China College Student Survey ( ccss ): Breaking Open the Black Box of the Process of Learning

Abstract

Abstract

The paper uses macro-causal analysis combined with empirical data analysis to illustrate how Chinese scholars in higher education (he) have adapted the concept of “student engagement” and indigenize the instrument of the National Survey of Student Engagement (nsse), with the aim of revealing college teaching/learning activities and understanding the learning processes of Chinese students in he. The paper explains why and how the China College Student Survey (ccss) was developed by the research team in the Institute of Education (ioe), Tsinghua University, using the nsse as a basis, with the help of both domestic and foreign scholars. The paper analyzes data collected nationwide in 2012, under the framework of four trends influencing he in China. Cases are also studied in the paper illustrating the learning behaviors of college students, as well as the macro-level driving forces in a more general context.

As we become an increasingly knowledge-based society, the competition of mental capability and “soft power” of which higher education is highly representative becomes a crucial issue, especially in China. In recent years, China is approaching a key stage of social reform: its he has gone through rapid expansion with the total student number increasing from 600 million in 1998 to 3100 million in 2012. The core concern now for the government, society and he institutions has shifted from quantitative expansion to quality improvement. Student engagement in learning activities and the outcomes of their college lives have become important research issues.

1 Research Background: Four Trends and the Changing Context of he in China

Currently higher education in China is undergoing reforms in which the driving forces are mainly due to a rapidly changing society, with four trends particularly relevant to he: globalization, marketization, massification, and reforms in basic education.

Globalization in a knowledge-based society is a global phenomenon with wide-ranging effects on many different parts of the world 1 —he is particularly sensitive to the pressures of globalization and the fundamental driver in the worldwide knowledge-driven growth model, which has the effect of not only widening the existing gaps among individuals, regions and nation-states, but also making he crucial to the development of individuals, regions and nations. 2 Mainland China, despite coming onto the scene of modernization relatively late, nonetheless successfully captured the opportunities afforded by the open-door policy (dating from 1978), reoriented its goals of economic and social development in comparison with its international counterparts, and became a prominent member of the world community. This policy provided China with a much larger platform that today not only equips Chinese scholars with a wide global vision, but also supports them with international academic networks. The he system, as a knowledge-intensive sector in the current global era, is generally recognized as the main tool to cultivate future generations and the quality of he which is partly recognized by the process of student learning behaviors that can be captured and analyzed. 3 The Chinese College Student Survey (ccss) began as an adaptation and indigenization of the nsse. It represents the efforts of Chinese scholars in the general context of globalization, with the help of international academic networks, to measure the college student learning process and improving the quality of he.

Marketization represents the movement of the ideology and rules of free market from the economy to he, bringing profound changes in the resource allocation, educational environment and policy prioritization of both the government and higher education institutions (heis). With a view to make existing universities world-class institutions, the Chinese government initiated the “key construction projects” in late 20th century, namely the “985” and “211” Projects, providing the top universities and programs with abundant resources. A joint effort on the part of the government, the heis and society itself is necessary to achieve this goal of increasing the quality of he in China. Parallel to the economic reform, there is a shift from the traditional “industrial or plan-based model” to a new “flexibility or need-based model,” 4 with an increasing awareness of competencies required in the market as well as consumer satisfaction. Despite its reputation as “the last fort of planned economy,” 5 China’s he apparatus has undergone a systematic reform and structural reorganization which gradually influenced the changes in teaching/learning practices and pedagogical methods.

