another in central China (City B and City C). City A has a relatively low levels of economic development reflected in its per-capita GDP of 46,984 RMB (7,265 USD), which is lower than the average per-capita GDP of the province as a whole (47,729 RMB, or 7,404 USD) (City A Bureau of Statistics, [<reflink idref="bib69" id="ref48">69</reflink>]). The majority of City A residents (55%) have rural household registration status. In 2019, rural residents had a per-capita disposable income of 14,586 RMB (2,256 USD), one of the lowest of all prefecture-level cities in Province A and less than half the disposable income of City A's urban residents (35,043 RMB; 5,419 USD) (City A Bureau of Statistics, [<reflink idref="bib69" id="ref49">69</reflink>]). City A's population is about 1.5 million, with a mix of Han and ethnic minority residents.</p> <p>The second sample was collected in December 2019 in six low-income counties and districts of two cities with low levels of economic development in a different province in central China. Four of the six counties and districts were located in City B, and the remaining two counties and districts were located in City C. City B's per-capita GDP (51,733 RMB; 7,426 USD) is lower than the average in this province as a whole (56,388 RMB, or 8,095 USD) (Province B Bureau of Statistics, [<reflink idref="bib30" id="ref50">30</reflink>]; City B Bureau of Statistics, [<reflink idref="bib32" id="ref51">32</reflink>]). City C has an upper-middle level average income, with a GDP per capita (76,312 RMB, or 10,955 USD) (Province B Bureau of Statistics, [<reflink idref="bib30" id="ref52">30</reflink>]). About 50% of the residents in City B and C have rural household registration status and have considerably lower income levels than urban residents. The per capita disposable income levels of rural residents in City B (14,473 RMB; 2078 USD) and City C (18,558 RMB; 2664 USD) in 2019 were about half the disposable income of City B's urban residents (31,305 RMB; 4,494 USD) and City C's urban residents (34,376 RMB; 4,935 USD), respectively (City B Bureau of Statistics, [<reflink idref="bib32" id="ref53">32</reflink>]; City C Bureau of Statistics, [<reflink idref="bib68" id="ref54">68</reflink>]).</p> <p>Two steps were used to select the samples in the three cities. First, the local education bureau of each county or municipality provided a list of all primary schools, yielding a total of 105 schools in City A, 46 schools in City B, and 84 schools in City C. Schools using non-pep textbooks[<reflink idref="bib3" id="ref55">3</reflink>] and non-six-year schools[<reflink idref="bib4" id="ref56">4</reflink>] were excluded. Second, one class in fourth grade, one class in fifth grade, and one class in sixth grade were randomly selected from each school in the sample, and all students from the selected classes were enrolled in the study (8,622 students in City A, 4,902 students in City B, and 7,696 students in City C).</p> <hd id="AN0183282411-9">Data collection</hd> <p>Trained research teams administered surveys in March 2021 (for the southwestern China sample) and December 2019 (for the central China sample). Surveys covered student academic performance, demographic characteristics (student and teacher), as described in more detail below. In each sample school, the research team followed strict protocols in managing the survey process.</p> <hd id="AN0183282411-10">Academic performance</hd> <p>A standardized math test was used in the survey to measure the academic performance of the sample students. The research team collaborated with the local education bureau and primary school math teachers to design a math test based entirely on the mathematics curriculum taught in Chinese primary schools. To prevent teachers from preparing students for the test, only teachers who did not teach students taking the test participated in test design. Standardized math exams have been used widely in previous research papers in China to measure academic performance as an outcome variable (Jia et al., [<reflink idref="bib31" id="ref57">31</reflink>]; Ma et al., [<reflink idref="bib37" id="ref58">37</reflink>]). The math teachers of the surveyed classes did not participate in designing the test, and students were not prepared for the material on the test in advance. Students were required to complete the test in 30 minutes, with proctors to minimize cheating and strictly enforce time limits. Student scores were standardized for analysis by scaling raw scores into z-scores.</p> <hd id="AN0183282411-11">Demographic characteristics</hd> <p>The survey also included questionnaires to collect information about student demographic characteristics (individual and household characteristics) and teacher demographic characteristics. Five questions focused on individual characteristics, specifically the student's ethnicity, gender, age, whether they were an only child, and whether they were boarding students.</p> <p>We included two measures of math attitudes: math affinity and math intrinsic motivation. First, students were asked to indicate the degree to which they liked math (0–100). Then, math intrinsic motivation, or the degree to which the student studied math for internal or personal reasons, was evaluated by a four-item Likert scale: I like to read books about mathematics; I am looking forward to taking math class; I do math problems because I love math; I am interested in the content of mathematics. For each item, students were asked to indicate their agreement on a four-point scale (strongly disagree, disagree, agree, or strongly agree). Cronbach's alpha was 0.77, meaning that the scale was a reliable measure of math intrinsic motivation.</p> <p>Math anxiety, defined as a feeling of fear and tension towards math (Mutegi et al., [<reflink idref="bib40" id="ref59">40</reflink>]), was evaluated by a five-item scale: I often worry that I will have difficulty learning math; I am nervous when I do my math homework; I am nervous when I do math problems; I feel inferior when I do math problems; I worry that I will not do well in math in the future. For each item, students were asked to indicate their agreement on a four-point scale (strongly disagree, disagree, agree, or strongly agree). Cronbach's alpha for math anxiety was 0.73, indicating that the scale was a reliable measure of math anxiety.</p> <p>To collect data on student household characteristics, students were asked whether their parents migrated to another city for work during most of the previous semester and what the highest levels of schooling their parents had attained were. Household income was estimated by asking students whether their household has access to a computer with internet, a refrigerator, an air conditioner, a microwave, and a car, respectively. These answers helped create an asset-based wealth index using principal component analysis (PCA) (Kolenikov & Angeles, [<reflink idref="bib33" id="ref60">33</reflink>]).</p> <p>The survey administered to math teachers was designed to understand math teacher demographics and work-related characteristics. Demographic data included teacher gender, age, and education level. Data on tenure (i.e. teaching experience in years) were also collected.</p> <hd id="AN0183282411-12">Statistical analysis</hd> <p>Analysis of survey data included three steps. The first was to compare the household and individual characteristics and math academic achievement of sampled Han and ethnic minority students. Specifically, the household and individual characteristics were determined by student gender, age in years, only child status, boarding status, standardized math affinity score, math anxiety index, math intrinsic motivation index, whether their parents had out-migrated to work for most of the previous semester, parental educational attainment, household asset index, teacher gender, teacher age, teacher education level, and teacher tenure. T-tests were used to measure the difference of household and individual characteristics and academic achievement between Han and minority students in the sample.</p> <p>Second, to estimate the Han-minority gap in student achievement while controlling for the observed attributes, we first generated results based on ordinary least squares (OLS) as the benchmark. The following function was used for a multivariate analysis with a vector for control variables:</p> <olist> <item> <ephtml> <math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><mrow><msub><mi>Y</mi><mrow><mi mathvariant="italic">ijh</mi></mrow></msub><mo>=</mo><mi>α</mi><mo>+</mo><msub><mi>β</mi><mn>1</mn></msub><msub><mtext>minority</mtext><mrow><mi mathvariant="italic">ijh</mi></mrow></msub><mo>+</mo><mi>γ</mi><msub><mi>x</mi><mrow><mi mathvariant="italic">ijh</mi></mrow></msub><mo>+</mo><msub><mrow><mi>δ</mi><mi>t</mi><mi>e</mi><mi>a</mi><mi>c</mi><mi>h</mi><mi>e</mi><mi>r</mi></mrow><mrow><mi mathvariant="italic">jh</mi></mrow></msub><mo>+</mo><msub><mrow><mi>θ</mi><mi mathvariant="normal">Φ</mi></mrow><mi>c</mi></msub><mo>+</mo><msub><mrow><mi>τ</mi><mi>S</mi></mrow><mi>h</mi></msub><mo>+</mo><msub><mi>ϵ</mi><mrow><mi mathvariant="italic">ijh</mi></mrow></msub></mrow></math> </ephtml> </item> </olist> <p>Graph</p> <p>where the dependent variable <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>Y</mi><mrow><mi mathvariant="italic">ijh</mi></mrow></msub></math> </ephtml> represents the math test score of student <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>i</mi></math> </ephtml> in school <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>h</mi></math> </ephtml> and class <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>j</mi></math> </ephtml> , and the score represents the standardized math test score. The variable <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>x</mi><mrow><mi mathvariant="italic">ijh</mi></mrow></msub></math> </ephtml> represents student demographic information, including individual and household characteristics. Individual characteristics include student gender (1 = female), age in years, only child status (1 = student is an only child), boarding status (1 = student is boarding at school), standardized math affinity score, math anxiety index, and math intrinsic motivation index. For the math anxiety index and math intrinsic motivation index, student responses to the survey items were weighted according to the generalized least squares (GLS) weighting procedure described in Appendix A of Anderson ([<reflink idref="bib2" id="ref61">2</reflink>]). The higher the index number, the higher the level of math anxiety or math intrinsic motivation.</p> <p>Household characteristics included whether the student's mother out-migrated for work (1 = mainly did not live at home in the past semester), whether the student's father out-migrated for work (1 = mainly did not live at home in the past semester), the education level of the parents (1 = at least one parent is a high school graduate), and the household asset index. The variable <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mi mathvariant="italic">teacher</mi></mrow><mrow><mi mathvariant="italic">jh</mi></mrow></msub></math> </ephtml> represents teacher demographic and work-related characteristics, including teacher gender (1 = female), age in years, education level (1 = completed associate degree or above), and tenure in years. County-level fixed effects ( <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi mathvariant="normal">Φ</mi><mtext>c</mtext></msub></math> </ephtml> ) were added in order to control for differences such as county policies and finances. School-level fixed effects ( <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mtext>S</mtext><mtext>h</mtext></msub></math> </ephtml> ) were also added in order to control for differences between schools, such as school environment. <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>ϵ</mi><mrow><mi mathvariant="italic">ijh</mi></mrow></msub></math> </ephtml> is an error term. The error term is allowed to be correlated at the school level to account for clustering effects. We also constructed Huber-White standard errors (adjusted standard errors at the school cluster level, relaxing the assumption that disturbance terms were independent and identically distributed within schools) to account for the clustered nature of our sample. To ensure the robust of OLS results, we also conducted hierarchical linear model (HLM) analysis, and propensity score matching (PSM) analysis.</p> <p>Lastly, a decomposition analysis was undertaken to observe to evaluate how much of the gap can be attributed to the systematic differences in characteristics between Han and minority students. Decomposition analysis involves partitioning the total variance into components and allows us to identify groups of factors that explain a significant portion of the achievement gap (Zaefarian et al., [<reflink idref="bib73" id="ref62">73</reflink>]). First, we estimate educational productions functions, or achievement regressions (Eq. (<reflink idref="bib1" id="ref63">1</reflink>)), that quantify returns to individual, family, and teacher-level characteristics. Second, we use the traditional Oaxaca-Blinder decomposition method (Blinder, [<reflink idref="bib8" id="ref64">8</reflink>]; Oaxaca, [<reflink idref="bib41" id="ref65">41</reflink>]) to decompose the achievement gap between Han students and minority students. We decompose the Han-minority gap into two components. The first is a component that can be explained by differences in student, family, and teacher-level characteristics. The second component is due to between-group differences in returns to characteristics. The following equation was used:</p> <p>2 <ephtml> <math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mi>D</mi><mo>=</mo><mi>E</mi><mfenced close=")" open="("><msub><mi>Y</mi><mi>H</mi></msub></mfenced><mo>-</mo><mi>E</mi><mfenced close=")" open="("><msub><mi>Y</mi><mi>M</mi></msub></mfenced><mo>=</mo><msub><mi>β</mi><mi>H</mi></msub><mfenced close="}" open="{"><mi>E</mi><mfenced close=")" open="("><msub><mi>X</mi><mi>H</mi></msub></mfenced><mo>-</mo><mi>E</mi><mfenced close=")" open="("><msub><mi>X</mi><mi>M</mi></msub></mfenced></mfenced><mo>+</mo><mfenced close=")" open="("><msub><mi>β</mi><mi>H</mi></msub><mo>-</mo><msub><mi>β</mi><mi>M</mi></msub></mfenced><mrow /><mo>∗</mo><mi>E</mi><mfenced close=")" open="("><msub><mi>X</mi><mi>M</mi></msub></mfenced></mrow></math> </ephtml></p> <p>Graph</p> <p>where <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mtext>Y</mtext><mtext>H</mtext></msub></math> </ephtml> and <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mtext>Y</mtext><mtext>M</mtext></msub></math> </ephtml> are the predicted mean standardized test scores of Han and minority students, <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mtext>X</mtext><mtext>H</mtext></msub></math> </ephtml> and <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mtext>X</mtext><mtext>M</mtext></msub></math> </ephtml> are the mean characteristics of Han and minority students (student, family, and teacher-level characteristics), and <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi mathvariant="normal">β</mi><mtext>H</mtext></msub></math> </ephtml> and <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi mathvariant="normal">β</mi><mtext>M</mtext></msub></math> </ephtml> are the returns to characteristics for Han and minority students estimated using Eq. (<reflink idref="bib1" id="ref66">1</reflink>) above. Equation (<reflink idref="bib2" id="ref67">2</reflink>) represents the crude difference between Han and minority students conditional on characteristics, which can be decomposed into two components. The first is the portion attributable to differences in the quantity of characteristics, evaluated using Han returns: <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mrow><msub><mi mathvariant="normal">β</mi><mtext>H</mtext></msub><mfenced close="}" open="{"><msub><mtext>X</mtext><mtext>H</mtext></msub><mo>-</mo><msub><mtext>X</mtext><mtext>M</mtext></msub></mfenced></mrow></math> </ephtml> , which became the "explained gap." The other portion, <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mfenced close=")" open="("><msub><mi mathvariant="normal">β</mi><mtext>H</mtext></msub><mo>-</mo><msub><mi mathvariant="normal">β</mi><mtext>M</mtext></msub></mfenced><mrow /><mo>∗</mo><msub><mtext>X</mtext><mtext>M</mtext></msub></mrow></math> </ephtml> , is that attributable to differences in returns to the characteristics of Han and minority students, which became the "unexplained gap." All analyses were conducted in STATA 18. P-values less than 0.05 were considered statistically significant.