







































 

 

 

 

 

The Effect of Education Groups on the 

Quality of Urban and Rural Compulsory 

Education: An Empirical Analysis 

Based on CEPS Data

Gang Cheng , Sihui Du, Yanan Xu, Luoping Li 

 
Faculty of Education, Beijing Normal University, Beijing 100091, China. 

 

Abstract: We retrieved data from China Education Panel Survey (CEPS) 

and adopted the hierarchical linear model as well as the quantile 

regression method to examine the effect of education groups on the 
quality of urban and rural compulsory education. Student cognitive 

ability was cited as the evaluation index of education quality. According 

to the research findings, the overall effect of education groups on the 
quality of compulsory education is significantly positive, but there is 

urban-rural heterogeneity in this effect: the positive impact occurs only 

to urban member schools, while the education quality of rural member 
schools has not been improved by school grouping. In addition, there 

exists individual difference in this effect; school grouping significantly 
and negatively affects students from rural member schools and with 

medium and lower cognitive ability. 

Best Evidence in Chinese Education 2022; 11(2):1517-1522. 

Doi: 10.15354/bece.22.ab004 

How to Cite: Cheng, G., Du, S., Xu, Y., & Li, L. (2022). The effect of education 

groups on the quality of urban and rural compulsory education: An empirical 

analysis based on CEPS data. Best Evidence in Chinese Education, 11(2):1517-

1522. 

Keywords: Education Group, Compulsory Education, Education Quality, Urban- Rural 

Disparity 

 

 

 

 



Cheng et al. (Beijing). Education Groups and Education Quality of Urban and Rural China. 

BECE, Vol.11, No.2, 2022 1518 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 

Correspondence to: Gang Cheng, Faculty of Education, Beijing Normal University, Beijing 100091, China. E-mail: 
cg@bnu.edu.cn  

Conflict of Interests: None. 
 

© 2022 Insights Publisher. All rights reserved. 

Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the 

Creative Commons Attribution-NonCommercial 4.0 License 

(http://www.creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and distribution of 

the work without further permission provided the original work is attributed by the Insights Publisher. 

mailto:cg@bnu.edu.cn
http://www.creativecommons.org/licenses/by-nc/4.0/


Cheng et al. (Beijing). Education Groups and Education Quality of Urban and Rural China. 

BECE, Vol.11, No. 2, 2022 1519 

Introduction 

The education group at the compulsory education level is a form of school association 

composed of a certain number of primary and/or junior secondary schools that are con-

tractually tied to achieve scale effect. It is designed to expand access to high-quality 

educational resources through the alliance of advantaged and disadvantaged schools. An 

education group constituted by prestigious urban schools and ordinary rural schools is 

established to universalize the frontier educational concepts and school management 

skills of the former and to promote the balanced development of urban and rural com-

pulsory education. Based on the data from the China Education Panel Survey (CEPS) 

2013-2015, this study uses the hierarchical linear model (HLM) and quantile regression 

method to evaluate the effect of education groups on the quality of urban and rural 

compulsory education quality as well as on the development of cognitive competencies 

of students with differential cognitive levels. The purpose of this paper is to identify the 

specific effects of school grouping on bridging the quality gap between urban and rural 

schools and to optimize the strategies for education group operation. 

Data Source 

This study sampled students who participated in the CEPS baseline survey of 2013-

2014 academic year and the follow-up survey of 2014-2015 academic year and drew the 

data of 6520 students from 88 schools. In the stage of baseline survey, they were in 

grade 7. Five of the sampled schools were members of education groups and contribut-

ed 316 students for this investigation, which accounted for 4.85% of the total sample of 

students. Further analysis showed that in the districts where the five schools are located, 

there are altogether 12 schools (715 students) included in the sample and that in this 

case, the portion of students from member schools of education groups was 44.20%. 

Research Findings 

Statistical Descriptions of Member Schools of Educa-

tion Groups 

According to the sample analysis results, students from member schools of education 

groups, compared with their peers from non-member schools, displayed considerably 

higher cognitive ability and featured a lower proportion of rural “registered residence” 

(Hukou) as well as a higher share of “the only child in the family”. Member schools of 

education groups are more likely to be located in urban areas, with higher school rank-

ings, stronger school finances, and higher teacher qualifications. Nonetheless, there 

were no remarkable differences in basic facilities and student-teacher ratio between 

member and non-member schools. 

 



Cheng et al. (Beijing). Education Groups and Education Quality of Urban and Rural China. 

BECE, Vol.11, No. 2, 2022 1520 

The Impact of School Grouping on Student Cognitive 

Ability 

Full Sample Estimation 

OLS regression analysis results showed that education group membership of the school 

had a significant positive effect on the development of student cognitive ability, while 

school location in rural areas has a remarkable negative influence on the cultivation of 

student cognitive ability. The incorporation of coefficients of interaction terms into the 

analysis confirmed the urban-rural heterogeneity in the impact of school grouping. 