Massification of he in China has occurred at an astounding rate, possibly the highest in the world, despite limited resources. 6 The total number enrolled in he increased more than five times from 1998 to 2010 and with a total number of 31 million students studying in various heis in 2012. The expansion of he in China is mainly driven by external forces and follows an up-down pattern, representing a “perfect story,” as Altbach said, of “external pressures and internal responses.” 7 In recent years, there is a consensus in Chinese academia that the rapid expansion has resulted in a decline of overall quality across the system, where part of the quality problem is caused by different parties providing varying and sometimes contradictory criteria to assess the quality. 8

With a large number of new recruits, especially first-generation college students entering the campuses with diverse needs, from different backgrounds, being used to contrasting social norms and armed with divergent views on education, there arose increasing tensions, or “institutionalized distraction” in Trow’s words. 9 An obvious tension is that as traditional universities no longer have a monopoly in the tertiary education sector, the concept of excellence is challenged, so that the quality value rooted in the system of meritocracy has to be redefined. 10 Therefore, a few questions needs to be further studied, such as what massification means in China and how it affects the campus atmosphere. The changes of interpersonal relationships and student learning behaviors are also issues to be researched in detail.

The reform in basic education in China, in addition to he, has been part of the world-wide paradigm shift from being centered on teaching to being centered on student learning, and from being input-focused to process-and outcome-focused; there is also a shift from a static or one-point evaluation to a multi-dimensional and value-added assessment. 11 New trials in basic education, such as an integrated curriculum, a focus on problem-solving skills, and cooperative teamwork, etc. have added value to the traditional model of Chinese education, which is well known for its high expectations, timely practices and devoted, hardworking teachers using positive guidance and incentives, to create a better climate for the children to fulfill their potential. Whilst Shanghai’s outstanding performance in pisa 2009 and 2012 vividly “defied preconceptions about China’s education system being based on rote learning,” 12 Shanghai’s standing at the forefront of basic education reform, is strongly challenging the he system which has yet to undergo a similar transformation. Recently, China’s Ministry of Education released plans to reform the rigorous national college entrance exam, which Chinese universities used to rely on heavily for student admission. This clearly shows that China is moving towards a deeper reform in education which will profoundly influence the whole system, in particular he.

Figure 1 depicts the framework for macro background analysis. In the framework, globalization plays a role in stimulating the internal efforts of internationalization; it presents a new perspective and creates new academic resources/networks to work with. It greatly revitalizes Chinese academia. Massification brings diversity and changes to the ecological structures of he. The paper will take first-generation college students as an example to illustrate the phenomenon. Marketization is no doubt an important factor to he, yet its influence is usually indirect, filtered through public/private considerations, cost/benefit analysis, and increasing concern over the students’ learning outcomes and satisfaction. The most notable effect from the basic education reform to he is the recruitment, but more profound than that is a new generation getting accepted into the campus, with their preference and habits of learning framed by the reformed basic education, and overwhelming new technologies.

figure 1

Framework for macro background analysis

Generally speaking the four trends representing external forces, along with numerous ancillary factors, lie at the center of any consideration of he policy, reform and development. 13 The influence of these factors on Chinese education may be observed through policy interventions and resource allocation. The changing beliefs in education, reinforced by cultural conformity, are gradually becoming a part of the familiar ecological atmosphere.

2 Major Works in Constructing the ccss

The ccss was designed, both as a tool and a project, to help stimulate evidence-focused dialogue regarding students’ learning behaviors, representing by their engagement in university study, and provide institutions with information that they can use to monitor the teaching/learning process and enhance the quality of education they provide to their students. The ccss was first called the nsse-China which was first launched by the Institute of Education (ioe) in Tsinghua University together with American scholars from the nsse team of University of Indiana in 2007, and was later developed into the indigenized China College Student Survey (ccss).

27 heis volunteered to participate in the ccss in 2009. This number rose to 47 in 2010, 59 in 2011 and 2012, and 73 in 2013. The ccss was becoming a nationwide, cross disciplinary and institutional project with the functions of assessing, diagnosing and improving teaching/learning processes in he.