</p> <hd id="AN0183282411-13">Results</hd> <p></p> <hd id="AN0183282411-14">Demographic characteristics</hd> <p>We compared the characteristics of 16,747 sampled Han and ethnic minority students from 235 primary schools in two communities located in central and southwestern China. Table 1 shows the descriptive statistics of the demographic characteristics of the ethnic minority sample (N = 1,299), the rural southwest sample (N = 6,874), and the rural central sample (N = 8,574). The ethnic minority sample consisted of ethnic minority students and their teachers from City A in southwestern China. The rural southwest sample consisted of Han students and their teachers from City A, and the rural central sample consisted of Han students and their teachers from Cities B and C. As some teachers in City A taught both ethnic minority and Han students, these teachers were assigned to both the ethnic minority and the rural southwest samples.</p> <p>Table 1 Descriptive statistics of demographic characteristics</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="3"><p>Variables</p></th><th align="left"><p>Ethnic sample (N = 1,299)</p></th><th align="left"><p>Rural southwest sample (N = 6,874)</p></th><th align="left"><p>Rural central sample (N = 8,574)</p></th><th align="left" colspan="2"><p>Ethnic VS Rural southwest</p></th><th align="left" colspan="2"><p>Ethnic VS Rural central</p></th><th align="left" colspan="2"><p>Rural southwest VS Rural central</p></th></tr><tr><th align="left"><p>Mean/Share</p></th><th align="left"><p>Mean/Share</p></th><th align="left"><p>Mean/Share</p></th><th align="left"><p>Difference</p></th><th align="left"><p><italic>p</italic>-value</p></th><th align="left"><p>Difference</p></th><th align="left"><p><italic>p</italic>-value</p></th><th align="left"><p>Difference</p></th><th align="left"><p><italic>p</italic>-value</p></th></tr><tr><th align="left"><p>(1)</p></th><th align="left"><p>(2)</p></th><th align="left"><p>(3)</p></th><th align="left"><p>(4) = (2)–(1)</p></th><th align="left"><p>(5)</p></th><th align="left"><p>(6) = (3)—(1)</p></th><th align="left"><p>(7)</p></th><th align="left"><p>(8) = (3)—(2)</p></th><th align="left"><p>(9)</p></th></tr></thead><tbody><tr><td align="left" colspan="10"><p><italic>Personal characteristics</italic></p></td></tr><tr><td align="left"><p>Age (years)</p></td><td align="left"><p>11.84 (1.31)</p></td><td align="left"><p>11.27 (0.97)</p></td><td align="left"><p>1.99 (0.95)</p></td><td align="left"><p>− 0.58 [0.03]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>− 0.85 [0.03]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>− 0.27 [0.02]</p></td><td align="left"><p> < 0.001</p></td></tr><tr><td align="left"><p>Gender (1 = girl)</p></td><td align="left"><p>0.53 (0.5)</p></td><td align="left"><p>0.5 (0.5)</p></td><td align="left"><p>0.47 (0.5)</p></td><td align="left"><p>− 0.03 [0.02]</p></td><td align="left"><p>0.028</p></td><td align="left"><p>− 0.06 [0.01]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>− 0.02 [0.01]</p></td><td align="left"><p>0.006</p></td></tr><tr><td align="left"><p>Only child (1 = yes)</p></td><td align="left"><p>0.08 (0.28)</p></td><td align="left"><p>0.18 (0.39)</p></td><td align="left"><p>0.06 (0.23)</p></td><td align="left"><p>0.1 [0.01]</p></td><td align="left"><p>< 0.001</p></td><td align="left"><p>− 0.03 [0.01]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>− 0.13 [0.01]</p></td><td align="left"><p> < 0.001</p></td></tr><tr><td align="left"><p>Boarding at school (1 = yes)</p></td><td align="left"><p>0.33 (0.47)</p></td><td align="left"><p>0.05 (0.22)</p></td><td align="left"><p>0.26 (0.44)</p></td><td align="left"><p>− 0.28 [0.01]</p></td><td align="left"><p>< 0.001</p></td><td align="left"><p>− 0.07 [0.01]</p></td><td align="left"><p>< 0.001</p></td><td align="left"><p>0.21 [0.01]</p></td><td align="left"><p>< 0.001</p></td></tr><tr><td align="left"><p>Math affinity Z-score</p></td><td align="left"><p>− 0.09 (1.06)</p></td><td align="left"><p>− 0.01 (0.97)</p></td><td align="left"><p>0.04 (0.98)</p></td><td align="left"><p>0.08 [0.03]</p></td><td align="left"><p>0.008</p></td><td align="left"><p>0.13 [0.03]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>0.05 [0.02]</p></td><td align="left"><p>0.002</p></td></tr><tr><td align="left"><p>Math anxiety index</p></td><td align="left"><p>0.05 (0.67)</p></td><td align="left"><p>− 0.01 (0.71)</p></td><td align="left"><p>− 0.02 (0.68)</p></td><td align="left"><p>− 0.07 [0.02]</p></td><td align="left"><p>0.002</p></td><td align="left"><p>− 0.07 [0.02]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>0.00 [0.01]</p></td><td align="left"><p>0.738</p></td></tr><tr><td align="left"><p>Math intrinsic motivation index</p></td><td align="left"><p>− 0.01 (0.77)</p></td><td align="left"><p>0 (0.78)</p></td><td align="left"><p>0.03 (0.75)</p></td><td align="left"><p>0.01 [0.02]</p></td><td align="left"><p>0.751</p></td><td align="left"><p>0.03 [0.02]</p></td><td align="left"><p>0.151</p></td><td align="left"><p>0.02 [0.01]</p></td><td align="left"><p>0.045</p></td></tr><tr><td align="left" colspan="3"><p><italic>Household characteristics</italic></p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p>Parental Education (1 = at least one completed high or above)</p></td><td align="left"><p>0.11 (0.31)</p></td><td align="left"><p>0.31 (0.46)</p></td><td align="left"><p>0.44 (0.5)</p></td><td align="left"><p>0.2 [0.01]</p></td><td align="left"><p>< 0.001</p></td><td align="left"><p>0.33 [.01]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>0.13 [0.01]</p></td><td align="left"><p> < 0.001</p></td></tr><tr><td align="left"><p>Mother out-migrated for work</p></td><td align="left"><p>0.2 (0.4)</p></td><td align="left"><p>0.23 (0.42)</p></td><td align="left"><p>0.24 (0.43)</p></td><td align="left"><p>0.03 [0.01]</p></td><td align="left"><p>0.024</p></td><td align="left"><p>0.04 [0.01]</p></td><td align="left"><p>0.002</p></td><td align="left"><p>0.01 [0.01]</p></td><td align="left"><p>0.146</p></td></tr><tr><td align="left"><p>Father out-migrated for work</p></td><td align="left"><p>0.38 (0.49)</p></td><td align="left"><p>0.43 (0.5)</p></td><td align="left"><p>0.51 (0.5)</p></td><td align="left"><p>0.05 [0.01]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>0.14 [0.01]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>0.08 [0.01]</p></td><td align="left"><p> < 0.001</p></td></tr><tr><td align="left"><p>Asset index (PCA score)</p></td><td align="left"><p>− 0.77 (1.19)</p></td><td align="left"><p>0.15 (1.28)</p></td><td align="left"><p>0.07 (1.23)</p></td><td align="left"><p>0.92 [0.04]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>0.84 [0.04]</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>− 0.08 [0.02]</p></td><td align="left"><p> < 0.001</p></td></tr><tr><td align="left" colspan="2"><p><italic>Teacher characteristics</italic></p></td><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td align="left"><p>Teacher age (year)</p></td><td align="left"><p>39.51 (11.04)</p></td><td align="left"><p>39.49 (11.3)</p></td><td align="left"><p>4.28 (8.93)</p></td><td align="left"><p>–</p></td><td align="left"><p>–</p></td><td align="left"><p>0.77 (0.86)</p></td><td align="left"><p>0.369</p></td><td align="left"><p>0.79 (0.77)</p></td><td align="left"><p>0.303</p></td></tr><tr><td align="left"><p>Teacher gender (1 = female)</p></td><td align="left"><p>0.52 (0.5)</p></td><td align="left"><p>0.52 (0.5)</p></td><td align="left"><p>0.63 (0.48)</p></td><td align="left"><p>–</p></td><td align="left"><p>–</p></td><td align="left"><p>0.11 (0.04)</p></td><td align="left"><p>0.011</p></td><td align="left"><p>0.11 (0.04)</p></td><td align="left"><p>0.003</p></td></tr><tr><td align="left"><p>Associate education (1 = yes)</p></td><td align="left"><p>0.61 (0.49)</p></td><td align="left"><p>0.61 (0.49)</p></td><td align="left"><p>0.52 (0.5)</p></td><td align="left"><p>–</p></td><td align="left"><p>–</p></td><td align="left"><p>− 0.09 (0.04)</p></td><td align="left"><p>0.047</p></td><td align="left"><p>− 0.09 (0.04)</p></td><td align="left"><p>0.021</p></td></tr><tr><td align="left"><p>Tenure (year)</p></td><td align="left"><p>17.62 (12.9)</p></td><td align="left"><p>17.64 (13.11)</p></td><td align="left"><p>18.53 (1.98)</p></td><td align="left"><p>–</p></td><td align="left"><p>–</p></td><td align="left"><p>0.91 (1.03)</p></td><td align="left"><p>0.378</p></td><td align="left"><p>0.89 (0.91)</p></td><td align="left"><p>0.330</p></td></tr></tbody></table> </ephtml> </p> <p> <emph>Notes</emph>: Standard deviations in parentheses. Standard errors in square brackets. "Mean" represents the average for continuous variables, and "Share" represents the proportion for categorical variables</p> <p> <emph>Source</emph>: authors' survey</p> <p></p> <p>Significant differences in demographic characteristics were observed between the ethnic minority sample and the rural southwest and rural central samples. As shown in Table 1, in comparison to the ethnic minority sample, the other two samples were younger by 0.58 and 0.85 years on average (<emph>p</emph> < 0.001), had lower proportions of female students by 3% (<emph>p</emph> = 0.028) and 6% (<emph>p</emph> < 0.001), and had lower proportions of boarding students by 28% and 7% (<emph>p</emph> < 0.001), respectively. However, the rural southwest sample, compared to the ethnic minority sample, exhibited a higher proportion of only children by 10% (<emph>p</emph> < 0.001), while the rural central sample showed a lower proportion of only children by 3% (<emph>p</emph> < 0.001). In terms of math-related variables, the ethnic Han samples exhibited higher math affinity Z-scores by 0.08 (<emph>p</emph> = 0.008) and 0.13 (<emph>p</emph> < 0.001) and lower math anxiety indices by 0.07 (<emph>p</emph> = 0.002) and 0.07 (p < 0.001).</p> <p>In terms of household characteristics, the results in Table 1 show notable gaps between the ethnic minority sample and the rural southwest and rural central samples. Compared to the ethnic minority sample, the rural southwest and rural central samples had a higher proportion of parents with higher education by 20% and 33% (<emph>p</emph> < 0.001), a higher proportion of parents who out-migrated for work by 3% (<emph>p</emph> = 0.024) and 14% (<emph>p</emph> < 0.001), and a higher average household asset index (PCA score) by 0.92 and 0.84 (<emph>p</emph> < 0.001), respectively.</p> <p>Regarding teacher characteristics, as shown in Table 1, the rural central sample had a higher percentage of female teachers by 11% (<emph>p</emph> = 0.011) and a lower education attainment by 9% (<emph>p</emph> < 0.05) than the ethnic minority sample. As there was overlap between teachers in the ethnic minority and rural southwest samples, we did not conduct comparisons between them.</p> <hd id="AN0183282411-15">The Han-minority achievement gap</hd> <p>Table 2 shows the math performance differences between the ethnic minority sample, the rural southwest sample, and the rural central sample. Kernel density plots showed that math performance was normally distributed across all samples (Appendix Figs. 1, 2, 3, 4). The ethnic minority sample had significantly lower math performance than both the rural southwest and rural central samples at all grade levels. Moreover, the size of this achievement gap grew considerably between 4th and 5th grade, and again between 5th and 6th grade. As shown in Columns 7 and 9, the overall ethnic minority sample had lower math performance than either rural sample across all grades, with a difference of 0.264 (<emph>p</emph> < 0.001) and 0.251 (<emph>p</emph> < 0.001). The rural southwest and rural central samples did not differ significantly in math performance at any grade level.</p> <p>Table 2 Math performance differences between ethnic, rural southeast and rural central children</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="3" colspan="2"><p>Groups</p></th><th align="left" colspan="2"><p>Ethnic sample (N = 1,299)</p></th><th align="left" colspan="2"><p>Rural southwest sample (N = 6,874)</p></th><th align="left" colspan="2"><p>Rural central sample (N = 8,574)</p></th><th align="left" colspan="2"><p>Ethnic VS Rural southwest</p></th><th align="left" colspan="2"><p>Ethnic VS Rural central</p></th><th align="left" colspan="2"><p>Rural southwest VS Rural central</p></th></tr><tr><th align="left"><p>Obs</p></th><th align="left"><p>Mean (SD)</p></th><th align="left"><p>Obs</p></th><th align="left"><p>Mean (SD)</p></th><th align="left"><p>Obs</p></th><th align="left"><p>Mean (SD)</p></th><th align="left"><p>Difference</p></th><th align="left"><p><italic>p</italic>-value</p></th><th align="left"><p>Difference</p></th><th align="left"><p><italic>p</italic>-value</p></th><th align="left"><p>Difference</p></th><th align="left"><p><italic>p</italic>-value</p></th></tr><tr><th align="left"><p>(1)</p></th><th align="left"><p>(2)</p></th><th align="left"><p>(3)</p></th><th align="left"><p>(4)</p></th><th align="left"><p>(5)</p></th><th align="left"><p>(6)</p></th><th align="left"><p>(7) = (4)—(2)</p></th><th align="left"><p>(8)</p></th><th align="left"><p>(9) = (6)—(2)</p></th><th align="left"><p>(10)</p></th><th align="left"><p>(11) = (6)—(4)</p></th><th align="left"><p>(12)</p></th></tr></thead><tbody><tr><td align="left"><p>(1)</p></td><td align="left"><p>Grade 4</p></td><td align="left"><p>385</p></td><td align="left"><p>−.071</p></td><td align="left"><p>2,088</p></td><td align="left"><p>0.052</p></td><td align="left"><p>2,645</p></td><td align="left"><p>0.027</p></td><td align="left"><p>0.123</p></td><td align="left"><p>0.024</p></td><td align="left"><p>0.098</p></td><td align="left"><p>0.067</p></td><td align="left"><p>− 0.025</p></td><td align="left"><p>0.381</p></td></tr><tr><td align="left" /><td align="left" /><td align="left" /><td align="left"><p>(0.935)</p></td><td align="left" /><td align="left"><p>(0.989)</p></td><td align="left" /><td align="left"><p>(0.983)</p></td><td align="left"><p>(0.054)</p></td><td align="left" /><td align="left"><p>(0.053)</p></td><td align="left" /><td align="left"><p>(0.029)</p></td><td align="left" /></tr><tr><td align="left"><p>(2)</p></td><td align="left"><p>Grade 5</p></td><td align="left"><p>473</p></td><td align="left"><p>− 0.210</p></td><td align="left"><p>2,293</p></td><td align="left"><p>0.069</p></td><td align="left"><p>3,104</p></td><td align="left"><p>0.057</p></td><td align="left"><p>0.279</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>0.267</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>− 0.012</p></td><td align="left"><p>0.655</p></td></tr><tr><td align="left" /><td