School rural location weakened the marginal effect of school grouping by 0 334 units, 

which reversed the positive impact of 0.208 units contributed by the variable of school 

grouping (0.208-0.334 = -0.126). Consistent with the OLS estimation results, in the hi-

erarchical linear model, the coefficient of the variable of school grouping was signifi-

cantly positive, while those of school location and interaction terms are significantly 

negative. Moreover, a comparison of the coefficients indicated that the cognitive ability 

of rural students was not positively affected by the factor of school grouping, but rather 

declined (0.176-0.271 = -0.095). 

Estimation of the Sample from the Districts Where 

Member Schools of Education Groups are located 

The OLS regression analysis results of the sample from the districts where member 

schools of education groups are located also pointed to prominent urban-rural heteroge-

neity in the effect of school grouping. Overall, school grouping had a significant posi-

tive impact on student cognitive ability, but the coefficients of interaction terms demon-

strated that the cognitive ability of students from rural member schools of education 

groups was remarkably lower than that of their urban counterparts by 0.914 units, which 

directly turned the overall positive influence of school grouping into a substantial nega-

tive impact (0.284-0.914 = -0.630). The hierarchical linear model also demonstrated 

high appropriateness. The result of chi-square test of zero model is significant. Dispari-

ties between schools accounted for about 27.1% of the gap in cognitive ability between 

students; Individual and family factors can partially explain the gap in student cognitive 

ability. The estimation results showed that the coefficients of the variable of school 

grouping and interaction terms are of significant levels, and there was positive-negative 

differentiation in their values. As a result, the influence of school grouping on rural stu-

dents’ cognitive ability was substantially negative (0.250-0.740 = -0.490). 

The Influence of School Grouping on Different Student 

Cohorts 

The regression coefficients of the variables of school grouping and interaction terms are 

significant at both the 25% and 50% quantiles. School grouping benefited urban stu-

dents with below-average and medium cognitive ability (QR _25 and QR_ 50) more 



Cheng et al. (Beijing). Education Groups and Education Quality of Urban and Rural China. 

BECE, Vol.11, No. 2, 2022 1521 

than any other student cohort, by an improvement of 0.476 and 0.537 units. For rural 

students with these levels of cognitive ability (QR_25 and QR_50), the effect of school 

grouping is reversed to -0.643 (0.476-1.119) and -0.440 (0.537-0.977) units, which 

means that the development of rural students with medium or below-average cognitive 

ability was severely impaired by the school’s membership in education group. 

Discussion and Conclusion 

The efficacy of education groups was confirmed by the research results in that the shar-

ing of staff, curriculum content, and campus culture among member schools have 

helped increase student education output. The yearly teacher in-service training of 

member schools of education groups increased from 10.3 to 22 times on average, and 

its baseline level, follow-up level and growth rate were all higher than the averages of 

non-member schools in the same district; and the growth rate of teacher training of rural 

member schools has exceeded 300%, much higher than that of their urban counterparts. 

Two types of costs should be taken into consideration in the exploration of the 

specific reasons for the heterogeneous effects of education groups on urban and rural 

compulsory education. (i) Explicit costs: in the practice of school grouping, urban 

member schools benefited from the sophisticated inner-city and inter-city transportation 

systems and ideal digital environments, which enabled the effective sharing of high-

quality courses and the smooth rotation of excellent teachers and administrators. In con-

trast, due to the geographical separation and the “digital divide” between urban and ru-

ral areas, high-quality educational resources were less accessible to rural member 

schools. (ii) Implicit costs: During the two years of the CEPS, the sampled member 

schools of education groups were in the early stage of cooperation, experiencing a diffi-

cult transition from heterogeneity to homogeneity. The survey results of school satisfac-

tion reflected the above costs in a direct way from the students’ perspective. Students at 

rural member schools had a much higher level of academic weariness and stronger in-

tention of school transfer than their urban counterparts. 

This study drew the following conclusions: first, the education group has a sig-

nificant positive influence on the quality of compulsory education and can effectively 

promote student cognitive ability; second, the influence of the education group is heter-

ogeneous in urban and rural areas with the positive effect existing only among urban 

member schools, and the quality of compulsory education in rural member schools may 

decline rather than improve; and third, there is individual heterogeneity in the impact of 

school grouping, and the impact on rural students with medium and lower cognitive 

ability is the most significantly negative. The education group is a new form of school 

organization, and the association of urban and rural schools will progressively improve 

in practice and in mutual adaptation. 
 

 

 

 

 

 

 

 



Cheng et al. (Beijing). Education Groups and Education Quality of Urban and Rural China. 

BECE, Vol.11, No. 2, 2022 1522 

 

The Chinese version of this article has been published in Journal of Educational Economics Re-

view, 2022, 7(2):20. The English version has been authorized for being publication in BECE by 

the author(s) and the Chinese journal. 

 

成刚, 杜思慧, 许亚男, & 李罗平. 集团化办学对城乡义务教育质量的影响研究—基于中国教

育追踪调查的实证分析. 教育经济评论, 2022, 7(2):20. 

 

 

Received: 26 June 2022 

Revised: 10 July 2022 

Accepted: 14 July 2022 

 

 

 

 

 


	Absbract1-GangCheng-BECE_TitleJuly_2022
	Absbract1-GangCheng-BECE_MaintextJuly2022 