Sharing the same theoretical foundations with the nsse, the ccss was also developed based on the core concept of ‘student engagement’ and other relevant pedagogical theories. Our understanding of the term ‘student engagement’ is based on the definitions found in the literature and discussion of the character of engagement and its alternatives:

Student engagement is concerned with the interaction between the time, effort and other relevant resources invested by both students and their institutions intended to optimize the student experience and enhance the learning outcomes and development of students and the performance, and reputation of the institution. 14

This term has its historic roots in a body of work concerned with student involvement, mainly in North America and Australia, where it has been firmly entrenched through annual, large-scale national surveys (nsse in the u.s. 15 ^, 16 and ausse in Australia.) 17 ^, 18 We also noticed that a body of work on student engagement has been produced in the uk (although rarely labeled as ‘student engagement’), which traces its roots back to other traditions such as student feedback, student representation, student approaches to learning, as well as institutional organization and learning development. 19 ^, 20

Although ccss was developed from the nsse, the Tsinghua team has made great efforts to make it suitable within the Chinese educational context and help break open the black box of the learning process of Chinese college students. There are four innovations found in the ccss.

2.1 Adding Psychological Variables to Deeper the Understanding of Behaviors

Engagement is more than involvement or participation. 21 Acting without feeling engaged is simply involvement or even compliance, while feeling engaged without acting is dissociation. Fredricks et al. 22 and Bloom 23 describe three dimensions of engagement: behavioral engagement, emotional engagement and cognitive engagement. Results from our data analysis and interviews show that Chinese learners’ engagement and the learning outcomes are largely influenced by their motivation, their feeling about the learning process, and their conceptions of learning. Therefore, the ccss instrument has developed three important psychological dimensions of engagement, namely emotional engagement, motivational engagement and cognitive/perceptional engagement. 24

Emotional engagement: Students who engage emotionally would experience affective reactions such as interest, enjoyment, or a sense of belonging. 25 Emotional engagement is reflected in four ways: one’s interest (or lack thereof) in one’s major, appreciate its usefulness, enjoy the learning process and feel happy while studying.

Motivational engagement: Student motivation is a key component of student learning, stimulating students to become engaged in certain learning activities. Students who have intrinsic motivation are more engaged behavior-wise and produce better results, while those who have extrinsic motivation are less engaged with poorer results. 26 This dimension of engagement is characterized as intrinsic motivation (interest in knowledge itself, self-development and challenging oneself) versus extrinsic motivation (satisfying teachers and parents, gaining high scores).

Cognitive engagement: Students’ conceptions of learning and of their educational situation have a large influence on their behavioral engagement. Students who see learning as a process of self-improvement are generally more invested in their learning, seek to go beyond the minimum requirements, and relish challenges more than those who see learning as a mere tool to increase knowledge. 27 Five levels of conception of learning have been identified: 1) learning to increase knowledge through memorization; 2) learning to help understand the world; 3) learning as a process of discovery and exploration; 4) learning as a process of vocational training; 5) learning as a process of growth into a fully developed person.

2.2 Developing the Diagnostic Function in the Chinese Context

To enhance the diagnostic function of the instrument in the Chinese educational context and to help Chinese universities improve education quality, four constructs have been developed as follows:

1. Course cognitive goals (memorize, apply, analyze, evaluate, synthesize)

2. Level of rigor of teachers’ academic requirements (number of writing and reading tasks, difficulty of exams, etc.)

3. Engagement in class (active learning behaviors in class)

4. Engagement outside of class (active learning behaviors outside of class)

The constructs agree well with the Chinese higher education pedagogy and emphasizes the operational factors of the teaching process. 28 It is reported by many project universities that the four constructs are better used in diagnosing the problems and improving the teaching and supervision of students. 29 ^, 30

2.3 Adapting the Instrument for Vocational heis

To meet the needs of the rapid development of vocational education, a field rarely researched in China by academia, we have developed an instrument aimed at evaluating the education quality of vocational colleges. 31 The core framework of the vocational instrument is identical to that of the ccss, with five benchmarks, four constructs, psychological variables and background information. However, the vocational instrument is more skill-oriented, concentrating on the educational practices that promote the development of vocational and transferrable skills.

