Sage Journals: Discover world-class research

Abstract

One unintended consequence of Chinese higher education expansion is an increasing socioeconomic gap in college achievement. Using 2011 College Freshmen Development Survey data, this paper engages in an empirical analysis of the association between one’s socioeconomic status, high school preparation, and college performance. This study finds that well prepared and academically engaged high school graduates achieve a significantly higher level of college development than their less prepared peers. Moreover, low-ses students have a significantly lower level of core competence and civic participation, due to their lack of academic preparation, learning engagement, and parental support in high school. This study provides a critical evidence for the life cycle skill formation argument that the deficit in early human capital investment can impair adolescent and adult attainment.

Keywords

high school preparation college success socioeconomic status

1 Introduction

Education policymakers have long been annoyed by the socioeconomic (ses thereafter) difference in college access, persistence and student development in developing countries. The rapid market reforms have produced a group of middle-income countries such as Brazil, Russia, and South Africa. One unintended consequence of these reforms is a booming but increasingly stratified postsecondary education sector, characterizing by unequal college access and completion.

China is no exception. Recent evidence demonstrates that the distribution of the higher education resources in general, the access to high quality tertiary education in particular, is extremely favoring male, middle- or high-income urban students with college educated parents. 1

For instance, a stream of literature based on newly available repeated cross-sectional data collected from multiple cohorts of undergraduate students in Beijing reveals that, the poor first-generation students have less access to selective colleges and universities even conditional on college entrance examination scores. 2 Once matriculating into college, freshmen from rural households face greater challenges in academic and social adaption. 3 Male students from metropolitan cities have much higher employment rates and starting salary than their female, rural, and low-income counterparts. 4

One cannot help asking why socioeconomic background still matters for college development, even after the strict ability sorting through college application and admission? 5 One hypothesis is that low-ses students accumulate disadvantages from preschool to high school, and this long-term deficit in human capital investment can damage child, adolescent, and adult achievement, because life cycle skill formation is a dynamic process in which early inputs strongly affect the productivity of later inputs. 6

The life cycle skill formation hypothesis implies a high correlation between one’s k-12 and postsecondary performance. This correlation may transfer one’s disadvantages in high school into college years. 7 Low-ses students may perform worse in colleges than others due to a lack of preparation in elementary school, middle school or high school.

To test this hypothesis, this paper engages in an empirical analysis of the association between one’s socioeconomic status, high school preparation, and college performance. Research questions that arise in this context are as follows: Is there a significant correlation between individual high school preparation and college development? How does socioeconomic status influence one’s college performance?

This paper develops its argument in the context of student engagement literature. It identifies a high correlation between high school engagement and college achievement. We find that well prepared and academically engaged high school graduates achieve a significantly higher level of college development than their less prepared peers. This is largely due to the fact that academic preparation and learning engagement in high school increase the level of college readiness, which is a critical threshold for college success. 8

Moreover, individual’s family background has direct and significant effects on one’s college performance. Low-ses college students have a significantly lower level of core competence and civic participation than their high-ses peers, due to their lack of academic preparation, learning engagement, and parental support in high school. This provides a critical evidence for the life cycle skill formation argument, i.e., the deficit in early human capital investment can impair adolescent and adult attainment.

This study has substantial implications for institution practice. If researchers can find a significant and strong correlation between college development and high school preparation, college admission officers will be able to predict students’ college success and make justified admission decision. Furthermore, educational policymakers can infer the quality of certain high schools by observing the variation in college achievement of their graduates.

The rest of the paper is organized as follows. The next section describes the Chinese context and reviews literature about the impact of high school experience on college access and performance. The third section establishes an analytical framework, and introduces the data and the sample. The following section reports the empirical findings and the final section closes this paper by discussing the implications of its findings.

2 The Intertwined Experience in High School and College

2.1 Chinese Context

Chinese education system is characterized with continuous sorting and competition. In China, the first wave of sorting occurs after the completion of compulsory education at the age of 15, and most students continue their education in academic or vocational high schools and a small group of graduates enter the labor market. The second wave of sorting occurs after the completion of the upper secondary education at the age of 18. At that time, vocational high school graduates join the workforce and academic graduates take the college entrance examination before entering three-year colleges or four-year universities.

The competition for elite universities is fierce, because only 112 out of more than 2240 postsecondary education institutions are considered the most selective ones. Each year, these so called “Project 985” and “Project 211” institutions take fewer than 7.5% of high school graduates. 9 They are considered the first-tier institutions in China. 10

In order to outperform others in the high-stake college entrance examination (gaokao in Chinese), the better-off families invest lots of money in school choice and private tutoring as early as first grade, even though the average effect of private tutoring is not significant for most students. 11 As for the low-income students, a high proportion of them retakes the college entrance examination after unsuccessful attempts, in order to enlarge their odds of admission to better universities. 12

Irrespective of those fancy strategies, high school years are the most critical time for achieving one’s college dream. Like in many Asian countries, Chinese parents are famous for pushing their children into the gaokao race, by providing private tutoring 13 and setting high expectation for gaokao scores. High schools also compete vigorously on college-going rate and number of students admitted by first-tier institutions, leading to the born of “super high schools.” 14 Nevertheless, low-ses students seem to be left behind in this race. For example, the proportion of rural students has dropped from one third to 10-17% in the topnotch institutions. 15