align="left" /><td align="left" /><td align="left"><p>(1.077)</p></td><td align="left" /><td align="left"><p>(0.959)</p></td><td align="left" /><td align="left"><p>(0.959)</p></td><td align="left"><p>(0.049)</p></td><td align="left" /><td align="left"><p>(0.048)</p></td><td align="left" /><td align="left"><p>(0.026)</p></td><td align="left" /></tr><tr><td align="left"><p>(3)</p></td><td align="left"><p>Grade 6</p></td><td align="left"><p>441</p></td><td align="left"><p>− 0.307</p></td><td align="left"><p>2,493</p></td><td align="left"><p>0.064</p></td><td align="left"><p>2,825</p></td><td align="left"><p>0.061</p></td><td align="left"><p>0.372</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>0.368</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>− 0.003</p></td><td align="left"><p>0.907</p></td></tr><tr><td align="left" /><td align="left" /><td align="left" /><td align="left"><p>(1.001)</p></td><td align="left" /><td align="left"><p>(0.984)</p></td><td align="left" /><td align="left"><p>(0.965)</p></td><td align="left"><p>(0.051)</p></td><td align="left" /><td align="left"><p>(0.050)</p></td><td align="left" /><td align="left"><p>(0.027)</p></td><td align="left" /></tr><tr><td align="left"><p>(4)</p></td><td align="left"><p>All samples</p></td><td align="left"><p>1,299</p></td><td align="left"><p>− 0.202</p></td><td align="left"><p>6,874</p></td><td align="left"><p>0.062</p></td><td align="left"><p>8,574</p></td><td align="left"><p>0.049</p></td><td align="left"><p>0.264</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>0.251</p></td><td align="left"><p> < 0.001</p></td><td align="left"><p>− 0.013</p></td><td align="left"><p>0.408</p></td></tr><tr><td align="left" /><td align="left" /><td align="left" /><td align="left"><p>(1.014)</p></td><td align="left" /><td align="left"><p>(0.977)</p></td><td align="left" /><td align="left"><p>(0.968)</p></td><td align="left"><p>(0.030)</p></td><td align="left" /><td align="left"><p>(0.029)</p></td><td align="left" /><td align="left"><p>(0.016)</p></td><td align="left" /></tr></tbody></table> </ephtml> </p> <p> <emph> Notes</emph>: Standard deviation in parentheses</p> <p> <emph>Source</emph>: authors' survey</p> <p></p> <hd id="AN0183282411-16">Correlations between ethnic minority status, math attitude, math anxiety and math performance</hd> <p>Table 3 presents the results of the multivariate analysis examining the relationship between ethnic minority status and math performance across different grade levels. We estimated four models to explore these relationships, with each model representing a different set of variables and controls.</p> <p>Table 3 OLS Multivariate correlations between demographic characteristics, ethnic minority and math performance</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="3" colspan="2"><p>Variables</p></th><th align="left" colspan="16"><p>Math performance</p></th></tr><tr><th align="left" colspan="4"><p>Grade 4</p></th><th align="left" colspan="4"><p>Grade 5</p></th><th align="left" colspan="4"><p>Grade 6</p></th><th align="left" colspan="4"><p>All sample</p></th></tr><tr><th align="left"><p>(1)</p></th><th align="left"><p>(2)</p></th><th align="left"><p>(3)</p></th><th align="left"><p>(4)</p></th><th align="left"><p>(5)</p></th><th align="left"><p>(6)</p></th><th align="left"><p>(7)</p></th><th align="left"><p>(8)</p></th><th align="left"><p>(9)</p></th><th align="left"><p>(10)</p></th><th align="left"><p>(11)</p></th><th align="left"><p>(12)</p></th><th align="left"><p>(13)</p></th><th align="left"><p>(14)</p></th><th align="left"><p>(15)</p></th><th align="left"><p>(16)</p></th></tr></thead><tbody><tr><td align="left"><p>(1)</p></td><td align="left"><p>Ethnic minority (1 = yes)</p></td><td align="left"><p>− 0.11</p></td><td align="left"><p>0.01</p></td><td align="left"><p>0.02</p></td><td align="left"><p>− 0.00</p></td><td align="left"><p>− 0.27***</p></td><td align="left"><p>− 0.15**</p></td><td align="left"><p>− 0.14**</p></td><td align="left"><p>− 0.17**</p></td><td align="left"><p>− 0.37***</p></td><td align="left"><p>− 0.12*</p></td><td align="left"><p>− 0.12*</p></td><td align="left"><p>− 0.17**</p></td><td align="left"><p>− 0.26***</p></td><td align="left"><p>− 0.17***</p></td><td align="left"><p>− 0.16***</p></td><td align="left"><p>− 0.17***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.10)</p></td><td align="left"><p>(0.11)</p></td><td align="left"><p>(0.10)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.06)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.04)</p></td></tr><tr><td align="left"><p>(2)</p></td><td align="left"><p>Personal characteristics</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(3)</p></td><td align="left"><p>Household characteristics</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(4)</p></td><td align="left"><p>Teacher characteristics</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(5)</p></td><td align="left"><p>School fixed effect</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(6)</p></td><td align="left"><p>County fixed effect</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(7)</p></td><td align="left"><p>Observations</p></td><td align="left"><p>5,118</p></td><td align="left"><p>5,118</p></td><td align="left"><p>5,118</p></td><td align="left"><p>5,118</p></td><td align="left"><p>5,870</p></td><td align="left"><p>5,870</p></td><td align="left"><p>5,870</p></td><td align="left"><p>5,870</p></td><td align="left"><p>5,759</p></td><td align="left"><p>5,759</p></td><td align="left"><p>5,759</p></td><td align="left"><p>5,759</p></td><td align="left"><p>16,747</p></td><td align="left"><p>16,747</p></td><td align="left"><p>16,747</p></td><td align="left"><p>16,747</p></td></tr><tr><td align="left"><p>(8)</p></td><td align="left"><p>R-squared</p></td><td align="left"><p>0.00</p></td><td align="left"><p>0.09</p></td><td align="left"><p>0.09</p></td><td align="left"><p>0.23</p></td><td align="left"><p>0.01</p></td><td align="left"><p>0.13</p></td><td align="left"><p>0.13</p></td><td align="left"><p>0.24</p></td><td align="left"><p>0.01</p></td><td align="left"><p>0.15</p></td><td align="left"><p>0.15</p></td><td align="left"><p>0.27</p></td><td align="left"><p>0.00</p></td><td align="left"><p>0.11</p></td><td align="left"><p>0.12</p></td><td align="left"><p>0.19</p></td></tr></tbody></table> </ephtml> </p> <p> <emph>Notes</emph>: Personal characteristics include the student's gender, age, whether they were an only child, whether they were boarding students, math affinity, math intrinsic motivation, and math anxiety. Household characteristics include whether their parents migrated to another city for work during most of the previous semester, the highest levels of schooling their parents had attained, and household asset. Teacher characteristics include teacher gender, age, education level, and teaching experience in years. Standard errors in parentheses; *** <emph>p</emph> < 0.01, ** <emph>p</emph> < 0.05, * <emph>p</emph> < 0.1. Standard errors (in parentheses) account for clustering at the school level</p> <p> <emph>Source</emph>: authors' survey</p> <p></p> <p>Model 1, represented by columns 1, 5, 9, and 13, investigates the relationship between ethnic minority status and math performance without any controls or fixed effects. The model results suggest that being an ethnic minority is significantly and negatively correlated with math performance across the whole sample, as well as in the 5th- and 6th-grade subsamples. Specifically, the coefficients in Model 1 range from -0.26 to -0.37 (<emph>p</emph> < 0.01). Model 2, represented by columns 2, 6, 10, and 14, includes personal and household characteristics as control variables while also excluding fixed effects. Again, ethnic minority status is significantly and negatively correlated with math performance across the whole sample and in the 5th-grade subsamples, with coefficients ranging from -0.15 to -0.17. Model 3 (columns 3, 7, 11, and 15), which incorporates teacher characteristics as control variables in addition to all characteristics included in Model 2, again shows that being part of an ethnic minority is negatively correlated with math performance across the whole sample and in the 5th-grade subsamples, with coefficients ranging from -0.14 to -0.16. Finally, Model 4 (columns 4, 8, 12, and 16) introduces county fixed effects and school fixed effects to capture unobserved county- and school-level characteristics that may affect math performance, such as local educational policies, resources, or socioeconomic factors. Even controlling for these fixed effects, we find that ethnic minority status is significantly and negatively correlated with math performance across the whole sample and in the 5th- and 6th-grade subsamples, with all the coefficients being -0.17. We also conducted HLM analysis to check the robustness of OLS results (Appendix Table 1). Results indicate that all the coefficients are consistent with the results of OLS analysis. Considering that some minority parents may send high-performing children to better educational opportunities elsewhere, we further conducted PSM to avoid overestimating the ethnic impact (Appendix Table 2). The tests of covariate matching quality, reported in Appendix Table 3, show that after matching, all the covariates in the two groups were no longer significantly different, compared to 14 covariates being significant pre-matching. Additionally, a sensitivity analysis using Rosenbaum bounds (Rosenbaum & Rubin, [<reflink idref="bib51" id="ref68">51</reflink>]) suggests that the treatment effects remain robust though these students have been allowed to differ in their odds of being treated up to (maximum value of gamma 2 with 0.25 increments) in terms of unobserved covariates (Appendix Table 4). These findings confirm the robustness of our results, which are consistent with the OLS analysis.</p> <p>Table 4 Oaxaca decomposition</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="3"><p>Variables</p></th><th align="left" colspan="2"><p>Full samples</p></th></tr><tr><th align="left"><p>No School/County FE</p></th><th align="left"><p>School and County FE</p></th></tr><tr><th align="left"><p>(1)</p></th><th align="left"><p>(2)</p></th></tr></thead><tbody><tr><td align="left"><p>Han residents</p></td><td align="left"><p>0.055***</p></td><td align="left"><p>0.055***</p></td></tr><tr><td align="left" /><td align="left"><p>(0.020)</p></td><td align="left"><p>(0.019)</p></td></tr><tr><td align="left"><p>Minority residents</p></td><td align="left"><p>− 0.202***</p></td><td align="left"><p>− 0.202***</p></td></tr><tr><td align="left" /><td align="left"><p>(0.045)</p></td><td align="left"><p>(0.041)</p></td></tr><tr><td align="left"><p>Difference</p></td><td align="left"><p>0.257***</p></td><td align="left"><p>0.257***</p></td></tr><tr><td align="left" /><td align="left"><p>(0.047)</p></td><td align="left"><p>(0.044)</p></td></tr><tr><td align="left"><p>Explained</p></td><td align="left"><p>0.096***</p></td><td align="left"><p>0.063**</p></td></tr><tr><td align="left" /><td align="left"><p>(0.020)</p></td><td align="left"><p>(0.026)</p></td></tr><tr><td align="left"><p>Unexplained</p></td><td align="left"><p>0.161***</p></td><td align="left"><p>0.194***</p></td></tr><tr><td align="left" /><td align="left"><p>(0.049)</p></td><td align="left"><p>(0.044)</p></td></tr><tr><td align="left"><p><italic>Explained</italic></p></td><td align="left" /><td align="left" /></tr><tr><td align="left"><p>Student characteristics</p></td><td align="left" /><td align="left" /></tr><tr><td align="left"><p>Age (years)</p></td><td align="left"><p>0.008</p></td><td align="left"><p>0.011</p></td></tr><tr><td align="left" /><td align="left"><p>(0.009)</p></td><td align="left"><p>(0.009)</p></td></tr><tr><td align="left"><p>Gender (1 = girl)</p></td><td align="left"><p>0.000</p></td><td align="left"><p>0.001</p></td></tr><tr><td align="left" /><td align="left"><p>(0.001)</p></td><td align="left"><p>(0.001)</p></td></tr><tr><td align="left"><p>Only child (1 = yes)</p></td><td align="left"><p>− 0.002*</p></td><td align="left"><p>− 0.003**</p></td></tr><tr><td align="left" /><td align="left"><p>(0.001)</p></td><td align="left"><p>(0.001)</p></td></tr><tr><td align="left"><p>Boarding at school (1 = yes)</p></td><td align="left"><p>0.006</p></td><td align="left"><p>0.002</p></td></tr><tr><td align="left" /><td align="left"><p>(0.009)</p></td><td align="left"><p>(0.008)</p></td></tr><tr><td align="left"><p>Math affinity Z score</p></td><td align="left"><p>0.020**</p></td><td align="left"><p>0.020**</p></td></tr><tr><td align="left" /><td align="left"><p>(0.010)</p></td><td align="left"><p>(0.010)</p></td></tr><tr><td align="left"><p>Math anxiety index</p></td><td align="left"><p>0.018***</p></td><td align="left"><p>0.018***</p></td></tr><tr><td align="left" /><td align="left"><p>(0.005)</p></td><td align="left"><p>(0.005)</p></td></tr><tr><td align="left"><p>Math intrinsic motivation index</p></td><td align="left"><p>0.001</p></td><td align="left"><p>0.001</p></td></tr><tr><td align="left" /><td align="left"><p>(0.002)</p></td><td align="left"><p>(0.002)</p></td></tr><tr><td align="left" colspan="3"><p>Parent characteristics</p></td></tr><tr><td align="left"><p>Parental education (1 = at least one high school graduate)</p></td><td align="left"><p>− 0.005</p></td><td align="left"><p>− 0.002</p></td></tr><tr><td align="left" /><td align="left"><p>(0.004)</p></td><td align="left"><p>(0.005)</p></td></tr><tr><td align="left"><p>Mother out-migrated for work</p></td><td align="left"><p>− 0.001</p></td><td align="left"><p>− 0.002</p></td></tr><tr><td align="left" /><td align="left"><p>(0.001)</p></td><td align="left"><p>(0.001)</p></td></tr><tr><td align="left"><p>Father out-migrated for work</p></td><td align="left"><p>0.001</p></td><td align="left"><p>0.000</p></td></tr><tr><td align="left" /><td align="left"><p>(0.002)</p></td><td align="left"><p>(0.002)</p></td></tr><tr><td align="left"><p>Asset index (PCA score)</p></td><td align="left"><p>0.037***</p></td><td align="left"><p>0.035***</p></td></tr><tr><td align="left" /><td align="left"><p>(0.008)</p></td><td align="left"><p>(0.007)</p></td></tr><tr><td align="left" colspan="3"><p>Teacher characteristics</p></td></tr><tr><td align="left"><p>Teacher age (year)</p></td><td align="left"><p>0.000</p></td><td align="left"><p>− 0.000</p></td></tr><tr><td align="left" /><td align="left"><p>(0.004)</p></td><td align="left"><p>(0.004)</p></td></tr><tr><td align="left"><p>Teacher gender (1 = female)</p></td><td align="left"><p>0.007</p></td><td align="left"><p>0.008</p></td></tr><tr><td align="left" /><td align="left"><p>(0.006)</p></td><td align="left"><p>(0.007)</p></td></tr><tr><td align="left"><p>Associate education (1 = yes)</p></td><td align="left"><p>0.004</p></td><td align="left"><p>0.003</p></td></tr><tr><td align="left" /><td align="left"><p>(0.004)</p></td><td align="left"><p>(0.004)</p></td></tr><tr><td align="left"><p>Tenure (year)</p></td><td align="left"><p>0.002</p></td><td align="left"><p>0.002</p></td></tr><tr><td align="left" /><td align="left"><p>(0.005)</p></td><td align="left"><p>(0.005)</p></td></tr><tr><td align="left"><p>School fixed effect</p></td><td align="left" /><td align="left"><p>− 3.186**</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(1.244)</p></td></tr><tr><td align="left"><p>County fixed effect</p></td><td align="left" /><td align="left"><p>3.154**</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(1.249)</p></td></tr><tr><td align="left"><p>Observations</p></td><td align="left"><p>16,747</p></td><td align="left"><p>16,747</p></td></tr></tbody></table> </ephtml> </p> <p> <emph> Notes</emph>: Standard errors in parentheses; *** <emph>p</emph> < 0.01, ** <emph>p</emph> < 0.05, * <emph>p</emph> < 0.1. Standard errors (in parentheses) account for clustering at the school level</p> <p> <emph>Source</emph>: authors' survey.