







































 

 

 

 

 

The “Theatre Effect” of Shadow Education: 

A Study Based on Data from China 

Education Panel Survey

Haiping Xue, Dancheng Xu 

 
Capital Normal University, Beijing 100089, China  

 

 

Abstract: Based on data from the China Education Panel Survey in the 

academic years of 2013-2014 and 2014-2015, this article sought to 
investigate whether shadow education can yield the “theater effect” and 

whether the “theater effect” of shadow education is heterogeneous 
between families of different socio-economic status as well as between 

urban and rural families, using the hierarchical linear Bernoulli model. 

Best Evidence in Chinese Education 2023; 13(2):1709-1713. 

Doi: 10.15354/bece.23.ar038 

How to Cite: Xue, H., & Xu, D. (2023). The “theatre effect” of shadow education: A 

study based on data from China Education Panel Survey. Best Evidence in Chinese 

Education, 13(2):1709-1713. 

Keywords: Shadow Education, Theatre Effect in Education, Peer Effects 

 

 

 

 

 

 

 

 

 

 

 

 



Xue & Xu. (Beijing). The “Theatre Effect” of Shadow Education. 

BECE, Vol.13, No.2, 2022 1710 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 

Correspondence to: Haiping Xue, Capital Normal University, Beijing 100089, China. E-mail: 
xuehaiping_416@163.com  

Conflict of Interests: None. 
 

© 2023 Insights Publisher. All rights reserved. 

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

Creative Commons Attribution-Non Commercial 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:xuehaiping_416@163.com
http://www.creativecommons.org/licenses/by-nc/4.0/


Xue & Xu. (Beijing). The “Theatre Effect” of Shadow Education. 

BECE, Vol.13, No. 2, 2023 1711 

Introduction 

When a student successfully improves his/her academic achievements with the help of 

shadow education, his/her classmates and their parents will follow suit and join in after-

school supplementary tutoring. An analogy is made between this phenomenon and the 

scenario in the theatre where if some viewers stand up from their seats to get better 

view, those sitting right behind will stand up too, and eventually all those sitting behind 

them will rise. This is how the term “theatre effect in education” was coined. This study 

established its theoretical framework based on the Reference Group theory and the Ef-

fectively Maintained Inequality (EMI) theory. Data from the China Education Panel 

Survey (CEPS) in the academic years of 2013-2014 and 2014-2015 were utilized to 

investigate whether shadow education can produce the “theater effect” and the differ-

ences in the “theater effect” of shadow education between families of different socio-

economic status (SES) as well as between urban and rural families. The hierarchical 

linear Bernoulli model was used in the analysis. In addition, the lag phase of shadow 

education participation rates of class and school peers was incorporated in the regres-

sion model to test the robustness of the “theatre effect” of shadow education.  

Research Hypotheses 

In this study, classmates and schoolmates were treated as the reference group of the 

individual student. A reference group is typically used by an individual as a standard for 

evaluating themselves and their behaviors and plays the role of “norm-setter”.  

The individual student has the most communication with their classmates dur-

ing the school days. Classmates play the major role of norm-setting of the reference 

group. Therefore, hypothesis one (H1) was proposed: The higher the shadow education 

participation rate of class and school peers, the greater the probability of the individual 

student participating in shadow education, and class peers pose a greater influence on 

individual’s participation in supplementary tutoring than schoolmates.  

Families of middle SES are typically more engaged in the educational competi-

tion, so more likely to participate in shadow education due to the peer effects. Hence, 

suggested were hypothesis two (H2): The higher the shadow education participation 

rate of classmates and schoolmates, the more likely it is that students with parents of the 

middle occupation level will participate in shadow education; and hypothesis three (H3): 

The higher the shadow education participation rates of classmates or schoolmates, the 

greater the probability that students from middle-income families will participate in off-

campus supplementary tutoring. 

Urban parents and students are more ready to make decisions about participat-

ing in shadow education, because they have more chances to be influenced by their 

peers due to their constant interactions. Thus, hypothesis four (H4) was proposed: 

Classmates’ or schoolmates’ shadow education participation has a greater effect on ur-

ban students’ participation in after-school tutoring than that of rural students, that is, 

there is heterogeneity between the class or school peer effects on urban and rural stu-

dents’ participation in shadow education. 



Xue & Xu. (Beijing). The “Theatre Effect” of Shadow Education. 

BECE, Vol.13, No. 2, 2023 1712 

Data Sources and Variables 

Data Sources 

This study sourced data from CEPS, which set the 2013-2014 academic year as the 

baseline and the grade 7 and grade 9 in that year as the initial points of investigation. A 

follow-up survey was conducted to gather information for the 2014-2015 academic year.  

Using multi-stage probability-proportional-to-size sampling (PPS), the study randomly 

selected 28 county-level regions as investigation sites, from which 483 classes from 112 

schools were randomly sampled. All students from the sampled classes were included 

as research subjects. The samples for the two academic years were made up of 14376 

and 6726 students, respectively. 

Variables 

 The Independent Variable: Participating in shadow education or not. 

 Dependent Variables: The shadow education participation rate of class peers and 

that of school peers. 

 Control Variables: Individual-, family-, and school-related factors that may influ-

ence student shadow education participation. 

Research Methods 

Null Model 

The total difference in the shadow education participation rate was decomposed into 

individual and school levels in order to examine whether there are significant gaps in 

student shadow education participation rate between schools. 

Full Model 

Based on the zero model, the full model was constructed by incorporating student- and 

school-related variables to examine their influences on students’ decisions on shadow 

education engagement. 

Conclusions 

According to the descriptive statistics, in the 2013-2014 academic year, 47% of junior 

secondary school students engaged in shadow education, and the shadow education par-

ticipation rates of students’ class and school peers were both 47%; in the 2014-2015 

academic year, 50% of students participated in shadow education, and the shadow edu-

cation participation rates of students’ class and school peers were 46%. Research find-

ings based on the hierarchical linear Bernoulli model show that: (i) The “theater effect” 

exists in shadow education, which is supported by the fact that the shadow education 

participation rates of class and school peers have robust and significant positive effects 



Xue & Xu. (Beijing). The “Theatre Effect” of Shadow Education. 

BECE, Vol.13, No. 2, 2023 1713 

on students’ participation in after-school tutoring; (ii) The “theater effect” of shadow 

education of classmates has a greater impact than that of schoolmates, which is evi-

denced by the analysis result that the coefficient of the variable of the shadow education 

participation rate of class peers is greater than that of shadow education participation 

rate of school peers; (iii) The “theater effect” of shadow education is heterogeneous 

among families of distinct SES, with the strongest impact on students from families of 

middle SES; (iv) The “theatre effect” of shadow education varies between urban and 

rural families, with more significant influences on urban students than on their rural 

counterparts. 

 
 

 

The Chinese version of this article has been published in China Economics of Education Review, 

2022, 7(1):75-98. The English version has been authorized for being publication in BECE by the 

author(s) and the Chinese journal. 

 

薛海平 & 徐丹诚. 影子教育的剧场效应研究—基于中国教育追踪调查数据分析. 教育经济评论, 

2022, 7(1):75-98. 

 

 

Received: 14 January 2023 

Revised: 17 February 2023 

Accepted: 22 February 2023 

 

 

 

 

 


	AR3-HaipingXue-BECE_TitleMarch_2023
	AR3-HaipingXue-BECE_MaintextMarch2023