2.2 Role of High School Preparation for College Success

In the past twenty years, higher education researchers have constructed comprehensive frameworks to capture the determinants of college success. For instance, Perna and Thomas and Perna 16 analyze the longitudinal process of college success from the sociological perspective, emphasizing the role of cultural and social capital in shaping individual college access and attainment. Paulsen and St. John 17 apply the economic theory to understand the link between financial aid and retention. Tinto & Pusser 18 develop an institutional action model for promoting college persistence. Although scholars differ in their analytical perspectives, they reach one consensus that there is a high correlation between high school performance and college development.

High school academic intensity and curriculum quality are considered the most critical predictors for college access, persistence and graduation. High school grade point average (gpa thereafter) and class ranking are strong predictors of college gpa. 19 Adelman 20 finds that academic intensity and curriculum quality have a greater impact on reducing dropout risk and increasing college gpa than that of high school gpa and sat/act scores. Kuh, Kinzie, Buckley, Bridges and Hayek 21 point out that high school academic intensity is as important as parental background, educational expectation and family support for college success. The number and the degree of challenges of high school math courses increase the probability of enrolling in four-year college. 22 The number of remedial courses in freshmen year also decreases if academic intensity in high school increases, and college students are less likely to drop out if they attend high schools with high level of academic intensity. 23

Parental style and support are also critical for college success. Discussing education related topics within households, parental involvements in high school volunteer programs, parents’ contacting high schools for children’s academic issues increase the probability of enrolling in 4-year college. 24 Early planning (as early as 8th grade) for higher education has a positive effect on completing baccalaureate degree on time. 25 Appropriate parental support and involvement can mitigate the negative impacts of poverty on higher education completion. 26 In addition, parents and high school peers can influence college access and persistence. 27

Similar evidences are found in Chinese context. Bao 28 indicates that attending selective high schools and having a higher class ranking increase the probability of going to the first-tier Chinese universities, but private tutoring cost in high school is negatively correlated with attending selective institutions. Using similar data, others report that high school teaching pedagogy, student-teacher interaction, and non-academic engagement influence students’ adaption to college environment.

2.3 Unresolved Issues

Although existing literature has identified the critical roles of academic intensity, curriculum quality, and parental support in high school for predicting college success, these studies have not reached agreement on the following issues. First, how to measure high school preparation? There are lots of discussions on the influence of number of credits enrolled and the level of course content, almost no one has examined the effects of students’ study effort or student-teacher interaction. These factors capture individual’s learning engagement in high school. This study explicitly introduces measures of high school learning engagement, in addition to indices of academic preparation.

Second, how to measure parental involvement and investment? Most American literature focuses on parental involvement in high school, but fails to notice the vital role of private tutoring and disciplinary competition supported by parents, which are prevalent in China. These measures are included in this paper to evaluate parental support outside high school.

figure 1

Finally, there is no consensus on how to measure college success. Existing literature predominately puts emphasis on academic achievement, retention and timely degree completion, not much focus on skills formation and moral development. We consider all three indices of college performance in this study.

3 Analytical Framework, Empirical Model, and Data

3.1 Framework

In keeping with previous literature, this paper constructs a simple model to test the association between socioeconomic status, high school preparation, and college development (see Figure 1).

This model hypothesizes that one’s socioeconomic status influences her high school experience, i.e., advantaged students tend to be better prepared in high school, be more engaged in learning, and their parents tend to provide more support. Next, high school experience is assumed to predict college achievement. Students with better high school preparation are assumed to have higher grades and accumulate more skills in college. In addition, this paper hypothesizes socioeconomic status has direct effects on college development.

In this model, individual high school experience is decomposed into three parts. Academic preparation is a composite measure based on indicators for whether students attended selective high school, participated in disciplinary competition at province or national level, and attended high schools using innovative teaching pedagogy. Learning engagement evaluates the quality of student’s study effort and teacher-student interaction, which is built on measures of daily study time, student’s interaction with teachers within and outside classroom. Parental support is a cumulative measure of parental involvement. This paper defines it as a function of household expenditure for private tutoring and parental involvement in college institution and major choice.

We also construct a socioeconomic status index based on father’s education, father’s occupation status, and household income. 29 College academic achievement is defined as the first-year college major ranking. Skills formation is reflected by college freshmen’s core competence and civic participation.

3.2 Empirical Model

In line with the economics of education literature, this study applies an education production function, which intends to identify the correlation between educational inputs and outputs, capturing the cumulative effects of family, school, and social inputs for education. 30 In this paper, we intend to estimate the effect of ses and high school preparation on college development.

A_t = α₀ + α₁SES_t-1 + α₂A_t-1 + α₃SES_t-1 • A_t-1 + ε (1)

Where A_t refers to one’s college major ranking, core competence or civic participation. A_t-1 stands for one’s academic preparation, learning engagement, and parental support in high school. SES_t-1 refers to student’s socioeconomic status in high school. The interaction term captures the potential intermediate effect of ses on the correlation between high school preparation and college achievement. ε is a random error term.