The vocational instrument is still in its early stages. A dozen vocational colleges over the country piloted the instrument in 2013. The validity and reliability of the instrument will be described in a later report.

2.4 Student Background Matters

Along with the massification and marketization of the Chinese higher education system, students enrolled in heis are becoming increasingly diversified, whilst the dual urban-rural structure in the he system has become even more complicated. 32 The ccss has therefore paid particular attention to the background information of the college students. We have developed a background questionnaire consisting of 35 items, including social/economic status information, pre-college education (including secondary school and high middle school), enrollment information, financial support information, etc. Information regarding the heis themselves (such as school type, location, educational resources provision etc.) is also collected, albeit separately. Typology analysis of heis and college student engagement is conducted by connecting the background information to the student engagement data, with interesting results that to be reported later.

3 Validity, Reliability, and Credibility of the ccss Data 2009-2012

The quality of ccss relies on the validity, reliability and credibility of the instrument. Table 1 shows that the ccss has achieved acceptable criteria validity across the following years according to the regression coefficients and their significance levels. 33 34 35 36 37

Table 2 shows that the five benchmarks achieved a high degree of consistency during the four wave data collection. The lowest Cronbach’s alpha coefficient is eee in 2009, coming in at a still acceptable 0.60. The other benchmarks’ alpha coefficients are all considerable higher, ranging from almost 0.70 to 0.82, which means they are suitable for analysis.

table 1

Criteria validity of five benchmarks on self-reported attainments

*p < 0.05 **p < 0.01 ***p < 0.001

table 2

Summary of five benchmarks’ Cronbach’s alpha coefficients from 2009-2012

In order to test the quality of the items of the ccss in measurement, we examined the reliability of ccss through the Item Response Theory by using the 2010 ccss data. The major findings were that the five benchmarks were suitable for China’s situation, and the reliability of the learning diagnostic scales developed by the Tsinghua team based on the context of Chinese higher education were adequate from the measurement perspective. On the whole, the ccss is an effective and efficient tool for the undergraduate survey. Therefore, the data collected from the survey may aid the institutions in performing a self-diagnosis of the teaching/learning process. 38

4 Basic Description of ccss 2012 Data

The paper uses three types of data collected by the ccss: 1) the observable learning behaviors of college students (i.e., lac, acl, sfi, eee, sce); 2) psychological variables (e.g., motivation, interests of learning), self-reported learning attainments and educational satisfaction; 3) the students’ social and economic status and pre-college educations. Most of the items in the first category were adapted from the nsse after a careful culturally-cognitive revision. 39 Items in categories 2 and 3 were mainly developed by the Tsinghua team. The ccss has achieved reasonable psychometric properties and moderate potential for predicting students’ learning outcomes, according to previous studies. 40 ^, 41

table 3

Reliability of five benchmarks in 2010

The whole sampling process consists of two stages: 1) institutes were chosen by stratum sampling in which the data collected could be inferred nation-wide by proper weighting; 2) students were randomly chosen on the basis of school years with an average of 400 students per grade in one institution.

We distributed 100,644 questionnaires in 2012; of these, 72131 were completed and returned, resulting in a response rate of 71.61%. Student background variables were coded as gender, ethnicity, grade, discipline, first generation college student, rural, only child, region, local student and enrollment type. We categorized heis into three types: “985 project” universities—first tier universities; 211 project universities—second tier universities; local universities—third tier universities. The student’s se Index was based on the parents’ education, occupation, and the conditions of the family’s residential housing. This indicator’s Cronbach’s alpha reaches 0.7064. The student’s pre-college education experience was also represented by a set of variables, including academic ranking (coded into five levels), and high school type (coded into eight levels). The indicator for precollege attainments was conceptualized from 3 items. (Table 1)

The sample included more male students (54.6%) than female students, with a greater percentage majoring in Engineering (46.4%) than the Humanities and Social Sciences (25.1%) or the pure Sciences (7.8%). It revealed that 70.6% of the students are not from the local area, which means that most of the students migrated from another area to the place where their university was located.

