The purpose of this study is to investigate the relationship between private tutoring and Chinese students’ Gaokao (National College Entrance Exam, NCEE) performance including other potential factors (e.g. gender, registered residence status, family SES, academic track, etc.). Specifically, this study examines three goals: a) the difference between students who choose to participate in private tutoring and students who does not choose to participate in private tutoring in terms of their background information, b) the factors influencing students’ Gaokao performance, and c) the mechanism of how private tutoring and other factors influence students’ Gaokao performance.

This study adopts three theoretical perspectives to explore the relationship between private tutoring and Chinese students’ Gaokao performance: 1) Dang and Rogers’ (2008) supply and demand framework in education with private tutoring; 2) Beruvides’ (1997) learning curve model with threshold region; and 3) the educational production function (Hanushek, 1979). The supply and demand framework in education with private tutoring helps people understand the interaction of supply and demand for public education (and private tutoring) in different situations. It indicates that households can consume more education after public education cannot meet their demand when the private tutoring is available (Dang & Rogers, 2008). The learning theory and the learning curve model provide a crucial framework to explain why parents and students demand private tutoring. By adopting the educational production function, this study was able to investigate the relationship between private tutoring and students’ Gaokao performance.

In this study, a survey on students’ shadow education participation was developed as the instrument. This study employed a quantitative approach. Descriptive statistics were used to examine students’ background characteristics, students’ family socioeconomic status, and private tutoring participation information. Independent samples t-tests and chi-square tests were conducted to identify the differences between students who participated in private tutoring before college and those who did not. Multiple regression analyses were administrated to investigate the factors (including private tutoring) that influence students’ Gaokao score. A Path analysis was employed to measure the robustness of a proposed model.

The results indicated that students who participated in private tutoring had the characteristics that much more being female, more had non-agricultural registered residence status, lived in the urban areas and chose the social science track in high school. Less of them were first generation college students. In terms of the variables measuring participants’ financial status, students who participated in private tutoring were more likely to have an apartment or a house and to have internet access and air conditioners at home. However, less of these students reported being top 20% of the class during high school. The results of independent samples t-tests and Pearson chi-square tests indicated that there is statistically significant difference in background characteristics including, parental education level, class rank in high school, gender, registered residence status, location of residence, and academic track in high school, between students who participated in private tutoring and students who did not participate in private tutoring. The multiple regression analysis showed that gender, class rank in high school, academic track, parental educational level, and private tutoring were statistically significant factors on students’ Gaokao performance. The path analysis indicated that students with longer private tutoring participation time had lower Gaokao score. In author’s opinion, the negative effects of private tutoring on Gaokao score suggested that private tutoring was employed by students for a compensatory or remedial purpose. Rather than helping good students become better students, private tutoring served a role that helping students with lower academic achievement to prevent their study from being worse. Further research is needed to have a better understanding about this phenomenon.