This paper runs a path analysis with the Structural Equation Modeling in stata. The estimation model is illustrated in Figure 2. In this model, we assume that individual socioeconomic background influences one’s high school experience, but it also has an independent effect on one’s college performance. Academic preparation in high school can improve a college student’s major ranking, core competence and civic participation; learning engagement may affect the two measures of skills formation; and parental support only associates with one’s first-year major ranking. We also hypothesize that the error terms for three dependent variables are highly correlated.

figure 2

3.3 Data and Sample

The data is derived from a longitudinal survey of college students from 28 colleges and universities in Beijing, sponsored by Peking University and the Beijing Education Working Commission. The College Freshmen Development Survey surveyed 5000 freshmen in the Spring of 2011. The following up sophomore and junior surveys were conducted in the Spring of 2012 and 2013. This study is based on the freshmen survey, which collected information on students’ family background, high school experience, college application, academic and social engagement in college, evaluation of institutions’ teaching and support, and skill development.

Among 5000 surveyed students, 4244 returned complete surveys, an 85% response rate. Among them, 81% enrolled in 4-year universities and 19% from 3-year vocational colleges. 44% were male and 56% were female students. About 30% were graduated from high school in Beijing. The average years of father’s schooling are 11.34 years, and the median household annual income is 40,000 rmb.

To construct the sample, we first drop 1678 students lacking household socioeconomic status information. Next, we delete another 403 observations who failed to provide high school information. Finally, we cut another 614 students who did not have accurate college information. The final sample includes 1,549 college freshmen.

4 How High School Experiences Affect College Development?

Preliminary analysis confirms the argument that high school experience matters for college development. We find strong evidence supporting the importance of high school academic preparation and learning engagement in nurturing college success, but a weaker role of parental support. Table 1 provides the standardized coefficients from the path analysis using the Structural Equation Modeling. The goodness-of-fit indices show that the model fits the data pretty well. 31

Well prepared and academically engaged high school graduates achieve a significantly higher level of college development than their less prepared peers. From the third column of Table 1, we find that one standard deviation increase in academic preparation corresponds to a 0.056 points increment in major ranking, 0.085 points increment in core competence, and 0.154 points increment in civic participation.

The same is true for learning engagement. One standard deviation increase in learning engagement is associated with 0.064 points increment in core competence, and 0.055 points increment in civic participation. After trying different specifications of the path analysis, we find that parental support only affects college major ranking, and has no significant impact on core competence and civic participation. Thus, the links between parental support and these two measures are dropped from the model.

Why well prepared high school students do better in college? This may due to the fact that academic preparation and learning engagement are good predicators of college readiness, which is critical for smooth high school to college transition and college success. 32 College readiness refers to the level of preparation a student needs in order to enroll and succeed in a credit-bearing general education course at a post-secondary institution that offers a baccalaureate degree. In order to become succeed in college, Conley 33 argued high school students should master key cognitive strategies, key content knowledge, academic behaviors, and contextual skills and awareness, which are the building stones of college readiness.

table 1

Structural equation model for college development

note: the table reports standardized coefficients from the sem models.

In our model, academic preparation measures the quality of high school teaching (being selective high school and using innovative teaching pedagogy) and the cognitive ability of individuals (participated in disciplinary competition at province or national level), while learning engagement measures the quality of student’s study effort and teacher-student interaction. When capable students attend high-quality high schools (i.e. academically prepared), if they work hard and get along with their teachers (i.e. engaged in learning), they are more likely facing academic challenges need to be solved by key cognitive strategies, acquiring more content knowledge, and obtaining self-management skills such as time management, strategic study skills, and awareness of one’s true performance, persistence and the ability to utilize study groups. With these trainings, high-ses students are ready for college success.

5 Does Socioeconomic Status Matters for College Development?

The path analysis reveals that individual socioeconomic status may produce direct and indirect effects on college success. Table 2 decomposes the total effects into direct and indirect effects. We report the standardized coefficient in this table and the Z score indicates the significance of coefficients.

Individual’s family background has direct and significant effects on one’s college performance. On the one hand, high-ses students are better prepared in high school. As indicated by Panel A, socioeconomic status is positively and significantly correlated with one’s academic preparation (the coefficient is 0.25), learning engagement (0.086), and parental support (1.427). On the other hand, high-ses students accumulate more core competence (0.079) and have a higher level of civic participation in college (0.061), but lower major ranking in the first year (–0.067).

table 2

Decomposition of total effects

note: the table reports standardized coefficients from the sem models.

Similarly, high school experiences directly improve one’s college achievement. College freshmen with better academic preparation get higher major ranking (0.055), higher level of core competence (0.068), and higher level of civic participation (0.129). The same is true for the impacts of learning engagement on core competence (0.045), and civic participation (0.041), and the effect of parental support on freshmen major ranking (0.013).