5 Further Investigation of the ccss 2012 Data

To better understand the behaviors of Chinese students in college, we need to first realize that the college students are living in a rapidly changing society influenced by the global trends with their learning behaviors being the products of their prior learning experience, current learning goals, college education norms and overall perceptions towards learning. Based on the data collected in 2012, three cases will be taken to discuss the issues and to depict college student learning in a macro contexts.

table 4

Description of demographic variables and institutional characteristics

5.1 Comparison of Student Learning Behaviors in 5 Benchmarks: China’s 985 Project Universities with American Students in Top Research Universities

With the aim of forming world-class universities, the 985 Project University in China has undertaken considerable effort in improving campus facilities, reforming teaching/learning and increasing quality of education. 42 What is the situation of student learning engagement compared with the American top research universities? Table 5 shows the student’s scores in the five benchmarks from China’s top research universities (985 project universities) and those of his counterpart in the u.s.: Research Universities with high research activity (ru/h).

We find that the mean scores of Chinese students are significantly lower than American students in lac and sfi, while the scores in eee and sce was higher than the American counterparts. The difference between the acl scores is statistically insignificant. Score comparison alone is an inadequate measure of the differences between students learning in two countries, a rather more meticulous study is necessary to reveal the underlying reasons. The Tsinghua team has designed some specific studies to explore these, and though they do not quite uncover all the details, some interesting clues have been found for, which may possibly assist in future research. For example, the low scores of Chinese students in Student Faculty Interaction (sfi) are not just attributable to the low frequencies and low levels of interaction with their faculties. Some Chinese students explained in the interviews their unwillingness of discussing examination scores with teachers, simply because they believed this sort of behavior is frowned upon in Chinese society. Moreover, they did not want to challenge the teachers’ authority by arguing with them on the issue. It leads us to consider the cultural differences in the two societies—the low scores of the Chinese students in sfi may be rooted in the social norms of Chinese society. 43

table 5

T-test and effect size analyses of five benchmarks

* p < 0.05 **p < 0.01 ***p < 0.001

data resource: http://nsse.iub.edu/institutional_report/benchmarks/fy%20bench%20by%20carn.pdf

5.2 High Impact Educational Practices (hips) in Chinese College

Experienced educators and researchers conducting empirical studies often find that certain activities in college tend to be more purposeful and influential upon students. Kuh recommends these activities as high impact practices (hips), 44 believing that participation in these practices can create life-changing results. 45 From the wealth of previous experience with and research into the college learning process, hips are frequently characterized by the following: (1) demanding considerable time and effort, (2) facilitating learning outside of the classrooms, (3) requiring meaningful interactions with the faculty and other students, (4) encouraging collaboration with diverse others, and (5) providing frequent and substantive feedback. The nsse constructed the benchmark of hips on the basis of six practices that meet these requirements: (1) practica, internships or fieldwork, (2) organizing or joining in the learning community or student association activities, (3) working with a faculty member on a research project, (4) studying abroad, (5) participating in an academic or design competition, (6) contributing to an academic journal or conference.

The questions for us to focus on are whether the benchmarks of hips constructed in American culture fit Chinese students, and if there are unique features of the practices for Chinese students. Akin to what was shown in the nsse Annual Results 2012, students in the “985 Project” universities in China also participated in a variety of hips, and differences by school year and discipline were also observed. However, the structure of the practices are different between the two countries, with the concepts being associated with different things.

As shown in Figure 2, the hips in China can be divided into three categories. The first is closely related to the academic development of students, the second is strongly correlated with their potential career developments, while the third is intended to improve communication skills and connection to society.