</p> <p></p> <p>We also present the multivariate correlations between demographic characteristics and math performance in the ethnic minority sample, rural southwest sample, and rural central sample, respectively (Appendix Tables 5, 6, and 7). In the ethnic minority sample, math affinity is positively and significantly correlated with math performance in 5th and 6th grade, with a coefficient of 0.21 (<emph>p</emph> < 0.01); meanwhile, math anxiety is negatively and significantly correlated with math performance in the 4th- and 5th-grade subsamples, with a coefficient of -0.18 (<emph>p</emph> < 0.01) in the overall sample. In the rural southwest sample, math affinity is positively and significantly correlated with math performance in all grades, with a coefficient of 0.20 (<emph>p</emph> < 0.01); meanwhile, math anxiety is negatively and significantly correlated with math performance in all grades, with a coefficient of -0.30 (<emph>p</emph> < 0.01) in the overall sample. Math intrinsic motivation is positively and significantly correlated with math performance in 6th grade, with a coefficient of 0.05 (<emph>p</emph> < 0.01) in the overall sample. In rural central sample, math affinity is positively and significantly correlated with math performance in all grades, with a coefficient of 0.16 (<emph>p</emph> < 0.01); meanwhile, math anxiety is negatively and significantly correlated with math performance in all grades, with a coefficient of -0.25 (<emph>p</emph> < 0.01); math intrinsic motivation is positively and significantly correlated with math performance in all grades, with a coefficient of 0.07 (<emph>p</emph> < 0.01) in the overall sample.</p> <hd id="AN0183282411-17">Decomposing the Han-minority achievement gap</hd> <p>To examine the proportion of Han-minority achievement gap explained by the student, household, and teacher characteristics, we used decomposition analysis to decompose the mean achievement gap between ethnic minority and Han students in our sample into an explained component and an unexplained component (Table 4). The explained component includes the components of the achievement gap that can be attributed to measured characteristics and is further decomposed into contributions from each measured student, parent, and teacher characteristic. Column 1 presents the results without accounting for school or county fixed effects, and column 2 accounts for both school and county fixed effects. Both columns show similar results.</p> <p>The analysis shows that being a Han resident is associated with a 0.257 (<emph>p</emph> < 0.01) standard deviation (SD) higher math score than being a minority resident. The explained component and the unexplained component of this achievement gap are both statistically significant. Column 1, which does not include school or county fixed effects, indicates that 37.35% (0.096 /0.257) of the gap can be explained by differences in student, parent, and teacher characteristics, while the remaining 62.65% (0.161/0.257) is unexplained. Specifically, the student characteristics explained 14.79% (0.038/0.257) of the gap, and parent characteristics explained 14.40% (0.037/0.257) of the gap. In column 2, which includes both school and county fixed effects, 24.51% (0.063/0.257) of the observed gap can be attributed to variations in student, parent, and teacher characteristics. Specifically, the student characteristics explained 13.62% (0.035/0.257) of the gap, and parent characteristics explained 13.62% (0.035/0.257) of the gap.</p> <p>In addition, the decomposition analysis also identified key specific variables accounting for Han-minority achievement gap. In both columns, the explained components of the achievement gap were composed of a consistent and comparable set of outcomes. The largest contribution came from household asset index (<emph>p</emph> < 0.01), which explained between 13.62% (0.035/0.257) and 14.40% (0.037/0.257) of the gap. The next largest contribution was from math affinity Z-score (<emph>p</emph> < 0.05), which explained 7.78% (0.020/0.257) of the gap. Math anxiety (<emph>p</emph> < 0.01) explained 7% (0.018/0.257) of the gap.</p> <hd id="AN0183282411-18">Discussion</hd> <p>The overall goal of this study is to examine and explain the academic achievement gap between Han and minority students in primary schools located in the central and southwestern regions of China. To achieve this, we conducted a survey encompassing 235 schools in the two regions, involving a total of 16,747 students across three grades. Our analysis revealed substantial disparities in academic performance between Han and minority students. On average, ethnic minority students scored 0.25 standard deviations lower in math compared to their rural central and rural southwest counterparts.</p> <p>Theoretically, we are among the first works to suggest and test that the achievement gap between socio-culturally different groups can be explained by cognitive and emotional factors. By integrating a two-factor model of math attitude and math anxiety, our findings showed that math affinity and math anxiety can contribute to explaining differences in Han-minority achievement gaps. Second, considering this finding, we believe that integrating conceptualizations based on the math affinity and math anxiety into the students' education can provide new directions for intervention research.</p> <p>We examined various student, household, and teacher characteristics, along with county and school fixed effects, to understand the factors associated with math scores. The results revealed that math affinity Z-score, and math anxiety index were all significantly correlated with student math performance outcomes among ethnic minority and Han students. Moreover, the decomposition analysis revealed that up to 15% of the achievement gap could be explained by variations in math affinity Z-score and math anxiety index.</p> <p>Building upon previous research on the Han-minority academic achievement gap (Wang et al., [<reflink idref="bib62" id="ref69">62</reflink>]; Yang et al., [<reflink idref="bib71" id="ref70">71</reflink>]), our findings provide further evidence that ethnic minority students in China continue to exhibit significantly lower academic achievement on average compared to Han students. While previous studies have primarily focused on social factors such as socioeconomic status, educational resources, parental involvement, and emotional factors (Assari et al., [<reflink idref="bib3" id="ref71">3</reflink>]; Azmitia et al., [<reflink idref="bib4" id="ref72">4</reflink>]; Brandmiller et al., [<reflink idref="bib9" id="ref73">9</reflink>]; Ceballo et al., [<reflink idref="bib12" id="ref74">12</reflink>]; Halpern-Manners et al., [<reflink idref="bib26" id="ref75">26</reflink>]; Samuelsson, [<reflink idref="bib53" id="ref76">53</reflink>]; Sonnenschein & Sun, [<reflink idref="bib55" id="ref77">55</reflink>]), our study contributes to the existing knowledge on the Han-minority achievement gap by confirming the crucial role of student math attitudes and math anxiety in explaining this disparity. Consistent with prior research (Connell et al., [<reflink idref="bib16" id="ref78">16</reflink>]; Eccles et al., [<reflink idref="bib19" id="ref79">19</reflink>]; Fuligni, [<reflink idref="bib23" id="ref80">23</reflink>]; Guay et al., [<reflink idref="bib25" id="ref81">25</reflink>]; Valentine et al., [<reflink idref="bib60" id="ref82">60</reflink>]), our findings underscore the significant relationship between math affinity, math anxiety and student's math achievement. Previous studies have shown that students with a higher affinity for math demonstrated greater engagement, improved learning, and higher performance, particularly in mathematics and science (Otis et al., [<reflink idref="bib43" id="ref83">43</reflink>]; Taylor et al., [<reflink idref="bib59" id="ref84">59</reflink>]). Correspondingly, we find that lower math affinity scores among ethnic minority students contribute to the Han-minority math achievement gap. Furthermore, our study contributes to the empirical evidence for the significance of anxiety in mathematics learning, which aligns with prior research highlighting its role as a mediating factor influencing academic performance (Bandura, [<reflink idref="bib5" id="ref85">5</reflink>]). Since lower levels of anxiety are associated with increased self-confidence, leading to enhanced self-regulation and improved academic outcomes (Hattie & Timperley, [<reflink idref="bib29" id="ref86">29</reflink>]), the high levels of math anxiety we observed likely contributed to the Han-minority achievement gap. These insights highlight the importance of addressing low math affinity and mitigating anxiety in mathematics education to narrow the achievement gap between Han and minority students.</p> <p>Our research provides a novel perspective, as previous investigations on the Han-Minority Achievement gap did not specifically focus on student's math attitudes and math anxiety. The utilization of decomposition analysis in our study adds further depth and nuance to the understanding of the relationship between math attitudes, math anxiety, and achievement. Our findings call for increasing attention to the cognitive and emotional causes of the Han-minority achievement gap, especially math affinity and math anxiety. The Stereotype Threat Theory suggested that math attitude and math anxiety serve as the mediate variable for the effects of stereotype threat on math performance, which suggested that math affinity and anxiety may be the critical risk factors contributing to the Han-minority achievement gap. Practitioners and educators should prioritize strategies that support the development of math affinity and reduce math anxiety among ethnic minority students to narrow the academic achievement gap. Incorporating classroom activities that foster math affinity and reduce math anxiety is essential (Grothérus et al., [<reflink idref="bib24" id="ref87">24</reflink>]; Otis et al., [<reflink idref="bib43" id="ref88">43</reflink>]; Wang et al., [<reflink idref="bib62" id="ref89">62</reflink>]). This can be accomplished through the creation of a learning environment that recognizes and celebrates student achievements (Person, [<reflink idref="bib46" id="ref90">46</reflink>]), cultivates a sense of safety and belonging (Akey, [<reflink idref="bib1" id="ref91">1</reflink>]), presents appropriate challenges that stimulate growth (Phillips & Lindsay, [<reflink idref="bib47" id="ref92">47</reflink>]), facilitates classroom discussions and supplementary activities (Okolo et al., [<reflink idref="bib42" id="ref93">42</reflink>]), establishes connections between lesson content and real-world applications (Molnar & Garcia, [<reflink idref="bib39" id="ref94">39</reflink>]), encourages collaborative work among students (Bartholomew, [<reflink idref="bib7" id="ref95">7</reflink>]), and provides opportunities for students to exercise autonomy and make choices (Bartholomew, [<reflink idref="bib7" id="ref96">7</reflink>]).</p> <p>While this study offers important insights into China's Han-minority achievement gap, several limitations warrant mention. First, the study relied on self-reported measures of students' math attitudes and math anxiety, which are subject to potential biases and inaccuracies. Second, while efforts were made to control for various student and household characteristics, there may still be unobserved factors that could influence the relationship between math attitudes, math anxiety and mathematics achievement, such as the language used by the students at home. Because the language of math instruction is Mandarin Chinese in schools, a different language spoken at home could negatively affect students' ability to engage with math classrooms. And since some, but not all, ethnic minority families speak Mandarin Chinese at home, the language spoken at home could be a factor that disproportionately affects ethnic minority students. Future research should address these limitations to further advance our understanding of the role of math attitudes and math anxiety in academic outcomes.</p> <p>Nevertheless, this paper offers important evidence that the Han-minority achievement gap is significant, persistent, and largely explainable by student math attitudes. Our findings can be used to guide future efforts to improve student math attitudes, reduce math anxiety, and promote positive engagement with mathematics as a way to closing the achievement gap and fostering equitable educational outcomes, both in rural China and abroad.</p> <hd id="AN0183282411-19">Declarations</hd> <p></p> <hd id="AN0183282411-20">Conflict of interest</hd> <p>None.</p> <hd id="AN0183282411-21">Appendix</hd> <p>See Figs. 1, 2, 3, 4 and Appendix Tables A1, A2, A3, A4, A5, A6, A7.</p> <p>Graph: Fig. 1 Kernel density estimate of standardized score for math performance among grade 4 students</p> <p>Graph: Fig. 2 Kernel density estimate of standardized score for math performance among grade 5 students</p> <p>Graph: Fig. 3 Kernel density estimate of standardized score for math performance among grade 6 students</p> <p>Graph: Fig. 4 Kernel density estimate of standardized score for math performance among students</p> <p>Appendix Table A1 HLM Multivariate correlations between demographic characteristics, ethnic minority, and math performance</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="3" colspan="2"><p>Variables</p></th><th align="left" colspan="16"><p>Math performance</p></th></tr><tr><th align="left" colspan="4"><p>Grade 4</p></th><th align="left" colspan="4"><p>Grade 5</p></th><th align="left" colspan="4"><p>Grade 6</p></th><th align="left" colspan="4"><p>All sample</p></th></tr><tr><th align="left"><p>(1)</p></th><th align="left"><p>(2)</p></th><th align="left"><p>(3)</p></th><th align="left"><p>(4)</p></th><th align="left"><p>(5)</p></th><th align="left"><p>(6)</p></th><th align="left"><p>(7)</p></th><th align="left"><p>(8)</p></th><th align="left"><p>(9)</p></th><th align="left"><p>(10)</p></th><th align="left"><p>(11)</p></th><th align="left"><p>(12)</p></th><th align="left"><p>(13)</p></th><th align="left"><p>(14)</p></th><th align="left"><p>(15)</p></th><th align="left"><p>(16)</p></th></tr><tr><th align="left"><p>(1)</p></th><th align="left"><p>Ethnic minority(1 = yes)</p></th><th align="left"><p>-0.11</p></th><th align="left"><p>0.01</p></th><th align="left"><p>0.02</p></th><th align="left"><p>0.00</p></th><th align="left"><p>− 0.27***</p></th><th align="left"><p>− 0.15**</p></th><th align="left"><p>− 0.14**</p></th><th align="left"><p>− 0.16**</p></th><th align="left"><p>− 0.37***</p></th><th align="left"><p>− 0.12*</p></th><th align="left"><p>− 0.12*</p></th><th align="left"><p>− 0.15**</p></th><th align="left"><p>− 0.26***</p></th><th align="left"><p>− 0.17***</p></th><th align="left"><p>− 0.16***</p></th><th align="left"><p>− 0.17***</p></th></tr></thead><tbody><tr><td align="left" /><td align="left" /><td align="left"><p>(0.10)</p></td><td align="left"><p>(0.11)</p></td><td align="left"><p>(0.10)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.06)</p></td><td align="left"><p>(0.06)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.06)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.04)</p></td></tr><tr><td align="left"><p>(2)</p></td><td align="left"><p>Personal characteristics</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(3)</p></td><td