Moreover, socioeconomic status also has indirect effect on college performance through its influence on high school preparation (Panel B of Table 2). We estimate that ses has a positive effect on academic preparation (0.25) and parental support (1.427), while academic preparation and parental support also increase one’s college major ranking (0.055 and 0.013, respectively). Thus, ses’s indirect effect for major ranking is the summation of the products of the two paths: 0.25*0.055+ 1.427*0.013=0.033, as indicated in Panel B. In addition, ses’s indirect effect for core competence is 0.021 (0.25*0.068+0.086*0.045=0.021) and for civic participation is 0.036 (0.25*0.129 +0.086*0.041=0.036).

figure 3

Hence, the total effect of ses is the summation of its direct and indirect effect on college development (Panel C of Table 2). The total effect of ses on first-year major ranking is –0.034 (–0.067+0.033=–0.034); 0.1 (0.079+0.021=0.10) for core competence and 0.097 (0.061+0.036=0.097) for civic engagement.

How to interpret the socioeconomic gap in college development? The life cycle skill formation may provide some clue. It argues that low-ses students cumulate disadvantages over time. 34 They often face hardship in their early environment, which may produce negative impacts on individuals’ cognitive and non-cognitive development in later life.

For instance, our descriptive analysis shows clear socioeconomic performance gaps at both high school and college level (see Figure 3). Students from the bottom ses quartile have less parental support, are less engaged in learning, and are less prepared academically in high school. Although they have similar first-year major rankings as their high-ses peers, they have a lower level of core competence and civic participation, both are desired outcomes in college.

Based on our findings, we argue that the substantial contrast in college performance by socioeconomic status may be explained by two factors—a variation in academic preparation and learning engagement within high school and a difference in parental support outside school. First, disadvantaged students were less prepared and engaged in high school as indicated by Table 1. Because academic preparation and learning engagement are key indicators for college readiness, low-ses students are possibly not ready for college at time of matriculation. Their chances of being success in college are therefore diminished.

By comparison, high-ses students are well prepared in high school with key cognitive strategy, content knowledge and appropriate academic behaviors. With these advanced training, they are able to understand what is expected in a college course, can cope with the content knowledge that is presented and can develop as a result of the course the key intellectual lessons and dispositions the course was designed to convey.

Second, parents in low-ses households were not able to provide sufficient support for their children as the high-ses parents did. In our sample, the top 10% ses group spends on average 8,095 rmb for private tutoring in senior year, while the bottom 10% ses group only spends 1,433 rmb. Early studies also find rural and low-income students are less likely receive private tutoring in high school, 35 and their parents are less involved in their college application and preparation. In brief, lower level of support from parents in high school may become barriers for college success for low-ses students.

6 Conclusion

6.1 Low-Level Development Cycle

Chinese higher education expansion in the late 1990s opens college access and drives the gross enrollment rate from 9.8 percent in 1998 to 30 percent in 2012. Many female, minority, first-generation, rural students from low-income families are thus able to receive public higher education for the first time. In 2011, 88% of the vocational college graduates were first-generation college students, 12.7% came from low-income areas, and 16.2% came from ethnic minority concentration regions. 36

Unfortunately, the increasing college access fails to level the playground at the finishing line. High-ses college students tend to have better ranking and gpa, be more likely to attend graduate school, and have higher earnings after graduation. For instance, graduates with one cadre parent earn 15% more than others. 37

Why students from wealthier, urban households with educated parents tend to do better in college, even after the rigorous ability sorting through college application and admission? Our study is one of the first attempts to disentangle this puzzle and provide empirical tests based on a large scale college freshmen survey.

Our findings show that, first, there is a positive and significant correlation between individual’s high school academic preparation, learning engagement, parental support and college performance. Well prepared high school graduates are more likely to be college ready, which is critical for adapting to campus life and becoming successful academically. Next, socioeconomic status has both direct and indirect effects on college achievement. Particularly, socioeconomic status indirectly influences college major ranking and skill formation through its impact on high school preparation and engagement.

In conclusion, this study provides a critical evidence for the life cycle skill formation argument, i.e., the deficit in early human capital investment can impair adolescent and adult attainment. In China, housing price reflects the quality of local public schools and low-income families cannot afford to live in communities with high quality public education. 38 Children from low-ses households are less likely to attend selective high schools due to residential segregation. When they enter mediocre or underperformed high schools, they are not likely meet academic challenges necessary for developing cognitive strategy, mastering high-level content knowledge, acquiring self-management skills, and adopting contextual skills and awareness.

Even though some disadvantaged students can pass the college entrance examination and are admitted to postsecondary institutions, they are not college ready. They may have a higher probability of being maladapted. These adaption issues in freshmen year may “cool out” low-ses students’ academic aspiration and they will become less engaged in learning and campus life. Low level of academic and social engagement can lead to course failure and marginalization on campus, significant predictors for dropout, low achievement and low satisfaction. These disengaged and frustrated students will acquire inadequate human capital and are less likely to be ready for the labor market. Thus, they are trapped in a low-level development cycle.

6.2 Potential Success Strategies

The most important implication of our study is that educational policymakers can help low-ses students to succeed in college, if their help comes early enough and target at the right group.