In the 985 project universities, the junior students are more active in the third type of hips, especially in organizing or participating in student association activities, (t = 4.36, effective size = 0.13, p < 0.001) while the seniors are more active in the first and second types of hips, especially working with a faculty member on a research project (t = –15.44, effective size = –0.50, p < 0.001), and practicum (t = –13.84, effective size = –0.44, p < 0.001). The final years for undergraduates in college in China is more of a transforming or preparing period for a new life than a natural continuity of the previous school years. Senior students who have passed the early selection of post graduate positions will join the research team led by the advisors and do a practicum or graduating assignment in the field, which is the requirement for almost all senior students in China. In order to investigate the correlation of the hips with students’ learning outcomes in China, we perform a multiple linear regression, which controls four original, comparable benchmarks, in order to analyze the impact of on students’ academic performance and learning outcomes.

Restructuring the items of hips based on the college life in China

academic performance and learning outcomes

= β₀ + β₁ hips + β₂ lac + β₃ acl + β₄ sfi + β₅ sce

We find that hips do not strongly affect students’ academic performance, but have a great influence on student learning outcomes (β ₁ = 0.154, t = 11.15, p < 0.001), especially in acquiring a broad general education (β ₁ = 0.095, t = 5.84, p < 0.001), speaking clearly and effectively (β ₁ = 0.086, t = 5.52, p < 0.001), thinking critically (β ₁ = 0.074, t = 4.45, p < 0.001), solving complex, real-world problems (β ₁ = 0.098, t = 6.20, p < 0.001), learning effectively on their own (β ₁ = 0.139, t = 8.49, p < 0.001), and understanding oneself (β ₁ = 0.118, t = 7.11, p < 0.001).

To better understand which kind of hips improves what aspect of student learning outcomes, we conducted interviews. Making a comparison of detailed learning outcomes between the students who have experienced various hips and those who have not, we find out their corresponding relationship, which is shown in Table 6.

table 6

Corresponding relationship between HIPs and learning outcomes

* p < 0.05 ** p < 0.01 *** p < 0.001

In conclusion, universities in China, just as in the u.s., provide education not only through classroom teaching/learning, but also through extra-curricular activities. The hips are indispensable for college students so he institutions should encourage students to actively participate in practices matching their growth stage and professional background. Although hips may not improve academic performance, they drive students to absorb, explore and accumulate knowledge, form good habits of learning autonomously and thinking critically, gain a better understanding of their abilities and interests, and overcome difficulties by challenging themselves and promoting cooperation. The explanation of why hips have a significant impact on student learning outcomes and the specific mechanism in Chinese universities need to be investigated in further detail.

5.3 Learning Features of First-Generation College Students

The first-generation college students here refer to the students whose parents have not received any degrees in he. Among our respondents 75.6% students belong to this category. According to the general demography, the number of first-generation students has increased with the years, which means that a growing number of young people have had the opportunity to enter universities and colleges (χ2=67.29, p < .001).

We tested the distribution of first generation and non-first generation students using Chi-square test and found that a statistically significant differences exist between the two groups in terms of school type, gender, grades, ethnicity, number of siblings and hometown.

From the distribution table, the first-generation college students are mainly enrolled into local universities (χ2=1059.39, p < .001). Most of the first-generation college students (77.8%) are male (χ2=202.71, p < .001). The Han ethnic group has a much higher representation (76.5%) than minority ethnic groups in terms of first-generation students (χ2=338.04, p < .001). Among the first-generation students, 90.6% comes from families with more than one child (χ2=9818.63, p < .001). The data also revealed that, compared to non-first-generation students, more first-generation students come from the countryside (χ2=9571.36, p < .001).

Table 8 gives the T-test results, by category, of the learning behaviors observed in independent samples of college students. As the table shows, the first-generation college students tend to be “good” students from the traditional Chinese perspective. They complete assignments in time and listen to their teachers’ instruction carefully. However, they do not communicate with their faculties by asking questions or actively responding to them, let alone by making a presentation or questioning the teachers’ viewpoint. Learning in this fashion meets the outside requirements, but does not activate their personal, learning desires.

table 7

Chi-square analysis

* p < 0.05 ** p < 0.01 *** p < 0.001

table 8

T-test of study behaviors

* p < 0.05 ** p < 0.01 *** p < 0.001

In terms of time allocation, the first-generation college students spent a lot of time on the Internet; for example, playing games, chatting online, or watching online videos. They participated less in extra-curricular courses, sports activities. Based on our observation, first-generation students are more likely to be within a small social circle due to financial constraints or lack of self-confidence. Therefore, they do not usually spend much time exploring their potential abilities or new territory in the region.