align="left"><p>Household characteristic</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(4)</p></td><td align="left"><p>Teacher characteristics</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(5)</p></td><td align="left"><p>School random effect</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(6)</p></td><td align="left"><p>County random effect</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>No</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(7)</p></td><td align="left"><p>Observations</p></td><td align="left"><p>5,118</p></td><td align="left"><p>5,118</p></td><td align="left"><p>5,118</p></td><td align="left"><p>5,118</p></td><td align="left"><p>5,870</p></td><td align="left"><p>5,870</p></td><td align="left"><p>5,870</p></td><td align="left"><p>5,870</p></td><td align="left"><p>5,759</p></td><td align="left"><p>5,759</p></td><td align="left"><p>5,759</p></td><td align="left"><p>5,759</p></td><td align="left"><p>16,747</p></td><td align="left"><p>16,747</p></td><td align="left"><p>16,747</p></td><td align="left"><p>16,747</p></td></tr></tbody></table> </ephtml> </p> <p> <emph>Notes</emph>: Standard errors in parentheses account for clustering at the school level; ***<emph> p</emph> < 0.01, ** <emph>p</emph> < 0.05, *<emph> p</emph> < 0.1. Personal characteristics include the student's gender, age, whether they were an only child, and whether they were boarding students, math affinity, math intrinsic motivation, and math anxiety. Household characteristics include whether their parents migrated to another city for work during most of the previous semester, the highest levels of schooling their parents had attained, and household asset. Teacher characteristics include teacher gender, age, education level, and teaching experience in years</p> <p> <emph>Source</emph>: authors' survey</p> <p></p> <p>Appendix Table A2 The Multivariate correlations between ethnic minority and math achievement after PSM (N = 2,224)</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left"><p>Dependent Variable</p></th><th align="left"><p>Coefficient</p></th><th align="left"><p>Standard Errors (SE)</p></th><th align="left"><p><italic>t</italic>-Value</p></th></tr></thead><tbody><tr><td align="left"><p>Math achievement</p></td><td align="left"><p>− 0.168<sup>***</sup></p></td><td align="left"><p>0.050</p></td><td char="." align="char"><p>3.349</p></td></tr><tr><td align="left"><p>Observations</p></td><td align="left"><p>2,224</p></td><td align="left"><p>2,224</p></td><td char="." align="char" /></tr></tbody></table> </ephtml> </p> <p> <emph>Note</emph>: *** <emph>p</emph> < 0.01, ** <emph>p</emph> < 0.05, *<emph> p</emph> < 0.1. The regression equation controlled the personal characteristics of students, household, and teachers</p> <p>Source: Authors' survey</p> <p></p> <p>Appendix Table A3 Tests of covariate matching quality</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="2"><p>Variable</p></th><th align="left"><p>Unmatched</p></th><th align="left" colspan="2"><p>Mean</p></th><th align="left" rowspan="2"><p>%bias</p></th><th align="left" rowspan="2"><p>%reduct |bias|</p></th><th align="left" colspan="2"><p>t-test</p></th></tr><tr><th align="left"><p>Matched</p></th><th align="left"><p>Treated</p></th><th align="left"><p>Control</p></th><th align="left"><p>t</p></th><th align="left"><p>p >|t|</p></th></tr></thead><tbody><tr><td align="left"><p>Age (years)</p></td><td align="left"><p>U</p></td><td align="left"><p>11.84</p></td><td align="left"><p>11.12</p></td><td align="left"><p>63.2</p></td><td char="." align="char" /><td align="left"><p>25.27</p></td><td align="left"><p> < 0.001</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>11.81</p></td><td align="left"><p>11.78</p></td><td align="left"><p>2.4</p></td><td char="." align="char"><p>96.2</p></td><td align="left"><p>0.61</p></td><td align="left"><p>0.543</p></td></tr><tr><td align="left"><p>Gender (1 = girl)</p></td><td align="left"><p>U</p></td><td align="left"><p>0.53</p></td><td align="left"><p>0.48</p></td><td align="left"><p>9.1</p></td><td char="." align="char" /><td align="left"><p>3.16</p></td><td align="left"><p>0.002</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>0.53</p></td><td align="left"><p>0.52</p></td><td align="left"><p>2.5</p></td><td char="." align="char"><p>72.6</p></td><td align="left"><p>0.63</p></td><td align="left"><p>0.527</p></td></tr><tr><td align="left"><p>Only child (1 = yes)</p></td><td align="left"><p>U</p></td><td align="left"><p>0.08</p></td><td align="left"><p>0.11</p></td><td align="left"><p>− 10.2</p></td><td char="." align="char" /><td align="left"><p>− 3.35</p></td><td align="left"><p>0.001</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>0.08</p></td><td align="left"><p>0.10</p></td><td align="left"><p>− 3.9</p></td><td char="." align="char"><p>61.5</p></td><td align="left"><p>− 1.03</p></td><td align="left"><p>0.301</p></td></tr><tr><td align="left"><p>Boarding at school (1 = yes)</p></td><td align="left"><p>U</p></td><td align="left"><p>0.33</p></td><td align="left"><p>0.16</p></td><td align="left"><p>39.3</p></td><td char="." align="char" /><td align="left"><p>15.21</p></td><td align="left"><p>< 0.001</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>0.32</p></td><td align="left"><p>0.31</p></td><td align="left"><p>1.7</p></td><td char="." align="char"><p>95.8</p></td><td align="left"><p>0.38</p></td><td align="left"><p>0.702</p></td></tr><tr><td align="left"><p>Like math Z score</p></td><td align="left"><p>U</p></td><td align="left"><p>− 0.09</p></td><td align="left"><p>0.02</p></td><td align="left"><p>− 10.4</p></td><td char="." align="char" /><td align="left"><p>− 3.71</p></td><td align="left"><p>< 0.001</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>− 0.09</p></td><td align="left"><p>− 0.07</p></td><td align="left"><p>− 1.9</p></td><td char="." align="char"><p>81.5</p></td><td align="left"><p>− .48</p></td><td align="left"><p>0.634</p></td></tr><tr><td align="left"><p>Math anxiety index</p></td><td align="left"><p>U</p></td><td align="left"><p>0.05</p></td><td align="left"><p>− 0.02</p></td><td align="left"><p>10.2</p></td><td char="." align="char" /><td align="left"><p>3.48</p></td><td align="left"><p>0.001</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>0.05</p></td><td align="left"><p>0.09</p></td><td align="left"><p>− 5.7</p></td><td char="." align="char"><p>44.4</p></td><td align="left"><p>− 1.41</p></td><td align="left"><p>0.158</p></td></tr><tr><td align="left"><p>Math intrinsic motivation index</p></td><td align="left"><p>U</p></td><td align="left"><p>− 0.01</p></td><td align="left"><p>0.02</p></td><td align="left"><p>− 2.8</p></td><td char="." align="char" /><td align="left"><p>− .96</p></td><td align="left"><p>0.336</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>− 0.01</p></td><td align="left"><p>− 0.03</p></td><td align="left"><p>2.5</p></td><td char="." align="char"><p>8.2</p></td><td align="left"><p>0.62</p></td><td align="left"><p>0.533</p></td></tr><tr><td align="left"><p>Parental education (1 = at least one high school graduate)</p></td><td align="left"><p>U</p></td><td align="left"><p>0.11</p></td><td align="left"><p>0.38</p></td><td align="left"><p>− 66.4</p></td><td char="." align="char" /><td align="left"><p>− 19.82</p></td><td align="left"><p>< 0.001</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>0.11</p></td><td align="left"><p>0.14</p></td><td align="left"><p>− 6.1</p></td><td char="." align="char"><p>90.8</p></td><td align="left"><p>− 1.91</p></td><td align="left"><p>0.056</p></td></tr><tr><td align="left"><p>Mother out-migrated for work</p></td><td align="left"><p>U</p></td><td align="left"><p>0.20</p></td><td align="left"><p>0.23</p></td><td align="left"><p>− 8.2</p></td><td char="." align="char" /><td align="left"><p>− 2.78</p></td><td align="left"><p>0.005</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>0.20</p></td><td align="left"><p>0.19</p></td><td align="left"><p>3</p></td><td char="." align="char"><p>63.1</p></td><td align="left"><p>0.8</p></td><td align="left"><p>0.425</p></td></tr><tr><td align="left"><p>Father out-migrated for work</p></td><td align="left"><p>U</p></td><td align="left"><p>0.38</p></td><td align="left"><p>0.48</p></td><td align="left"><p>− 20.1</p></td><td char="." align="char" /><td align="left"><p>− 6.88</p></td><td align="left"><p>< 0.001</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>0.38</p></td><td align="left"><p>0.37</p></td><td align="left"><p>1.9</p></td><td char="." align="char"><p>90.5</p></td><td align="left"><p>0.49</p></td><td align="left"><p>0.624</p></td></tr><tr><td align="left"><p>Asset index (PCA score)</p></td><td align="left"><p>U</p></td><td align="left"><p>− 0.77</p></td><td align="left"><p>0.10</p></td><td align="left"><p>− 71.4</p></td><td char="." align="char" /><td align="left"><p>− 24.16</p></td><td align="left"><p>< 0.001</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>− 0.75</p></td><td align="left"><p>− 0.76</p></td><td align="left"><p>1</p></td><td char="." align="char"><p>98.6</p></td><td align="left"><p>0.25</p></td><td align="left"><p>0.803</p></td></tr><tr><td align="left"><p>Teacher age (year)</p></td><td align="left"><p>U</p></td><td align="left"><p>39.08</p></td><td align="left"><p>39.88</p></td><td align="left"><p>− 7.6</p></td><td char="." align="char" /><td align="left"><p>− 2.81</p></td><td align="left"><p>0.005</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>38.96</p></td><td align="left"><p>38.77</p></td><td align="left"><p>1.8</p></td><td char="." align="char"><p>76.2</p></td><td align="left"><p>0.43</p></td><td align="left"><p>0.664</p></td></tr><tr><td align="left"><p>Teacher gender (1 = female)</p></td><td align="left"><p>U</p></td><td align="left"><p>0.53</p></td><td align="left"><p>0.61</p></td><td align="left"><p>− 16.5</p></td><td char="." align="char" /><td align="left"><p>− 5.77</p></td><td align="left"><p>< 0.001</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>0.53</p></td><td align="left"><p>0.52</p></td><td align="left"><p>1.7</p></td><td char="." align="char"><p>89.5</p></td><td align="left"><p>0.44</p></td><td align="left"><p>0.663</p></td></tr><tr><td align="left"><p>Associate education (1 = yes)</p></td><td align="left"><p>U</p></td><td align="left"><p>0.63</p></td><td align="left"><p>0.55</p></td><td align="left"><p>16.7</p></td><td char="." align="char" /><td align="left"><p>5.69</p></td><td align="left"><p>< 0.001</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>0.63</p></td><td align="left"><p>0.63</p></td><td align="left"><p>0.6</p></td><td char="." align="char"><p>96.2</p></td><td align="left"><p>0.16</p></td><td align="left"><p>0.87</p></td></tr><tr><td align="left"><p>Tenure (year)</p></td><td align="left"><p>U</p></td><td align="left"><p>17.44</p></td><td align="left"><p>18.12</p></td><td align="left"><p>− 5.4</p></td><td char="." align="char" /><td align="left"><p>− 1.97</p></td><td align="left"><p>0.049</p></td></tr><tr><td align="left" /><td align="left"><p>M</p></td><td align="left"><p>17.30</p></td><td align="left"><p>17.31</p></td><td align="left"><p>0</p></td><td char="." align="char"><p>99.4</p></td><td align="left"><p>− 0.01</p></td><td align="left"><p>0.993</p></td></tr></tbody></table> </ephtml> </p> <p>Appendix Table A4 Sensitivity analysis with ROSENBAUM'S bounds</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" colspan="2"><p>Gamma</p></th><th align="left"><p>sig + </p></th><th align="left"><p>sig-</p></th><th align="left"><p>t-hat + </p></th><th align="left"><p>t-hat-</p></th><th align="left"><p>CI + </p></th><th align="left"><p>CI-</p></th></tr></thead><tbody><tr><td align="left" colspan="2"><p>1</p></td><td align="left"><p>0</p></td><td align="left"><p>0</p></td><td align="left"><p>25</p></td><td align="left"><p>25</p></td><td align="left"><p>25</p></td><td align="left"><p>25.5</p></td></tr><tr><td align="left" colspan="2"><p>1.25</p></td><td align="left"><p>0</p></td><td align="left"><p>0</p></td><td align="left"><p>24.5</p></td><td align="left"><p>25.5</p></td><td align="left"><p>24.5</p></td><td align="left"><p>26</p></td></tr><tr><td align="left" colspan="2"><p>1.5</p></td><td align="left"><p>0</p></td><td align="left"><p>0</p></td><td align="left"><p>24</p></td><td align="left"><p>26</p></td><td align="left"><p>24</p></td><td align="left"><p>26.5</p></td></tr><tr><td align="left" colspan="2"><p>1.75</p></td><td align="left"><p>0</p></td><td align="left"><p>0</p></td><td align="left"><p>23.5</p></td><td align="left"><p>26.5</p></td><td align="left"><p>23.5</p></td><td align="left"><p>26.5</p></td></tr><tr><td align="left" colspan="2"><p>2</p></td><td align="left"><p>0</p></td><td align="left"><p>0</p></td><td align="left"><p>23.5</p></td><td align="left"><p>27</p></td><td align="left"><p>23</p></td><td align="left"><p>27</p></td></tr></tbody></table> </ephtml> </p> <p>* gamma - log odds of differential assignment due to unobserved factors</p> <p>sig +  - upper bound significance level</p> <p>sig- - lower bound significance level</p> <p>t-hat +  - upper bound Hodges-Lehmann point estimate</p> <p>t-hat- - lower bound Hodges-Lehmann point estimate</p> <p>CI +  upper bound confidence interval (a = 0.95)</p> <p>CI- - lower bound confidence interval (a = 0.95)</p> <p>Appendix Table A5 Multivariate correlations between demographic characteristics and math performance in ethnic minority samples</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="3" colspan="2"><p>Variables</p></th><th align="left" colspan="4"><p>Math performance</p></th></tr><tr><th align="left"><p>Grade 4</p></th><th align="left"><p>Grade 5</p></th><th align="left"><p>Grade 6</p></th><th align="left"><p>All sample</p></th></tr><tr><th align="left"><p>(1)</p></th><th align="left"><p>(2)</p></th><th align="left"><p>(3)</p></th><th align="left"><p>(4)</p></th></tr></thead><tbody><tr><td align="left"><p>(1)</p></td><td align="left"><p>Age (years)</p></td><td align="left"><p>0.04</p></td><td align="left"><p>− 0.01</p></td><td align="left"><p>− 0.12**</p></td><td align="left"><p>− 0.04</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.03)</p></td></tr><tr><td align="left"><p>(2)</p></td><td align="left"><p>Gender (1 = girl)</p></td><td align="left"><p>− 0.11</p></td><td align="left"><p>− 0.25**</p></td><td align="left"><p>0.06</p></td><td align="left"><p>− 0.13***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.10)</p></td><td align="left"><p>(0.11)</p></td><td align="left"><p>(0.09)</p></td><td align="left"><p>(0.05)</p></td></tr><tr><td align="left"><p>(3)</p></td><td align="left"><p>Only child (1 = yes)</p></td><td align="left"><p>0.11</p></td><td align="left"><p>0.06</p></td><td align="left"><p>− 0.16</p></td><td align="left"><p>− 0.00</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.23)</p></td><td align="left"><p>(0.17)</p></td><td align="left"><p>(0.26)</p></td><td align="left"><p>(0.11)</p></td></tr><tr><td align="left"><p>(4)</p></td><td align="left"><p>Boarding at school (1 = yes)</p></td><td align="left"><p>0.32***</p></td><td align="left"><p>0.01</p></td><td align="left"><p>0.12</p></td><td align="left"><p>0.19***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.09)</p></td><td align="left"><p>(0.14)</p></td><td align="left"><p>(0.14)</p></td><td align="left"><p>(0.07)</p></td></tr><tr><td align="left"><p>(5)</p></td><td align="left"><p>Math affinity Z-score</p></td><td align="left"><p>0.11*</p></td><td align="left"><p>0.30***</p></td><td align="left"><p>0.31***</p></td><td align="left"><p>0.21***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.09)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.03)</p></td></tr><tr><td