Obviously, in order to improve college performance, we should level the playground in high school. Some high school interventions in China have focused on the college preparation and application stage. For instance, several researchers have conducted random experiments in rural high schools, using information intervention and counseling to leverage low-income high school graduates’ college expectation. 39 These studies find that information does not have significant effects on student outcomes. Others try to provide early college financial aid commitment to rural high school students, in order to affect the college decision making of poor students in rural China. 40 The authors conclude that if early commitments are made early enough; and they are large enough, students will make less distorting college decisions.

The other option is to help disadvantaged students enter high-quality high schools. One proposal is to reform the current high school recruitment system, by randomly assigning a proportion of slots in selective high schools to graduates from non-selective middle schools. Beijing has planned to implement similar policies in the near future. 41 Earlier policy evaluation finds such reforms may increases educational equity and efficiency, in terms of opening quality high school education to low-ses students and improving overall college enrollment rate in the reformed region. 42

In addition to high school intervention, colleges can also play important roles in helping their low-ses students, by targeting right students and providing right services. Ford Foundation has recently sponsored four colleges and universities in southwestern China to implement a project, titled “Improving Disadvantaged Freshmen’s College Adjustment through Enhancing Student Engagement”. The program intends to help poor freshmen build social network with peers and administrators, acquire necessary learning strategies, and immerse to diverse campus activities. Preliminary analysis shows that the low-income students in the treatment group are less likely to be maladapted to campus life and have better academic achievement than their peers in the control group. 43

In summary, more studies are needed to identify the effective high school and college interventions for promoting college success of disadvantaged students.

Footnotes

* Yang Po is an associate professor at the Department of Economics of Education and Administration, Graduate School of Education, Peking University, China. She can be reached at Graduate School of Education, Box 68, Peking University, China, 100871. This study is sponsored by the China National Science Foundation (Project number 71103008).

1 Yang Po, “Who gets more financial aid in China? A multilevel analysis,” International Journal of Educational Development 30, no. 6 (2010): 560-569. Li, Hongbin, et al. “Does having a cadre parent pay? Evidence from the first job offers of Chinese college graduates,” Journal of Development Economics 99, No. 2 (): 513-520.

2 Bao Wei 鲍威, “Diyidai nongcun daxuesheng de shengxue xuanze 第一代农村大学生的升学选择 [College Choice of the First-generation College Students in China],” Jiaoyu Xueshu Yuekan 教育学术月刊 [Education Research Monthly], , 1:3-12.

3 Yang Po, Mao Dan 杨钋, 毛丹, “Shiying: Daxue xinsheng fazhan de guangjianci “适应” 大学新生发展的关键词——基于首都高校学生发展调查的实证分析 [Adaptation: The Key Word for College Freshmen Development],” Zhongguo gaojiao yanjiu 中国高教研究 [China Higher Education Research], (3): 16-24.

4 Bao Wei, Li Binglong 鲍威, 李炳龙. “Shui shiye, shui jiuye, shui shengxue? 谁失业, 谁就业, 谁升学——2003 年-2009 年中国高校学生毕业后发展路径选择的观察 [Who Gets Unemployed, Employed, or Admitted to the Graduate School?],” Qinghua daxue jiaoyu yanjiu 清华大学教育研究 [Tsinghua Journal of Education] 33, No. 1 (): 72-80.

5 L. Hongbin, P. Loyalka, S. Rozelle, et al. “Unequal Access to College in China: How Far Have Poor, Rural Students Been Left Behind?” SSRN Working Paper, No. 2293344, 2013. .

6 James T. Heckman, “Skill formation and the economics of investing in disadvantaged children,” Science 312, no. 5782 (2006): 1900-1902.

7 G.D. Kuh, J. Kinzie, J.A. Buckley, et al. “What matters to student success: A review of the literature,” Commissioned Report for the National Symposium on Postsecondary Student Success: Spearheading a Dialog on Student Success, 2006.

8 David T. Conley, “Rethinking college readiness,” New Directions for Higher Education 2008, no. 144 (): 3-13.

9 Sina News, “Shushuo gaokao 数说高考：不足2%考生能考上 985 高校 [Discussing College Entrance Examination with Data], accessed February 27, 2014,

10 Less selective public 4-year institutions belong to the second tier, non-selective public or private 4-year colleges occupy the third tier and vocational 3-year private or public colleges comprise the fourth tier. Prashant Loyalka, Yingquan Song, and Jianguo Wei. “The effects of attending selective college tiers in China,” Social science research 41, no. 2 (2012): 287-305.

11 Zhang, Yu. “Does private tutoring improve students’ National College Entrance Exam performance?—A case study from Jinan, China,” Economics of Education Review 32, (2013): 1-28.

12 Prashant Loyalka, Yingquan Song, and Jianguo Wei. “Zhongguo xianxing gaoxiao xuesheng zizhu zhengce pinggu 中国现行高校学生资助政策评估 [Evaluation of Current Chinese College Financial Aid Policy]. Beijingdaxue Jiaoyu Pinglun 北京大学教育评论 [Peking University Education Review], 2011, 9(1): 68-79.

13 Mark Bray, et al. “Differentiated demand for private supplementary tutoring: Patterns and implications in Hong Kong secondary education,” Economics of Education Review 38 (2014): 24-37.