As for personal relationships on the campus, first-generation students reported a better relationship with their classmates compared to non-first-generation students. Non-first-generation students received more support from other important people like their faculty, administrators and class assistants who have authority on resource distribution and student benefits.

From the above data, a portrait of the first-generation college students seems to present themselves: they try hard to be good students in terms of classroom learning, but do not invest much energy in higher order learning activities. They participate in fewer extracurricular activities with others and protect themselves by staying in their comfort zone, playing online games or watching television. They hardly communicate with important people on campus and prefer to ask for help from their peers. First-generation students are a group who are easily overlooked. Compared with the non-first-generation college students, they make less efficient use of resources provided by the college.

table 9

T-test of time allocation

* p < 0.05 ** p < 0.01 *** p < 0.001

table 10

T-test of personal relationships

* p < 0.05 ** p < 0.01 *** p < 0.001

5.4 The Impact of Ascribed and Achieved Status Factors to College Student Learning

Living in a rapidly changing society with tremendous geographic mobility, students enter colleges with abundant personal background knowledge (pbk), including ascribed status factors, such as a cultural or social heritage from their parents or elder generations, and achieved status factors, which they gain mainly through their own educational efforts. In order to assess the ascribed and achieved status factors affecting student learning and development (i.e. learning behavior, psychological learning status, self-reported attainment and satisfaction), a statistical nested model is designed: (1) the reduced model includes demographic variables, institutional characteristics, pre-college experience variables and dependent variables; (2) in the full model, the ses Index representing family socioeconomic status is added to the previous model as an independent variable, since the ses Index has been shown to be significantly associated with pre-college experience as well as college student experience.

In the reduced model, the gross effects of high school attainment and high school types on student learning and development are estimated. With the full model we are able to estimate the size of the effect of precise pre-college on student development by controlling the ses effect. We may also compare the effects of pre-college experience and the ses Index on student development using their standardized coefficients. Furthermore, the coefficients change in the full models after entering the ses Index, which serves to roughly illustrate the correlation between the ses Index and other variables.

Our overall findings are as following: a student’s pre-college experience is the most important factor influencing one’s college learning and is the strongest predictor on the college learning outcome of that student, whether in the reduced model or in the full model (p< 0.001).With the entry of the ses factors, the student’s pre-college experience coefficient is reduced slightly. However, compared with the impact of the ses, the effect of pre-college experience is still very strong. In the full model, it shows that the ses is also highly correlated with student learning activities and their outcomes.

Table 11 shows that after controlling student backgrounds and institutional characteristics, pre-college experience is closely associated with Active and Collaborative Learning (acl), Student and Faculty Interaction (sfi) and Enriching Educational Experience (eee). It suggests that students have gained learning habits through pre-college activities, which continuously influences their higher educational learning process. According to the statistics, the ses Index has moderate effects on acl, sfi and eee. Although previous studies argued that the impact of ses on education has gradually decreased as the general education level increased, it is inappropriate to entirely ignore its influence. The same phenomenon is also exhibited in the analyses of learning status and learning outcomes, which supports our preliminary assumption (Table 12 and Table 13).

Control variables are gender, ethnic, grade, major, university type, location, local students.