align="left"><p>(6)</p></td><td align="left"><p>Math anxiety index</p></td><td align="left"><p>− 0.18**</p></td><td align="left"><p>− 0.31***</p></td><td align="left"><p>− 0.11</p></td><td align="left"><p>− 0.18***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.09)</p></td><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.10)</p></td><td align="left"><p>(0.05)</p></td></tr><tr><td align="left"><p>(7)</p></td><td align="left"><p>Math intrinsic motivation index</p></td><td align="left"><p>0.05</p></td><td align="left"><p>− 0.01</p></td><td align="left"><p>− 0.06</p></td><td align="left"><p>− 0.01</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.07)</p></td><td align="left"><p>(0.10)</p></td><td align="left"><p>(0.09)</p></td><td align="left"><p>(0.04)</p></td></tr><tr><td align="left"><p>(8)</p></td><td align="left"><p>Parental education (1 = at least one high school graduate)</p></td><td align="left"><p>0.10</p></td><td align="left"><p>0.04</p></td><td align="left"><p>− 0.01</p></td><td align="left"><p>0.04</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.19)</p></td><td align="left"><p>(0.21)</p></td><td align="left"><p>(0.23)</p></td><td align="left"><p>(0.10)</p></td></tr><tr><td align="left"><p>(9)</p></td><td align="left"><p>Mother out-migrated for work</p></td><td align="left"><p>− 0.05</p></td><td align="left"><p>0.02</p></td><td align="left"><p>0.07</p></td><td align="left"><p>− 0.03</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.11)</p></td><td align="left"><p>(0.15)</p></td><td align="left"><p>(0.12)</p></td><td align="left"><p>(0.08)</p></td></tr><tr><td align="left"><p>(10)</p></td><td align="left"><p>Father out-migrated for work</p></td><td align="left"><p>0.04</p></td><td align="left"><p>− 0.19</p></td><td align="left"><p>0.04</p></td><td align="left"><p>− 0.05</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.12)</p></td><td align="left"><p>(0.13)</p></td><td align="left"><p>(0.14)</p></td><td align="left"><p>(0.07)</p></td></tr><tr><td align="left"><p>(11)</p></td><td align="left"><p>Asset index (PCA score)</p></td><td align="left"><p>− 0.09</p></td><td align="left"><p>− 0.02</p></td><td align="left"><p>− 0.15***</p></td><td align="left"><p>− 0.06**</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.06)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.03)</p></td></tr><tr><td align="left"><p>(12)</p></td><td align="left"><p>Teacher age (year)</p></td><td align="left"><p>− 0.12***</p></td><td align="left"><p>4.05***</p></td><td align="left"><p>− 1.91***</p></td><td align="left"><p>− 0.01</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.62)</p></td><td align="left"><p>(0.20)</p></td><td align="left"><p>(0.03)</p></td></tr><tr><td align="left"><p>(13)</p></td><td align="left"><p>Teacher gender (1 = female)</p></td><td align="left"><p>− 1.12***</p></td><td align="left"><p>6.19***</p></td><td align="left"><p>− 7.65***</p></td><td align="left"><p>0.41***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.39)</p></td><td align="left"><p>(0.88)</p></td><td align="left"><p>(0.84)</p></td><td align="left"><p>(0.11)</p></td></tr><tr><td align="left"><p>(14)</p></td><td align="left"><p>Associate education (1 = yes)</p></td><td align="left"><p>0.28</p></td><td align="left"><p>− 4.75***</p></td><td align="left"><p>1.21***</p></td><td align="left"><p>− 0.00</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.25)</p></td><td align="left"><p>(0.91)</p></td><td align="left"><p>(0.29)</p></td><td align="left"><p>(0.19)</p></td></tr><tr><td align="left"><p>(15)</p></td><td align="left"><p>Tenure (year)</p></td><td align="left"><p>0.06*</p></td><td align="left"><p>− 3.03***</p></td><td align="left"><p>1.55***</p></td><td align="left"><p>0.01</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.47)</p></td><td align="left"><p>(0.16)</p></td><td align="left"><p>(0.− 02)</p></td></tr><tr><td align="left"><p>(16)</p></td><td align="left"><p>School fixed effect</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(17)</p></td><td align="left"><p>County fixed effect</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(18)</p></td><td align="left"><p>Observations</p></td><td align="left"><p>385</p></td><td align="left"><p>473</p></td><td align="left"><p>441</p></td><td align="left"><p>1,299</p></td></tr><tr><td align="left"><p>(19)</p></td><td align="left"><p>R-squared</p></td><td align="left"><p>0.36</p></td><td align="left"><p>0.29</p></td><td align="left"><p>0.33</p></td><td align="left"><p>0.21</p></td></tr></tbody></table> </ephtml> </p> <p> <emph>Notes</emph>: Standard errors in parentheses account for clustering at the school level; *** <emph>p</emph> < 0.01, **<emph> p</emph> < 0.05, *<emph> p</emph> < 0.1</p> <p> <emph>Source</emph>: authors' survey</p> <p></p> <p>Appendix Table A6 Multivariate correlations between demographic characteristics and math performance in rural southwest samples</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="3" colspan="2"><p>Variables</p></th><th align="left" colspan="4"><p>Math performance</p></th></tr><tr><th align="left"><p>Grade 4</p></th><th align="left"><p>Grade 5</p></th><th align="left"><p>Grade 6</p></th><th align="left"><p>All sample</p></th></tr><tr><th align="left"><p>(1)</p></th><th align="left"><p>(2)</p></th><th align="left"><p>(3)</p></th><th align="left"><p>(4)</p></th></tr></thead><tbody><tr><td align="left"><p>(1)</p></td><td align="left"><p>Age (years)</p></td><td align="left"><p>0.05</p></td><td align="left"><p>0.03</p></td><td align="left"><p>− 0.16***</p></td><td align="left"><p>0.02</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(2)</p></td><td align="left"><p>Gender (1 = girl)</p></td><td align="left"><p>− 0.16***</p></td><td align="left"><p>− 0.11**</p></td><td align="left"><p>− 0.16***</p></td><td align="left"><p>− 0.15***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.03)</p></td></tr><tr><td align="left"><p>(3)</p></td><td align="left"><p>Only child (1 = yes)</p></td><td align="left"><p>0.03</p></td><td align="left"><p>− 0.02</p></td><td align="left"><p>0.02</p></td><td align="left"><p>0.02</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.06)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(4)</p></td><td align="left"><p>Boarding at school (1 = yes)</p></td><td align="left"><p>0.17</p></td><td align="left"><p>0.15</p></td><td align="left"><p>− 0.01</p></td><td align="left"><p>0.11</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.13)</p></td><td align="left"><p>(0.14)</p></td><td align="left"><p>(0.10)</p></td><td align="left"><p>(0.07)</p></td></tr><tr><td align="left"><p>(5)</p></td><td align="left"><p>Math affinity Z score</p></td><td align="left"><p>0.20***</p></td><td align="left"><p>0.26***</p></td><td align="left"><p>0.17***</p></td><td align="left"><p>0.20***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(6)</p></td><td align="left"><p>Math anxiety index</p></td><td align="left"><p>− 0.27***</p></td><td align="left"><p>− 0.27***</p></td><td align="left"><p>− 0.33***</p></td><td align="left"><p>− 0.30***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(7)</p></td><td align="left"><p>Math intrinsic motivation index</p></td><td align="left"><p>0.04</p></td><td align="left"><p>0.05</p></td><td align="left"><p>0.10***</p></td><td align="left"><p>0.05***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(8)</p></td><td align="left"><p>Parental education (1 = at least one high school graduate)</p></td><td align="left"><p>0.12***</p></td><td align="left"><p>0.0− 6</p></td><td align="left"><p>0.06</p></td><td align="left"><p>0.07***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(9)</p></td><td align="left"><p>Mother out-migrated for work</p></td><td align="left"><p>− 0.04</p></td><td align="left"><p>− 0.05</p></td><td align="left"><p>− 0.03</p></td><td align="left"><p>− 0.04*</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.06)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(10)</p></td><td align="left"><p>Father out-migrated for work</p></td><td align="left"><p>0.04</p></td><td align="left"><p>0.02</p></td><td align="left"><p>0.02</p></td><td align="left"><p>0.02</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(11)</p></td><td align="left"><p>Asset index (PCA score)</p></td><td align="left"><p>0.01</p></td><td align="left"><p>0.02</p></td><td align="left"><p>0.03**</p></td><td align="left"><p>0.02**</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.01)</p></td><td align="left"><p>(0.02)</p></td><td align="left"><p>(0.01)</p></td><td align="left"><p>(0.01)</p></td></tr><tr><td align="left"><p>(12)</p></td><td align="left"><p>Teacher age (year)</p></td><td align="left"><p>− 0.01</p></td><td align="left"><p>− 0.25***</p></td><td align="left"><p>− 0.23***</p></td><td align="left"><p>− 0.01</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.02)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.01)</p></td></tr><tr><td align="left"><p>(13)</p></td><td align="left"><p>Teacher gender (1 = female)</p></td><td align="left"><p>0.07</p></td><td align="left"><p>− 0.75***</p></td><td align="left"><p>0.04</p></td><td align="left"><p>0.06</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.56)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.10)</p></td><td align="left"><p>(0.04)</p></td></tr><tr><td align="left"><p>(14)</p></td><td align="left"><p>Associate education (1 = yes)</p></td><td align="left"><p>− 0.43</p></td><td align="left"><p>0.64***</p></td><td align="left"><p>0.14***</p></td><td align="left"><p>0.02</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.83)</p></td><td align="left"><p>(0.17)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.− 06)</p></td></tr><tr><td align="left"><p>(15)</p></td><td align="left"><p>Tenure (year)</p></td><td align="left"><p>− 0.02</p></td><td align="left"><p>0.18***</p></td><td align="left"><p>0.17***</p></td><td align="left"><p>0.02*</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.01)</p></td><td align="left"><p>(0.01)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(.01)</p></td></tr><tr><td align="left"><p>(16)</p></td><td align="left"><p>School fixed effect</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(17)</p></td><td align="left"><p>County fixed effect</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(18)</p></td><td align="left"><p>Observations</p></td><td align="left"><p>2,088</p></td><td align="left"><p>2,293</p></td><td align="left"><p>2,493</p></td><td align="left"><p>6,874</p></td></tr><tr><td align="left"><p>(19)</p></td><td align="left"><p>R-squared</p></td><td align="left"><p>0.23</p></td><td align="left"><p>0.24</p></td><td align="left"><p>0.26</p></td><td align="left"><p>0.19</p></td></tr></tbody></table> </ephtml> </p> <p> <emph> Notes</emph>: Standard errors in parentheses account for clustering at the school level; *** <emph>p</emph> < 0.01, ** <emph>p</emph> < 0.05, * <emph>p</emph> < 0.1</p> <p> <emph>Source</emph>: authors' survey</p> <p></p> <p>Appendix Table A7 Multivariate correlations between demographic characteristics and math performance in rural central samples</p> <p> <ephtml> <table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="3" colspan="2"><p>Variables</p></th><th align="left" colspan="4"><p>Math performance</p></th></tr><tr><th align="left"><p>Grade 4</p></th><th align="left"><p>Grade 5</p></th><th align="left"><p>Grade 6</p></th><th align="left"><p>All sample</p></th></tr><tr><th align="left"><p>(1)</p></th><th align="left"><p>(2)</p></th><th align="left"><p>(3)</p></th><th align="left"><p>(4)</p></th></tr></thead><tbody><tr><td align="left"><p>(1)</p></td><td align="left"><p>Age (years)</p></td><td align="left"><p>− 0.12***</p></td><td align="left"><p>− 0.21***</p></td><td align="left"><p>− 0.24***</p></td><td align="left"><p>− 0.04**</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(2)</p></td><td align="left"><p>Gender (1 = girl)</p></td><td align="left"><p>0.06</p></td><td align="left"><p>0.06*</p></td><td align="left"><p>0.14***</p></td><td align="left"><p>0.10***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.05)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(3)</p></td><td align="left"><p>Only child (1 = yes)</p></td><td align="left"><p>− 0.27***</p></td><td align="left"><p>− 0.16*</p></td><td align="left"><p>− 0.33***</p></td><td align="left"><p>− 0.29***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.08)</p></td><td align="left"><p>(0.08)</p></td><td align="left"><p>(0.09)</p></td><td align="left"><p>(0.05)</p></td></tr><tr><td align="left"><p>(4)</p></td><td align="left"><p>Boarding at school (1 = yes)</p></td><td align="left"><p>0.00</p></td><td align="left"><p>0.03</p></td><td align="left"><p>0.07</p></td><td align="left"><p>0.07*</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.09)</p></td><td align="left"><p>(0.09)</p></td><td align="left"><p>(0.08)</p></td><td align="left"><p>(0.04)</p></td></tr><tr><td align="left"><p>(5)</p></td><td align="left"><p>Math affinity Z score</p></td><td align="left"><p>0.18***</p></td><td align="left"><p>0.19***</p></td><td align="left"><p>0.15***</p></td><td align="left"><p>0.16***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.02)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(6)</p></td><td align="left"><p>Math anxiety index</p></td><td align="left"><p>− 0.27***</p></td><td align="left"><p>− 0.22***</p></td><td align="left"><p>− 0.24***</p></td><td align="left"><p>− 0.25***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(7)</p></td><td align="left"><p>Math intrinsic motivation index</p></td><td align="left"><p>0.06**</p></td><td align="left"><p>0.11***</p></td><td align="left"><p>0.10***</p></td><td align="left"><p>0.07***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.02)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(8)</p></td><td align="left"><p>Parental education (1 = at least one high school graduate)</p></td><td align="left"><p>− 0.04</p></td><td align="left"><p>− 0.05*</p></td><td align="left"><p>− 0.05</p></td><td align="left"><p>− 0.06**</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(9)</p></td><td align="left"><p>Mother out-migrated for work</p></td><td align="left"><p>− 0.02</p></td><td align="left"><p>0.03</p></td><td align="left"><p>− 0.02</p></td><td align="left"><p>− 0.02</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.03)</p></td></tr><tr><td align="left"><p>(10)</p></td><td align="left"><p>Father out-migrated for work</p></td><td align="left"><p>− 0.01</p></td><td align="left"><p>− 0.02</p></td><td align="left"><p>0.01</p></td><td align="left"><p>0.00</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.02)</p></td></tr><tr><td align="left"><p>(11)</p></td><td align="left"><p>Asset index (PCA score)</p></td><td align="left"><p>0.06***</p></td><td align="left"><p>0.04**</p></td><td align="left"><p>− .00</p></td><td align="left"><p>0.04***</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.02)</p></td><td align="left"><p>(0.02)</p></td><td align="left"><p>(0.01)</p></td><td align="left"><p>(0.01)</p></td></tr><tr><td align="left"><p>(12)</p></td><td align="left"><p>Teacher age (year)</p></td><td align="left"><p>− 3.06***</p></td><td align="left"><p>0.12***</p></td><td align="left"><p>0.04***</p></td><td align="left"><p>0.00</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.15)</p></td><td align="left"><p>(0.01)</p></td><td align="left"><p>(0.01)</p></td><td align="left"><p>(0.01)</p></td></tr><tr><td align="left"><p>(13)</p></td><td align="left"><p>Teacher gender (1 = female)</p></td><td align="left"><p>− 4.34***</p></td><td align="left"><p>0.41***</p></td><td align="left"><p>0.51***</p></td><td align="left"><p>0.05</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.36)</p></td><td align="left"><p>(0.08)</p></td><td align="left"><p>(0.10)</p></td><td align="left"><p>(0.05)</p></td></tr><tr><td align="left"><p>(14)</p></td><td align="left"><p>Associate education (1 = yes)</p></td><td align="left"><p>4.87***</p></td><td align="left"><p>1.09***</p></td><td align="left"><p>− 0.17***</p></td><td align="left"><p>− 0.11**</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.22)</p></td><td align="left"><p>(0.12)</p></td><td align="left"><p>(0.04)</p></td><td align="left"><p>(0.05)</p></td></tr><tr><td align="left"><p>(15)</p></td><td align="left"><p>Tenure (year)</p></td><td align="left"><p>2.54***</p></td><td align="left"><p>− 0.27***</p></td><td align="left"><p>− 0.01***</p></td><td align="left"><p>− 0.00</p></td></tr><tr><td align="left" /><td align="left" /><td align="left"><p>(0.12)</p></td><td align="left"><p>(0.03)</p></td><td align="left"><p>(0.00)</p></td><td align="left"><p>(0.00)</p></td></tr><tr><td align="left"><p>(16)</p></td><td align="left"><p>School fixed effect</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(17)</p></td><td align="left"><p>County fixed effect</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td><td align="left"><p>Yes</p></td></tr><tr><td align="left"><p>(18)</p></td><td align="left"><p>Observations</p></td><td align="left"><p>2,645</p></td><td align="left"><p>3,104</p></td><td align="left"><p>2,825</p></td><td align="left"><p>8,574</p></td></tr><tr><td align="left"><p>(19)</p></td><td align="left"><p>R-squared</p></td><td align="left"><p>0.25</p></td><td align="left"><p>0.26</p></td><td align="left"><p>0.32</p></td><td align="left"><p>0.21</p></td></tr></tbody></table> </ephtml> </p> <p> <emph>Notes</emph>: Standard errors in parentheses account for clustering at the school level; ***<emph> p</emph> < 0.01, ** <emph>p</emph> < 0.05, * <emph>p</emph> < 0.1</p> <p> <emph>Source</emph>: authors' survey.