14 Lei Lei 雷磊, “Hengshui Zhongxue de Fengshen zhi Lu 衡水中学的“封神”之路—超级高考工厂,” accessed on February 27, 2014, .

15 Pan Xiaoling 潘晓凌, “Qionghaizi Meiyou Chuntian 穷孩子没有春天？——寒门子弟为何离一线高校越来越远,” accessed on February 27, 2014 .

16 Laura W. Perna, “Improving College Access, Persistence, and Completion: Lessons Learned,” The State of College Access and Completion: Improving College Success for Students from Underrepresented Groups (2013): 208. L.W. Perna, S.L. Thomas, “A framework for reducing the college success gap and promoting success for all,” National Symposium on Postsecondary Student Success: Spearheading a Dialog on Student Success, 2006.

17 Michael B. Paulsen, and Edward P. St John. “Social class and college costs: Examining the financial nexus between college choice and persistence,” The Journal of Higher Education 73, no. 2 (2002): 189-236.

18 Vincent Tinto, and Brian Pusser. “Moving from theory to action: Building a model of institutional action for student success,” National Postsecondary Education Cooperative (2006): 1-51.

19 Gary P. Pike, and Joseph L. Saupe, “Does high school matter? An analysis of three methods of predicting first-year grades,” Research in higher education 43, no. 2 (2002): 187-207.

20 Clifford, Adleman, “Answers in the tool box: Academic intensity, attendance patterns, and Bachelor’s degree attainment,” Office of Educational Research and Improvement, u.s. Department of Education (1999).

21 George D. Kuh, Piecing Together the Student Success Puzzle: Research, Propositions, and Recommendations: ashe Higher Education Report. Vol. 116 (John Wiley & Sons, 2011).

22 Laura H. Horn, and Anne-Marie Nuñez, Mapping the road to college first-generation students’ math track, planning strategies, and context of support (diane Publishing, 2000).

23 Edward C. Warburton, Rosio Bugarin, and Anne-Marie Nunez, “Bridging the Gap: Academic Preparation and Postsecondary Success of First-Generation Students,” Education Statistics Quarterly 3.3 (2001): 73-77.

24 Laura Walter Perna, and Marvin A. Titus, “The relationship between parental involvement as social capital and college enrollment: An examination of racial/ethnic group differences,” Journal of Higher Education 76, no. 5 (2005): 485-518.

25 Walter Scott Swail, et al., Latino students and the educational pipeline. Part III: Pathways to the bachelor’s degree for Latino students (Washington, dc: Educational Policy Institute and Lumina Foundation for Education, 2005).

26 Janet H. Chrispeels, and Elvia Rivero, “Engaging Latino families for student success: How parent education can reshape parents’ sense of place in the education of their children,” Peabody Journal of Education 76.2 (2001): 119-169.

27 Barbara J. Bank, Ricky L. Slavings, and Bruce J. Biddle, “Effects of peer, faculty, and parental influences on students’ persistence,” Sociology of education 63, no. 3 (1990): 208-225.

28 Bao Wei 鲍威, “Diyidai noncun daxuesheng de shengxue xuanze 第一代农村大学生的升学选择 [College Choice of the First-generation College Students in China],” Jiaoyu Xueshu Yuekan 教育学术月刊 [Education Research Monthly], , 1:3-12.

29 Blau, Peter M., and Otis Dudley Duncan. The American occupational structure. New York: Wiley ().

30 D. Harris, “Education production functions: Concepts,” in Economics of Education, edited by D.J. Brewer, and P.J. McEwan (Amsterdam: Elsevier, 2010), 127-131. Jesse Rothstein, “Teacher quality in educational production: Tracking, decay, and student achievement,” The Quarterly Journal of Economics 125, no. 1 (2010): 175-214.

31 The chi-square from the likelihood ratio test indicates that we can reject at the 5% level that the model fits as well as the saturated model (P value for Chi-square is 0.00). When we look at the size of residuals, the standardized root mean squared residual (srmr) is 0.055, which is smaller than the threshold of a good fit at 0.08. Stata Press, “Structural Equation Modelling Reference Manual,” accessed on February 7, 2014: .

32 George L. Wimberly and Richard J. Noeth, “College Readiness Begins in Middle School. ACT Policy Report,” American College Testing ACT Inc. (2005).

33 David T. Conley, “Rethinking college readiness,” New Directions for Higher Education 2008.144 (): 3-13.

34 James J. Heckman, “Skill formation and the economics of investing in disadvantaged children,” Science 312.5782 (2006): 1900-1902.

35 Zhang, Yu. “Does private tutoring improve students’ National College Entrance Exam performance?—A case study from Jinan, China,” Economics of Education Review 32, (2013): 1-28.

36 Wang Congcong 王聪聪, “88.1% 2011 Gaozhi Biyesheng wei Jiating Diyidai Daxues 88.1% 2011 届高职毕业生为家庭第一代大学生,” accessed on February 7, 2014, .

37 Li, Hongbin, et al., “Does having a cadre parent pay? Evidence from the first job offers of Chinese college graduates,” Journal of Development Economics 99.2 (): 513-520.