Furthermore, we calculated the difference between the standardized coefficient of pre-college experience and the ses Index. The result showed that the pre-college experience coefficient is larger than the ses coefficient, ranging from 0.0425 to 0.1347. It implies that learning habits, learning strategies, and other behaviors formed in high school have a larger influence on their college learning than students’ family backgrounds. Good learning habits and positive learning experience gained in the pre-college period will greatly benefit their college learning. Therefore the efforts in improving learning in he should not be limited to the college level- it should include high school education. If we want to increase the quality of college education, we need to work closely with educators from the basic education sector.

table 11

Coefficients of demographic variables, institute characteristics, pre-college experience and SES on student learning behaviors

* p < 0.05 ** p < 0.01 *** p < 0.001

table 12

Coefficients of demographic variables, institute characteristics, pre-college experience and SES on student learning status

* p < 0.05 ** p < 0.01 *** p < 0.001

table 13

Coefficients of demographic variables, institute characteristics, pre-college experience and SES on student learning outcomes

* p < 0.05 ** p < 0.01 *** p < 0.001

Furthermore, the higher school factor is stronger than the family background. Its direct effect on higher education is not much stronger than proximal factors like high school impacts. When we compare the pre-college experience coefficient in the reduced model and the full model, there is a slight change. It demonstrates that the relationship between the pre-college period and the ses is fairly weak. On the other hand, the high school coefficient seemed much more sensitive to entry of the ses. We also conducted correlation analyses for the ses Index, the pre-college experience and high school type. We discovered that these variables were significantly correlated. However, the correlation coefficient between ses and high school was as high as 0.472 (p< 0.001) while the coefficient between ses and pre-college experience was only 0.140(p< 0.001).

To sum up, both pre-college experience and ses have considerable impact on college students’ learning and development. The pre-college experience has a strong impact on college students’ learning process and outcome, while the students’ ses mainly influences students’ access to a good high school and their performance in secondary school. The finding provides us much incentive to make all tiers of education, not just he, better suited to the students, whose development will decide the future.

6 Conclusions and the Future of the Project

During the past 5 years, the ccss has had a growing impact on Chinese higher education researchers and practioners for two reasons: one, it is the first time that China has initiated a student survey on a nation-wide scale, focusing on student-centered pedagogy and the student learning experience based on observable behaviors and tractable psychological features. Data accumulated from the survey can be used to support more “evidence-based” policy making at the institutional and national level. Furthermore, existing problems in a student’s learning process can be identified based on relatively large scale data analysis, so that targeted strategies can be put into actions in reforming classroom teaching/learning behaviors and campus-wide student life experience.

In addition, along with the expansion of participating heis in the ccss, the idea of “student-centered” and “educational process matters” embraced by the ccss is gradually accepted by Chinese heis and is becoming a popular concept. Although we never intended to create a university league table of student engagement, the ccss survey indeed provides a new perspective focusing on interaction between students and the educational environment the institution provides. It is believed that the more heis take on this perspective, the better the educational quality provided.

We have broached a number of topics in this paper, namely the comparison of Chinese top university students and their American counterparts in ru/h in student engagement and high impact practices specifically, learning features of the first-generation college students who comprise the majority of today’s Chinese college students, and the impact of ascribed and achieved status factors to college student learning. Although the treatment here is more hasty than we would like, due to the limitation of the length of the article, they nonetheless illustrate how the data can be further investigated to “unlock the black box” of Chinese college student learning.

As indicated in this article, we have made a great effort in indigenizing the survey instrument. We believe that with the increasing global role China is expected to play, a tool that allows international comparisons is indispensable. We have retained the five benchmarks of the nsse to allow for easier comparison between the data from the two countries. We have also innovated heavily to adapt it for the Chinese political, cultural, and educational context. In the future, we will further study the instrument in the following ways: first, to explore the key indicators which describe the characteristics of the development and learning of Chinese students; second, to construct the theory of Chinese college students learning and development based on empirical and normative research; last but not the least, to create the national database of Chinese student learning and development for longitudinal study.

Attempting to break open the black box of Chinese college student learning is an endless yet meaningful journey. We shall continue the journey with the confidence that a better understanding of the learning process of Chinese students will not just benefit China, but the whole world.