</p> <p></p> <hd id="AN0183282411-22">Publisher's Note</hd> <p>Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.</p> <ref id="AN0183282411-23"> <title> References </title> <blist> <bibl id="bib1" idref="ref2" type="bt">1</bibl> <bibtext> Akey TM. Using positive student engagement to increase student achievement. Center for Comprehensive School Reform and Improvement Newsletter. 2007; 4: 1-14</bibtext> </blist> <blist> <bibl id="bib2" idref="ref3" type="bt">2</bibl> <bibtext> Anderson ML. Multiple inference and gender differences in the effects of early intervention: A reevaluation of the abecedarian, perry preschool, and early training projects. Journal of the American Statistical Association. 2008; 103; 484: 1481-1495. 10.1198/016214508000000841</bibtext> </blist> <blist> <bibl id="bib3" idref="ref55" type="bt">3</bibl> <bibtext> Assari S, Boyce S, Bazargan M, Caldwell CH. Mathematical performance of American youth: Diminished returns of educational attainment of Asian-American parents. Education Sciences. 2020; 10; 2: 32. 10.3390/educsci10020032</bibtext> </blist> <blist> <bibl id="bib4" idref="ref56" type="bt">4</bibl> <bibtext> Azmitia M, Cooper CR, Brown JR. Support and guidance from families, friends, and teachers in Latino early adolescents' math pathways. The Journal of Early Adolescence. 2009; 29; 1: 142-169. 10.1177/0272431608324476</bibtext> </blist> <blist> <bibl id="bib5" idref="ref85" type="bt">5</bibl> <bibtext> Bandura A. Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review. 1977; 84; 2: 191-215. 10.1037/0033-295X.84.2.191</bibtext> </blist> <blist> <bibl id="bib6" idref="ref36" type="bt">6</bibl> <bibtext> Barroso C, Ganley CM, McGraw AL, Geer EA, Hart SA, Daucourt MC. A meta-analysis of the relation between math anxiety and math achievement. Psychological Bulletin. 2021; 147; 2: 134-168. 10.1037/bul0000307</bibtext> </blist> <blist> <bibl id="bib7" idref="ref95" type="bt">7</bibl> <bibtext> Bartholomew B. Why we can't always get what we want. Phi Delta Kappan. 2007; 88; 8: 593-598. 10.1177/003172170708800809</bibtext> </blist> <blist> <bibl id="bib8" idref="ref64" type="bt">8</bibl> <bibtext> Blinder AS. Wage discrimination: Reduced form and structural estimates. The Journal of Human Resources. 1973; 8; 4: 436. 10.2307/144855</bibtext> </blist> <blist> <bibl id="bib9" idref="ref73" type="bt">9</bibl> <bibtext> Brandmiller C, Dumont H, Becker M. Teacher perceptions of learning motivation and classroom behavior: The role of student characteristics. Contemporary Educational Psychology. 2020; 63. 10.1016/j.cedpsych.2020.101893101893</bibtext> </blist> <blist> <bibtext> Brodish AB, Devine PG. The role of performance–avoidance goals and worry in mediating the relationship between stereotype threat and performance. Journal of Experimental Social Psychology. 2009; 45; 1: 180-185. 10.1016/j.jesp.2008.08.005</bibtext> </blist> <blist> <bibtext> Carey E, Hill F, Devine A, Szücs D. The chicken or the egg? The direction of the relationship between mathematics anxiety and mathematics performance. Frontiers in Psychology. 2016. 10.3389/fpsyg.2015.01987</bibtext> </blist> <blist> <bibtext> Ceballo R, Maurizi LK, Suarez GA, Aretakis MT. Gift and sacrifice: Parental involvement in Latino adolescents' education. Cultural Diversity and Ethnic Minority Psychology. 2014; 20; 1: 116-127. 10.1037/a0033472</bibtext> </blist> <blist> <bibtext> Cheo R. Small rewards or some encouragement? Using an experiment in China to test extrinsic motivation on academic performance. The Singapore Economic Review. 2017; 62; 04: 797-808. 10.1142/S0217590817400276</bibtext> </blist> <blist> <bibtext> Cherng H-YS, Hannum E, Lu C, Kong PA, Yu XStevens PAJ, Dworkin AG. China: sociological perspectives on ethnicity and education: Views from Chinese and English literatures. The palgrave handbook of race and ethnic inequalities in education. 2019: Berlin; Springer International Publishing: 301-344. 10.1007/978-3-319-94724-2_8</bibtext> </blist> <blist> <bibtext> Ching BH-H, Kong KHC, Wu HX, Chen TT. Examining the reciprocal relations of mathematics anxiety to quantitative reasoning and number knowledge in Chinese children. Contemporary Educational Psychology. 2020; 63. 10.1016/j.cedpsych.2020.101919101919</bibtext> </blist> <blist> <bibtext> Connell JP, Spencer MB, Aber JL. Educational risk and resilience in African-American youth: Context, self, action, and outcomes in school. Child Development. 1994; 65; 2: 493. 10.2307/1131398</bibtext> </blist> <blist> <bibtext> Daskalogianni, K, & Simpson, A. (2000). Towards a definition of attitude: the relationship between the affective and the cognitive in pre- university students. In T. Nakahara & M. Koyama (Eds.), Proceedings of the 24th Conference of the International Group for the Psychology of Mathematics Education (Vol. 2, pp. 217– 224), Vol. 2, Hirishima, Japan.</bibtext> </blist> <blist> <bibtext> Dowker A, Sarkar A, Looi CY. Mathematics anxiety: What have we learned in 60 years?. Frontiers in Psychology. 2016. 10.3389/fpsyg.2016.00508</bibtext> </blist> <blist> <bibtext> Eccles J, Wigfield A, Harold RD, Blumenfeld P. Age and gender differences in children's self- and task perceptions during elementary school. Child Development. 1993; 64; 3: 830. 10.2307/1131221</bibtext> </blist> <blist> <bibtext> Erturan S, Jansen B. An investigation of boys' and girls' emotional experience of math, their math performance, and the relation between these variables. European Journal of Psychology of Education. 2015; 30; 4: 421-435. 10.1007/s10212-015-0248-7</bibtext> </blist> <blist> <bibtext> Fan X, Zhang X, Tang B, Chang F. Family background, parental involvement, and mathematics anxiety—empirical analysis based on the rural areas of Shaanxi and Taiwan. Journal of Mathematics Education. 2021; 30; 2: 14-19</bibtext> </blist> <blist> <bibtext> Fogliati VJ, Bussey K. Stereotype threat reduces motivation to improve: Effects of stereotype threat and feedback on women's intentions to improve mathematical ability. Psychology of Women Quarterly. 2013; 37; 3: 310-324. 10.1177/0361684313480045</bibtext> </blist> <blist> <bibtext> Fuligni AJ. The academic achievement of adolescents from immigrant families: The role of family background, attitudes, and behavior. Child Development. 1997; 68; 2: 351-363. 10.1111/j.1467-8624.1997.tb01944.x</bibtext> </blist> <blist> <bibtext> Grothérus A, Jeppsson F, Samuelsson J. Formative scaffolding: How to alter the level and strength of self-efficacy and foster self-regulation in a mathematics test situation. Educational Action Research. 2019; 27; 5: 667-690. 10.1080/09650792.2018.1538893</bibtext> </blist> <blist> <bibtext> Guay F, Marsh HW, Boivin M. Academic self-concept and academic achievement: Developmental perspectives on their causal ordering. Journal of Educational Psychology. 2003; 95; 1: 124-136. 10.1037/0022-0663.95.1.124</bibtext> </blist> <blist> <bibtext> Halpern-Manners A, Warren JR, Brand JE. Dynamic measures of primary and secondary school characteristics: Implications for school effects research. Social Science Research. 2009; 38; 2: 397-411. 10.1016/j.ssresearch.2008.11.002</bibtext> </blist> <blist> <bibtext> Hannum E. Educational stratification by ethnicity in China: Enrollment and attainment in the early reform years. Demography. 2002; 39; 1: 95-117. 10.1353/dem.2002.0005</bibtext> </blist> <blist> <bibtext> Hannum E, Wang M. China: A case study in rapid poverty reduction. In indigenous people, poverty, and development. 2012: Cambridge; Cambridge University Press</bibtext> </blist> <blist> <bibtext> Hattie J, Timperley H. The power of feedback. Review of Educational Research. 2007; 77; 1: 81-112. 10.3102/003465430298487</bibtext> </blist> <blist> <bibtext> Henan Bureau of Statistics. (2020). Henan Statistical Yearbook-2020. China Statistics Press.</bibtext> </blist> <blist> <bibtext> Jia X, Zhang X, Jing J, Zhou M, Li H, Friesen D, Ma Y. The impact of nonboarding on the development of disadvantaged boarding students in western rural China. Asia Pacific Education Review. 2022; 23; 1: 131-150. 10.1007/s12564-022-09742-z</bibtext> </blist> <blist> <bibtext> Kaifeng Bureau of Statistics. (2020). Kaifeng Statistical Yearbook-2020.</bibtext> </blist> <blist> <bibtext> Kolenikov S, Angeles G. Socioeconomic status measurement with discrete proxy variables: Is principal component analysis a reliable answer?. Review of Income and Wealth. 2009; 55; 1: 128-165. 10.1111/j.1475-4991.2008.00309.x</bibtext> </blist> <blist> <bibtext> Kriegbaum K, Becker N, Spinath B. The relative importance of intelligence and motivation as predictors of school achievement: A meta-analysis. Educational Research Review. 2018; 25: 120-148. 10.1016/j.edurev.2018.10.001</bibtext> </blist> <blist> <bibtext> Leibold J. Interior ethnic minority boarding schools: China's bold and unpredictable educational experiment. Asian Studies Review. 2019; 43; 1: 3-15. 10.1080/10357823.2018.1548572</bibtext> </blist> <blist> <bibtext> Lin S-H, Huang Y-C. Development and application of a Chinese version of the short attitudes toward mathematics inventory. International Journal of Science and Mathematics Education. 2016; 14; 1: 193-216. 10.1007/s10763-014-9563-8</bibtext> </blist> <blist> <bibtext> Ma Y, Fairlie R, Loyalka P, Rozelle S. Isolating the "Tech" from EdTech: experimental evidence on computer assisted learning in China. National Bureau of Economic Research. 2020. 10.3386/w26953</bibtext> </blist> <blist> <bibtext> Ma Y, Zhang X, Abbey C, Hu D, Lee O, Hung W, Chang C, Wu C-I, Friesen D, Rozelle S. Computer assisted learning and academic performance in rural Taiwan. Journal of Research on Educational Effectiveness. 2023. 10.1080/19345747.2023.2279167</bibtext> </blist> <blist> <bibtext> Molnar A, Garcia D. The battle over commercialized schools. Educational Leadership. 2006; 63; 7: 78-82</bibtext> </blist> <blist> <bibtext> Mutegi CM, Gitonga CM, Rugano P. Mathematics anxiety, attitude and performance among secondary school students in Kenya. Educational Research and Reviews. 2021; 16; 6: 226-235. 10.5897/ERR2021.4119</bibtext> </blist> <blist> <bibtext> Oaxaca R. Male-female wage differentials in urban labor markets. International Economic Review. 1973; 14; 3: 693. 10.2307/2525981</bibtext> </blist> <blist> <bibtext> Okolo CM, Bahr CM, Gardner JE. Increasing achievement motivation of elementary school students with mild disabilities. Intervention in School and Clinic. 1995; 30; 5: 279-286. 10.1177/105345129503000505</bibtext> </blist> <blist> <bibtext> Otis N, Grouzet FME, Pelletier LG. Latent motivational change in an academic setting: A 3-year longitudinal study. Journal of Educational Psychology. 2005; 97; 2: 170-183. 10.1037/0022-0663.97.2.170</bibtext> </blist> <blist> <bibtext> Pantoja N, Schaeffer MW, Rozek CS, Beilock SL, Levine SC. Children's math anxiety predicts their math achievement over and above a key foundational math skill. Journal of Cognition and Development. 2020; 21; 5: 709-728. 10.1080/15248372.2020.1832098</bibtext> </blist> <blist> <bibtext> Pekrun R, Lichtenfeld S, Marsh HW, Murayama K, Goetz T. Achievement emotions and academic performance: Longitudinal models of reciprocal effects. Child Development. 2017; 88; 5: 1653-1670. 10.1111/cdev.12704</bibtext> </blist> <blist> <bibtext> Person, P. (1990). Motivational theories as applied to ways to improve school attendance of high school students.</bibtext> </blist> <blist> <bibtext> Phillips N, Lindsay G. Motivation in gifted students. High Ability Studies. 2006; 17; 1: 57-73. 10.1080/13598130600947119</bibtext> </blist> <blist> <bibtext> Putwain DW, Schmitz EA, Wood P, Pekrun R. The role of achievement emotions in primary school mathematics: Control–value antecedents and achievement outcomes. British Journal of Educational Psychology. 2021; 91; 1: 347-367. 10.1111/bjep.12367</bibtext> </blist> <blist> <bibtext> Ramirez G, Chang H, Maloney EA, Levine SC, Beilock SL. On the relationship between math anxiety and math achievement in early elementary school: The role of problem solving strategies. Journal of Experimental Child Psychology. 2016; 141: 83-100. 10.1016/j.jecp.2015.07.014</bibtext> </blist> <blist> <bibtext> Ramirez G, Gunderson EA, Levine SC, Beilock SL. Math anxiety, working memory, and math achievement in early elementary school. Journal of Cognition and Development. 2013; 14; 2: 187-202. 10.1080/15248372.2012.664593</bibtext> </blist> <blist> <bibtext> Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika. 1983; 70; 1: 41-55. 10.1093/biomet/70.1.41</bibtext> </blist> <blist> <bibtext> Ryan RM, Deci EL. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist. 2000; 55; 1: 68-78. 10.1037/0003-066X.55.1.68</bibtext> </blist> <blist> <bibtext> Samuelsson J. Developing students' relationships with mathematics. Educational Action Research. 2023; 31; 2: 180-194. 10.1080/09650792.2021.1899012</bibtext> </blist> <blist> <bibtext> Seo E, Lee Y. Stereotype threat in high school classrooms: How it links to teacher mindset climate, mathematics anxiety, and achievement. Journal of Youth and Adolescence. 2021; 50; 7: 1410-1423. 10.1007/s10964-021-01435-x</bibtext> </blist> <blist> <bibtext> Sonnenschein S, Sun S. Racial/ethnic differences in kindergartners' reading and math skills: Parents' knowledge of children's development and home-based activities as mediators: Parent knowledge and home activities. Infant and Child Development. 2017; 26; 5. 10.1002/icd.2010e2010</bibtext> </blist> <blist> <bibtext> Steele CM, Aronson J. Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology. 1995; 69; 5: 797-811. 