38 Wen, Haizhen, Yan Zhang, and Ling Zhang, “Do educational facilities affect housing price? An empirical study in Hangzhou, China,” Habitat International 42 (2014): 155-163.

39 Prashant Loyalka, et al., “Can information and counseling help students from poor rural areas go to high school? Evidence from China,” Journal of Comparative Economics 41, no. 4 (2013): 1012-1025.

40 Liu, Chengfang, et al., “Early commitment on financial aid and college decision making of poor students: Evidence from a randomized evaluation in China,” Economics of Education Review 30, no. 4 (2011): 575-774.

41 “Beijing Zhongzhao: Youzhi Gaozhong 30% Minge Fengei Benqu Chuzhongxiao 北京中招：优质高中 30% 名额分给本区初中校,” accessed on February 2, 2014, .

42 Pan Kunfeng, Li Yang 潘昆峰, 李扬. “Gaozhong Jiaoyu Ziyuan Junheng Peizhi de Xiaoguo Jiqi Donglixue Fenxi 高中教育资源均衡配置的效果及其动力学分析 [Balanced Allocation of Educational Resources between High Schools: Effects and Its Dynamical Analysis]. Tsinghua Journal of Education 清华大学教育研究, , 5:14-22.

43 Yang, P. & Li, F. Program Summary Report. Working paper series, 201304, China Center for Education Demand Study, Southwestern Finance University.

References

Bank

Barbara J.

Slavings

Ricky L.

Biddle

Bruce J.

“Effects of peer, faculty, and parental influences on students’ persistence”

Sociology of education 1990 63 no. 3 208 225

Bao

Wei

“Diyidai nongcun daxuesheng de shengxue xuanze 第一代农村大学生的升学选择 [College Choice of the First-generation College Students in China]”

Jiaoyu Xueshu Yuekan 教育学术月刊 Education Research Monthly 2013 1 3 12

Bao

Wei

Binglong

“Shui shiye, shui jiuye, shui shengxue? 谁失业, 谁就业, 谁升学——2003 年-2009 年中国高校学生毕业后发展路径选择的观察 [Who Gets Unemployed, Employed, or Admitted to the Graduate School?]”

Qinghua daxue jiaoyu yanjiu 清华大学教育研究 Tsinghua Journal of Education 2012 33 No. 1 72 80

Blau

Peter M.

Duncan

Otis Dudley

The American occupational structure 1967

New York

Wiley

Bray

Mark

“Differentiated demand for private supplementary tutoring: Patterns and implications in Hong Kong secondary education”

Economics of Education Review 2014 38 24 37

Chrispeels

Janet H.

Rivero

Elvia

“Engaging Latino families for student success: How parent education can reshape parents’ sense of place in the education of their children”

Peabody Journal of Education 2001 76 2 119 169

Clifford

Adleman

“Answers in the tool box: Academic intensity, attendance patterns, and Bachelor’s degree attainment”

1999 Office of Educational Research and Improvement, u.s. Department of Education

Conley

David T.

“Rethinking college readiness”

New Directions for Higher Education 2008 no. 144 3 13 2008

Harris

David

Brewer

D. J.

McEwan

P. J.

“Education production functions: Concepts”

Economics of Education 2010

Amsterdam

Elsevier

10.

H. Horn

Laura

Nuñez

Anne-Marie

Mapping the road to college first-generation students’ math track, planning strategies, and context of support. 2000

DIANE Publishing

11.

Heckman

James J.

“Skill formation and the economics of investing in disadvantaged children”

Science 2006 312 no. 5782

12.

Hongbin

Loyalka

Rozelle

“Unequal Access to College in China: How Far Have Poor, Rural Students Been Left Behind?”

ssrn Working Paper, No. 2293344, 2013. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2293344

13.

Kuh

G. D.

Kinzie

Buckley

J. A.

“What matters to student success: A review of the literature”

2006 Commissioned Report for the National Symposium on Postsecondary Student Success: Spearheading a Dialog on Student Success

14.

Kuh

George D.

“Piecing Together the Student Success Puzzle: Research, Propositions, and Recommendations”

ASHE Higher Education Report 2011 Vol. 116

John Wiley & Sons

15.

Lei

“Hengshui Zhongxue de Fengshen zhi Lu 衡水中学的‘封神’之路”

accessed on February 27, 2014

http://www.infzm.com/content/94916

16.

Hongbin

“Does having a cadre parent pay? Evidence from the first job offers of Chinese college graduates”

Journal of Development Economics 2012 99 No. 2

17.

Liu

Chengfang

“Early commitment on financial aid and college decision making of poor students: Evidence from a randomized evaluation in China”

Economics of Education Review 2011 30 no. 4 575 774

18.

Loyalka

Prashant

“Can information and counseling help students from poor rural areas go to high school? Evidence from China”

Journal of Comparative Economics 2013 41 no. 4 1012 1025

19.

Loyalka

Prashant

Song

Yingquan

Wei

Jianguo

“The effects of attending selective college tiers in China”

Social science research 2012 41 no. 2 287 305

20.

