The Influence of Social Capital and Intergenerational Mobility on University Students’ Sustainable Development in China

Abstract

A person’s family of origin has a profound impact on his or her life; a student’s performance at university and their interpersonal skills are also influenced by the characteristics of their family. In order to explore how social capital and intergenerational mobility impact university students’ quality of interpersonal communication and experiences, whether there is a “Matthew effect” at the higher education level, how the external social capital and the parenting style within a family affect the comprehensive quality of university students separately, and how university students’ family characteristics impact their quality of interpersonal communication, which will in turn affect their performance at university, this paper starts from the perspective of social capital and intergenerational mobility, and uses the cross-sectional data of the China Family Tracking Survey (CFPS) from 2018 to conduct Kaiser-Meyer-Olkin analysis and logistic binary regression for 1037 university students nationwide, to check whether the selected variables can be subjected to principal component analysis. The results show that, on one hand, university students’ family characteristics have a significant positive impact on their quality of interpersonal communication; on the other hand, their family backgrounds also have significant impact on their academic performance and enthusiasm for participating in student unions, as well as part-time employment and internships. In addition, according to the results of the mediation test, interpersonal communication skills play a mediating role in the way that family characteristics impact students’ performance at university, including their participation in student organizations, as well as internships and part-time jobs. Finally, this paper, by referring to both theoretical and empirical analysis, presents relevant suggestions from the perspectives of the individual, family, and government, aiming to facilitate the sustainable development of university students.

1. Introduction

The training of university students, as the reserve force of China, occupies a special position in China’s talent training system. A series of policies and regulations have been introduced by China over the past few years regarding undergraduate education. In 2017, the State Council of China issued the “Opinions on Strengthening and Improving Ideological and Political Work in Colleges and Universities under the New Situation”, which pointed out that universities should adhere to the principle of educating students in an all-around way, by all relevant parties, and throughout the whole process [1]. The Ministry of Education of China proposed to speed up the development of higher-level undergraduate education by adhering to “taking undergraduate education as the root of university education” [2]. Therefore, one of the most widely discussed topics in society is how to further optimize the talent training model and improve university students comprehensively.

The overall quality of university students is affected by many factors. Parents are children’s first mentors, and children are exposed to their family characteristics from birth, so family has an influential role to play throughout children’s growth. The American sociologist Coleman pointed out in his report that family characteristics have a profound impact on students’ development. Factors contributing to family characteristics, such as financial status, parenting style, parents’ cultural backgrounds, and social capital, are coupled together to affect students’ performance at university and interpersonal communication skills [3]. As higher education becomes increasingly accessible in China, undergraduate education has turned from the elite model to the mass model. In addition, students’ family backgrounds are becoming increasingly complex and diverse. According to statistics, the gross enrollment rate of colleges and universities in China has been increasing [4]. However, unequal access to education has also become prominent. According to some scholars, students’ access to higher education resources is affected by family background, or whether they come from rural or urban areas [5]. Some scholars believe that in key higher education institutions, students from families with relatively abundant social capital account for a large proportion, while the proportion of students from rural areas or from disadvantaged backgrounds is on the decline [6]. In addition, resources for higher-quality education are slightly tilted in favor of families with perceived superior backgrounds, and the threshold for higher education is unequal as well. This, to some extent, indicates that social capital is characterized by intergenerational mobility in the field of education [7]. The majority of previous studies focused on the acquisition and quality of higher education [8], while ignoring the internal identity differences among educated people [9,10], such as family background, personal characteristics, etc.

The following questions remain unanswered: How do social capital and intergenerational mobility impact university students’ quality of interpersonal communication and experiences at university? Is there a “Matthew effect” at the higher education level? How will the external social capital and the parenting style within a family affect the comprehensive quality of university students? Will university students’ family characteristics impact their quality of interpersonal communication, which will in turn affect their performance at university? In response to the above questions, the following hypothesis is proposed in this study: the comprehensive level of family characteristics can promote the quality of university students’ interpersonal communication and school experience. Meanwhile, there is an intermediate influence mechanism between the interpersonal skills of university students and the comprehensive level of family characteristics. The study uses the data from the China Family Panel Studies (CFPS) of 2018, and takes 1037 university students, screened nationwide, as representatives. On the basis of theoretical analysis, principal component analysis is used to reduce the dimension of family characteristics, logistic binary regression research is applied to make quantitative analysis of university students’ interpersonal relationship and school performance, and the mediation effect was used to test the influence mechanism. With the aforementioned methods, the study aims to explore the impact of students’ family characteristics on their interpersonal communication skills and experiences at university.

The innovation and marginal contribution of this research mainly lie in that: theoretically, this paper will enrich the research on the impact of university students’ family backgrounds on their comprehensive performance and work out a fairly complete intergenerational mobility mechanism of social capital; and realistically, based on the theoretical and empirical results, this paper gives relevant suggestions for improving the comprehensive quality of Chinese university students from the perspectives of the “individual, family, and government”, so as to facilitate the further implementation of the “three-dimensional comprehensive talent cultivation” policy and improve the talent training system.

This paper is structured as follows: Section 2 reviews the existing literature and puts forth the theoretical basis and hypothesis development. Section 3 describes data sources, research samples, and statistical models. Section 4 is mainly devoted to descriptive statistics of variables, regression analysis, robustness tests, analysis of the impact mechanism, and heterogeneity analysis. Finally, Discussions, Conclusions, and Recommendations are provided.