10.1037/0022-3514.69.5.797</bibtext> </blist> <blist> <bibtext> Stevens T, Olivárez A, Hamman D. The role of cognition, motivation, and emotion in explaining the mathematics achievement gap between Hispanic and white students. Hispanic Journal of Behavioral Sciences. 2006; 28; 2: 161-186. 10.1177/0739986305286103</bibtext> </blist> <blist> <bibtext> Su A, He W. Exploring factors linked to the mathematics achievement of ethnic minority students in China for sustainable development: A multilevel modeling analysis. Sustainability. 2020; 12; 7: 2755. 10.3390/su12072755</bibtext> </blist> <blist> <bibtext> Taylor G, Jungert T, Mageau GA, Schattke K, Dedic H, Rosenfield S, Koestner R. A self-determination theory approach to predicting school achievement over time: The unique role of intrinsic motivation. Contemporary Educational Psychology. 2014; 39; 4: 342-358. 10.1016/j.cedpsych.2014.08.002</bibtext> </blist> <blist> <bibtext> Valentine JC, DuBois DL, Cooper H. The relation between self-beliefs and academic achievement: A meta-analytic review. Educational Psychologist. 2004; 39; 2: 111-133. 10.1207/s15326985ep3902_3</bibtext> </blist> <blist> <bibtext> Vukovic RK, Kieffer MJ, Bailey SP, Harari RR. Mathematics anxiety in young children: Concurrent and longitudinal associations with mathematical performance. Contemporary Educational Psychology. 2013; 38; 1: 1-10. 10.1016/j.cedpsych.2012.09.001</bibtext> </blist> <blist> <bibtext> Wang L, Min W, Zhang S, Shi Y, Rozelle S. Math and Chinese-language learning. Asian Survey. 2018; 58; 5: 797-821. 10.1525/as.2018.58.5.797</bibtext> </blist> <blist> <bibtext> Wang P, Cui H, Ye L. Office for the seventh national census of the state council. 2020 China census yearbook. 2022: Beijing; China Statistics Press</bibtext> </blist> <blist> <bibtext> Wu MLee WO, Brown P, Goodwin AL, Green A. The impact of measures aimed at achieving equity and quality in Chinese ethnic minority education. International handbook on education development in Asia-Pacific. 2022: Berlin; Springer Nature Singapore: 1-14</bibtext> </blist> <blist> <bibtext> Wu M, Parkay FW, Pitre PE. Student-faculty interaction: A key to academic integration and success for ethnic minority students at a major university in southwest China. Frontiers of Education in China. 2016; 11; 4: 483-502. 10.1007/BF03397137</bibtext> </blist> <blist> <bibtext> Wu SS, Barth M, Amin H, Malcarne V, Menon V. Math anxiety in second and third graders and its relation to mathematics achievement. Frontiers in Psychology. 2012. 10.3389/fpsyg.2012.00162</bibtext> </blist> <blist> <bibtext> Wu SS, Willcutt EG, Escovar E, Menon V. Mathematics achievement and anxiety and their relation to internalizing and externalizing behaviors. Journal of Learning Disabilities. 2014; 47; 6: 503-514. 10.1177/0022219412473154</bibtext> </blist> <blist> <bibtext> Xuchang Bureau of Statistics. (2019). 2019 Xuchang city national economic and social development statistical bulletin.</bibtext> </blist> <blist> <bibtext> Ya'an Bureau of Statistics. (2020). Ya'an statistical yearbook-2020.</bibtext> </blist> <blist> <bibtext> Yang J, Sicular T, Lai D. The changing determinants of high school attainment in rural China. China Economic Review. 2014; 30: 551-566. 10.1016/j.chieco.2013.10.001</bibtext> </blist> <blist> <bibtext> Yang Y, Wang H, Zhang L, Sylvia S, Luo R, Shi Y, Wang W, Rozelle S. The Han-minority achievement gap, language, and returns to schools in rural China. Economic Development and Cultural Change. 2015; 63; 2: 319-359. 10.1086/679070</bibtext> </blist> <blist> <bibtext> Yi H, Zhang L, Luo R, Shi Y, Mo D, Chen X, Brinton C, Rozelle S. Dropping out: Why are students leaving junior high in China's poor rural areas?. International Journal of Educational Development. 2012; 32; 4: 555-563. 10.1016/j.ijedudev.2011.09.002</bibtext> </blist> <blist> <bibtext> Zaefarian G, Iurkov V, Koval M. Variance decomposition analysis: What is it and how to perform it – a complete guide for B2B researchers. Industrial Marketing Management. 2022; 107: 315-322. 10.1016/j.indmarman.2022.10.020</bibtext> </blist> <blist> <bibtext> Zhang J, Zhao N, Kong QP. The relationship between math anxiety and math performance: A meta-analytic investigation. Frontiers in Psychology. 2019; 10: 1613. 10.3389/fpsyg.2019.01613</bibtext> </blist> </ref> <ref id="AN0183282411-24"> <title> Footnotes </title> <blist> <bibtext> Han Chinese refers to the largest ethnic group in China, making up approximately 91% of the country's population (Wang et al., [63]).</bibtext> </blist> <blist> <bibtext> Ethnic minority refers to the various ethnic groups in China that are recognized by the Chinese government and are distinct from the majority Han Chinese population, such as Zhuang, Tibetan, etc.</bibtext> </blist> <blist> <bibtext> Non-pep textbooks refer to textbooks that do not follow the curriculum set by the People's Education Press (PEP) in China. The PEP is a major publisher of textbooks in China, and the majority of schools in China teach with a same set of textbooks published by PEP.</bibtext> </blist> <blist> <bibtext> Non-six-year schools refer to schools that do not follow China's standard six-year primary school education system. These schools have different curriculum structures and length for completing primary school education, possibly due to regional education policies.</bibtext> </blist> </ref> <aug> <p>By Xiyuan Jia; Xinwu Zhang; Andrew Rule and Yue Ma</p> <p>Reported by Author; Author; Author; Author</p> <p></p> <p>Xiyuan Jia Xiyuan Jia is a Ph.D. Candidate at Northwest University in Xi'an, China, with research focusing on rural education and healthcare in China. She has extensive experience in fieldwork projects.</p> <p>Xinwu Zhang Xinwu Zhang is a Postdoctoral Fellow at the Stanford Center on China's Economy and Institutions, Stanford University. His research examines the economics of education and human capital in rural China, and he has substantial experience in fieldwork management.</p> <p>Andrew Rule Andrew Rule is an academic editor at the Stanford Center on China's Economy and Institutions, Stanford University, contributing to the editing of research findings.</p> <p>Yue Ma Yue Ma is a Research Scholar at the Stanford Center on China's Economy and Institutions, Stanford University. His research focuses on the economics of education and human capital in rural China, with extensive experience in fieldwork management.</p> </aug> <nolink nlid="nl1" bibid="bib70" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib63" firstref="ref4"></nolink> <nolink nlid="nl3" bibid="bib71" firstref="ref6"></nolink> <nolink nlid="nl4" bibid="bib58" firstref="ref7"></nolink> <nolink nlid="nl5" bibid="bib14" firstref="ref8"></nolink> <nolink nlid="nl6" bibid="bib35" firstref="ref9"></nolink> <nolink nlid="nl7" bibid="bib64" firstref="ref10"></nolink> <nolink nlid="nl8" bibid="bib62" firstref="ref12"></nolink> <nolink nlid="nl9" bibid="bib72" firstref="ref13"></nolink> <nolink nlid="nl10" bibid="bib28" firstref="ref15"></nolink> <nolink nlid="nl11" bibid="bib65" firstref="ref16"></nolink> <nolink nlid="nl12" bibid="bib27" firstref="ref17"></nolink> <nolink nlid="nl13" bibid="bib21" firstref="ref19"></nolink> <nolink nlid="nl14" bibid="bib38" firstref="ref20"></nolink> <nolink nlid="nl15" bibid="bib13" firstref="ref21"></nolink> <nolink nlid="nl16" bibid="bib56" firstref="ref22"></nolink> <nolink nlid="nl17" bibid="bib22" firstref="ref23"></nolink> <nolink nlid="nl18" bibid="bib10" firstref="ref24"></nolink> <nolink nlid="nl19" bibid="bib54" firstref="ref25"></nolink> <nolink nlid="nl20" bibid="bib17" firstref="ref26"></nolink> <nolink nlid="nl21" bibid="bib36" firstref="ref27"></nolink> <nolink nlid="nl22" bibid="bib11" firstref="ref28"></nolink> <nolink nlid="nl23" bibid="bib45" firstref="ref29"></nolink> <nolink nlid="nl24" bibid="bib48" firstref="ref30"></nolink> <nolink nlid="nl25" bibid="bib52" firstref="ref31"></nolink> <nolink nlid="nl26" bibid="bib34" firstref="ref32"></nolink> <nolink nlid="nl27" bibid="bib57" firstref="ref33"></nolink> <nolink nlid="nl28" bibid="bib18" firstref="ref34"></nolink> <nolink nlid="nl29" bibid="bib40" firstref="ref35"></nolink> <nolink nlid="nl30" bibid="bib15" firstref="ref37"></nolink> <nolink nlid="nl31" bibid="bib20" firstref="ref38"></nolink> <nolink nlid="nl32" bibid="bib44" firstref="ref40"></nolink> <nolink nlid="nl33" bibid="bib50" firstref="ref41"></nolink> <nolink nlid="nl34" bibid="bib49" firstref="ref42"></nolink> <nolink nlid="nl35" bibid="bib61" firstref="ref43"></nolink> <nolink nlid="nl36" bibid="bib66" firstref="ref44"></nolink> <nolink nlid="nl37" bibid="bib67" firstref="ref45"></nolink> <nolink nlid="nl38" bibid="bib74" firstref="ref46"></nolink> <nolink nlid="nl39" bibid="bib69" firstref="ref48"></nolink> <nolink nlid="nl40" bibid="bib30" firstref="ref50"></nolink> <nolink nlid="nl41" bibid="bib32" firstref="ref51"></nolink> <nolink nlid="nl42" bibid="bib68" firstref="ref54"></nolink> <nolink nlid="nl43" bibid="bib31" firstref="ref57"></nolink> <nolink nlid="nl44" bibid="bib37" firstref="ref58"></nolink> <nolink nlid="nl45" bibid="bib33" firstref="ref60"></nolink> <nolink nlid="nl46" bibid="bib73" firstref="ref62"></nolink> <nolink nlid="nl47" bibid="bib41" firstref="ref65"></nolink> <nolink nlid="nl48" bibid="bib51" firstref="ref68"></nolink> <nolink nlid="nl49" bibid="bib12" firstref="ref74"></nolink> <nolink nlid="nl50" bibid="bib26" firstref="ref75"></nolink> <nolink nlid="nl51" bibid="bib53" firstref="ref76"></nolink> <nolink nlid="nl52" bibid="bib55" firstref="ref77"></nolink> <nolink nlid="nl53" bibid="bib16" firstref="ref78"></nolink> <nolink nlid="nl54" bibid="bib19" firstref="ref79"></nolink> <nolink nlid="nl55" bibid="bib23" firstref="ref80"></nolink> <nolink nlid="nl56" bibid="bib25" firstref="ref81"></nolink> <nolink nlid="nl57" bibid="bib60" firstref="ref82"></nolink> <nolink nlid="nl58" bibid="bib43" firstref="ref83"></nolink> <nolink nlid="nl59" bibid="bib59" firstref="ref84"></nolink> <nolink nlid="nl60" bibid="bib29" firstref="ref86"></nolink> <nolink nlid="nl61" bibid="bib24" firstref="ref87"></nolink> <nolink nlid="nl62" bibid="bib46" firstref="ref90"></nolink> <nolink nlid="nl63" bibid="bib47" firstref="ref92"></nolink> <nolink nlid="nl64" bibid="bib42" firstref="ref93"></nolink> <nolink nlid="nl65" bibid="bib39" firstref="ref94"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1461163
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: A Decomposition of the Han-Minority Achievement Gap in Rural China
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Xiyuan+Jia%22">Xiyuan Jia</searchLink><br /><searchLink fieldCode="AR" term="%22Xinwu+Zhang%22">Xinwu Zhang</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0002-4574-536X">0000-0002-4574-536X</externalLink>)<br /><searchLink fieldCode="AR" term="%22Andrew+Rule%22">Andrew Rule</searchLink><br /><searchLink fieldCode="AR" term="%22Yue+Ma%22">Yue Ma</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Social+Psychology+of+Education%3A+An+International+Journal%22"><i>Social Psychology of Education: An International Journal</i></searchLink>. 2025 28(1).
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Springer. Available from: Springer Nature. One New York Plaza, Suite 4600, New York, NY 10004. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-460-1700; e-mail: customerservice@springernature.com; Web site: https://link.springer.com/
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Elementary+Education%22">Elementary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink><br /><searchLink fieldCode="DE" term="%22Rural+Schools%22">Rural Schools</searchLink><br /><searchLink fieldCode="DE" term="%22Ethnic+Groups%22">Ethnic Groups</searchLink><br /><searchLink fieldCode="DE" term="%22Achievement+Gap%22">Achievement Gap</searchLink><br /><searchLink fieldCode="DE" term="%22Mathematics+Achievement%22">Mathematics Achievement</searchLink><br /><searchLink fieldCode="DE" term="%22Mathematics+Anxiety%22">Mathematics Anxiety</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Attitudes%22">Student Attitudes</searchLink><br /><searchLink fieldCode="DE" term="%22Elementary+School+Students%22">Elementary School Students</searchLink><br /><searchLink fieldCode="DE" term="%22Minority+Group+Students%22">Minority Group Students</searchLink><br /><searchLink fieldCode="DE" term="%22Correlation%22">Correlation</searchLink>
– Name: Subject
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22China%22">China</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1007/s11218-025-10021-z
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1381-2890<br />1573-1928
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Despite increased investment in rural education in recent years, the academic achievement of ethnic minority students in rural China continues to lag behind their Han Chinese peers. Few studies have used up-to-date data to measure the size of this achievement gap, and none have examined whether differential math attitudes and math anxiety play a role in perpetuating it. Using data from a survey of 16,747 students in 235 rural primary schools, we find a significant disparity in math performance between ethnic minority and Han students, with minority students scoring 0.25 standard deviations lower than their Han peers. Multivariate analysis reveals that differences in math affinity and math anxiety were significantly associated with math achievement among ethnic minority and Han students. Decomposition analysis further reveals that up to 37% of the achievement gap can be explained by variations in student, parent, and teacher characteristics. The difference in household asset was the largest explained contributor to the Han-minority achievement gap, followed by math affinity and math anxiety. This study underscores the importance of addressing students' math affinity and math anxiety in narrowing rural achievement gaps and promoting equitable educational outcomes in rural China and abroad.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2025
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1461163
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1461163
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1007/s11218-025-10021-z
    Languages:
      – Text: English
    Subjects:
      – SubjectFull: Foreign Countries
        Type: general
      – SubjectFull: Rural Schools
        Type: general
      – SubjectFull: Ethnic Groups
        Type: general
      – SubjectFull: Achievement Gap
        Type: general
      – SubjectFull: Mathematics Achievement
        Type: general
      – SubjectFull: Mathematics Anxiety
        Type: general
      – SubjectFull: Student Attitudes
        Type: general
      – SubjectFull: Elementary School Students
        Type: general
      – SubjectFull: Minority Group Students
        Type: general
      – SubjectFull: Correlation
        Type: general
      – SubjectFull: China
        Type: general
    Titles:
      – TitleFull: A Decomposition of the Han-Minority Achievement Gap in Rural China
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Xiyuan Jia
      – PersonEntity:
          Name:
            NameFull: Xinwu Zhang
      – PersonEntity:
          Name:
            NameFull: Andrew Rule
      – PersonEntity:
          Name:
            NameFull: Yue Ma
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 12
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 1381-2890
            – Type: issn-electronic
              Value: 1573-1928
          Numbering:
            – Type: volume
              Value: 28
            – Type: issue
              Value: 1
          Titles:
            – TitleFull: Social Psychology of Education: An International Journal
              Type: main
ResultId 1