Loyalka

Prashant

Song

Yingquan

Wei

Jianguo

“Zhongguo xianxing gaoxiao xuesheng zizhu zhengce pinggu 中国现行高校学生资助政策评估 [Evaluation of Current Chinese College Financial Aid Policy]”

Beijingdaxue Jiaoyu Pinglun 北京大学教育评论 Peking University Education Review 2011 9 1 68 79

21.

Pan

Kunfeng

Yang

“Gaozhong Jiaoyu Ziyuan Junheng Peizhi de Xiaoguo Jiqi Donglixue Fenxi 高中教育资源均衡配置的效果及其动力学分析 [Balanced Allocation of Educational Resources between High Schools: Effects and Its Dynamical Analysis]”

Tsinghua Journal of Education 清华大学教育研究 2010 5

22.

Pan

Xiaoling

“Qionghaizi Meiyou Chuntian 穷孩子没有春天？寒门子弟为何离一线高校越来越远”

accessed on February 7, 2014

http://www.infzm.com/content/61888

23.

Paulsen

Michael B.

St John

Edward P.

“Social class and college costs: Examining the financial nexus between college choice and persistence”

The Journal of Higher Education 2002 73 no. 2 189 236

24.

Perna

Laura W.

“Improving College Access, Persistence, and Completion: Lessons Learned”

2013 The State of College Access and Completion: Improving College Success for Students from Underrepresented Groups

25.

Perna

Laura W.

Titus

Marvin A.

“The relationship between parental involvement as social capital and college enrollment: An examination of racial/ethnic group differences”

Journal of Higher Education 2005 76 no. 5 485 518

26.

Perna

Laura W.

Thomas S.

“A framework for reducing the college success gap and promoting success for all”

2006 National Symposium on Postsecondary Student Success: Spearheading a Dialog on Student Success

27.

Pike

Gary R.

Saupe

Joseph L.

“Does high school matter? An analysis of three methods of predicting first-year grades”

Research in higher education 2002 43 no. 2 187 207

28.

Rothstein

Jesse

“Teacher quality in educational production: Tracking, decay, and student achievement”

The Quarterly Journal of Economics 2010 125 no. 1 175 214

29.

Stata Press

“Structural Equation Modelling Reference Manual”

Accessed on February 7, 2014

http://www.stata.com/bookstore/structural-equation-modeling-reference-manual/

30.

Swail

Walter Scott

“Latino students and the educational pipeline. Part III: Pathways to the bachelor’s degree for Latino students”

2005

Washington, DC

Educational Policy Institute and Lumina Foundation for Education

31.

Tinto

Vincent

Pusser

Brian

“Moving from theory to action: Building a model of institutional action for student success”

National Postsecondary Education Cooperative 2006 1 51

32.

Wang

Congcong

“88.1% 2011 Gaozhi Biyesheng wei Jiating Diyidai Daxues 88.1% 2011 届高职毕业生为家庭第一代大学生”

Accessed on February 7, 2014

http://zqb.cyol.com/html/2013-01/17/nw.D110000zgqnb_20130117_1-07.htm

33.

Warburton

Edward C.

Bugarin

Rosio

Nunez

Anne-Marie

“Bridging the Gap: Academic Preparation and Postsecondary Success of First-Generation Students”

Education Statistics Quarterly 2001 3 3 73 77

34.

Wen

Haizhen

Zhang

Yan

Zhang

Ling

“Do educational facilities affect housing price? An empirical study in Hangzhou, China”

Habitat International 2014 42 155 163

35.

Wimberly

George L.

Noeth

Richard J.

“College Readiness Begins in Middle School. ACT Policy Report”

2005 American College Testing act Inc

36.

Yang

“Who gets more financial aid in China? A multilevel analysis”

International Journal of Educational Development 2010 30 no. 6 560 569

37.

Yang

Mao

Dan

“Shiying: Daxue xinsheng fazhan de guangjianci “适应”大学新生发展的关键词——基于首都高校学生发展调查的实证分析 [Adaptation: The Key Word for College Freshmen Development]”

Zhongguo gaojiao yanjiu 中国高教研究 China Higher Education Research 2013 3

38.

Yang

Li Feng

Program Summary Report

Working paper series, 201304 China Center for Education Demand Study, Southwestern Finance University

39.

Zhang

“Does private tutoring improve students’ National College Entrance Exam performance?—A case study from Jinan, China”

Economics of Education Review 2013 32 1 28

Who Will Success in College? Effect of High School Experience and Socioeconomic Status

Abstract

Abstract

Keywords

1 Introduction

2 The Intertwined Experience in High School and College

2.1 Chinese Context

2.2 Role of High School Preparation for College Success

2.3 Unresolved Issues

figure 1

3 Analytical Framework, Empirical Model, and Data

3.1 Framework

3.2 Empirical Model

figure 2

3.3 Data and Sample

4 How High School Experiences Affect College Development?

Structural equation model for college development

5 Does Socioeconomic Status Matters for College Development?

Decomposition of total effects

figure 3

6 Conclusion

6.1 Low-Level Development Cycle

6.2 Potential Success Strategies

Footnotes

References