2. Literature Review and Hypothesis Development

2.1. Family Capital and Education

Researchers believe there are reciprocal cause-and-effect relationships between social capital and education. On the one hand, education is significant to the generation, maintenance, and disappearance of social capital [11]. On the other hand, social capital is believed to be a key explanatory factor of actors’ educational experience [12]. This paper focuses on the second aspect in order to explore to what degree and how social capital can explain actors’ educational experience. Educational experience in this paper undoubtedly refers to universities’ educational experience.

The concept of social capital, as is generally believed by the academic community, comes from the “social capital theory” proposed by French scholar Bourdieu, who defines it as “the actual or potential collection of resources comprised of relationships characterized by mutual acquiescence or recognition” [13]. Family capital is a sub-concept under social capital and is primarily composed of marriages, blood relationships, and inheritances [7]. It plays an instrumental role in determining children’s educational outcomes. The British “Plowton Report” proposes that family capital is a multi-dimensional comprehensive concept, which mainly encompasses parents’ education level, the family’s financial status, parents’ occupations, parents’ expectations for their children and parenting styles, etc. Individuals from families with more family capital are often provided with more resources and information during their growth [14], and thus have easier access to higher education. In his theory on capital habits, Bourdieu points out that, compared with personal factors such as diligence and intelligence, the upbringing standards of individuals from families with more capital have more commonality with the standards implemented by the ruling institutions. He believes that the fundamental reason for social inequality lies in the compatibility difference between a family’s parental standards and the society’s mainstream standards. He also believes that family capital is greatly transmitted through the precepts and deeds of the parents. The American scholar Coleman proposes in his family capital theory that family social capital refers to a closed network of relationships established both inside and outside of the family. This network plays an invaluable role in the development of children and enables a family’s human capital and cultural capital to be transmitted across generations. If a family’s social capital is absent, the mechanism for the transmission of other capital will be blocked [15].

Education is an important part of the societal system, and the mechanism for its acquisition also plays an influential role in the intergenerational mobility of family capital [16]. The current research on education and family capital mostly focuses on the weak reproduction model under the social reproduction paradigm, i.e., based on the BlauDuncan model of status acquisition, studying the availability and quality of children’s education, and discussing how they are influenced by family background factors such as parents’ academic degrees, family financial status, and social status [17,18]. Bowes proposes that non-cognitive capital within a family is the root of social inequality [19]. This non-cognitive capital is very sensitive to the financial status of a family; that is, parents’ investment in their children can significantly improve children’s non-cognitive capital. Liang Chen further demonstrates the views of previous researchers by arguing that parents’ financial investment in children’s education affects their comprehensive ability [16]. Han Yu points out that, in addition to a family’s financial capital, the parenting style, such as family rules and customs, also impacts children’s ability [20].

To sum up, previous empirical studies have mostly focused on a family’s financial capital, while less discussion can be found on intra-family communication. According to the Mindsponge Theory, the human mind is similar to a sponge that selectively assimilates or rejects different cultures and concepts [21,22]. It synthesizes and innovates such information in the brain to shape personality and thoughts. Family environment, as the earliest and longest contact environment for children in the process of growing up, plays an important role in shaping children’s personality and thinking modes. Coleman points out in his previous research that, compared with parents’ behavioral or regulatory involvement, their emotional involvement has a greater impact on children [23]. The social capital within a family should be grounded in an emotionally warm style of intergenerational interaction. If researchers only focus on a family’s financial capital and social connections while ignoring the emotional characteristics of family connections, they will misunderstand the impact of family capital on children’s education [24]. Therefore, when studying the influence of family characteristics on education and children’s ability, it is necessary not only to observe the factors at the macro social structure level but also to analyze the parent–child interaction within the family at the micro level [15].

In this sense, the study summarizes the past definitions of family characteristics, and puts them into two categories: a family’s external social capital, represented by a family’s financial status and social connections, as well as parents’ academic degrees; and a family’s internal social capital, represented by parent–child communication and intergenerational relationships. In addition, university students’ behavior performance as discussed in this paper falls into two categories, i.e., quality of interpersonal communication and performance in university. The latter includes academic performance, degree of participation in student unions, and internship and part-time experiences.

2.2. Family Characteristics and University Students’ Interpersonal Communication

Interpersonal communication refers to a behavioral process in which individuals conduct emotional transmission, personal expression, and exchange of opinions through verbal or behavioral media [25]. The existing literature and studies tend to link an individual’s interpersonal communication skills with emotions and intelligence, and there is a belief that interpersonal communication is a subject’s natural response to external stimuli provided by others, such as emotions and motivations [26]. There are many factors that affect the interpersonal communication skills of university students, among which family characteristics, as one of the key factors, plays a fundamental role in the generation of university students’ quality of interpersonal communication [14]. Chinese traditional culture is a relational culture, exploring the significance of individual existence in social relations. Confucius said, young people should be filial to their parents at home and respectful to their brothers when they are with them. They should be serious and trustworthy, love the populace extensively, and be close to those who are humane. When all this is done and there is time for other things, they should use it for the study of the classics [27].

People’s minds affect the formation of their personality characteristics and produce a mechanism similar to the sponge mechanism (namely Mindsponge Theory) when absorbing or rejecting external information [22]. Such a mechanism is based on the interaction between the subjective world (such as mentality, buffer zone, and multiple filtering systems in the world), and the objective world, i.e., the external environment. The family of origin, as a key component of the external cultural environment, gives people life-long influence. A family’s upbringing style and atmosphere, parents’ occupations and education, as well as financial or income level will affect the children’s personality, habits, and interpersonal communication skills indirectly or directly.

Children raised in different family backgrounds have different characteristics, communication habits, and styles of handling interpersonal relationships. Individuals from families with more capital tend to be more positively trained during their growth. Individuals from a discordant family of origin are more likely to experience feelings of inferiority and suspicion in the process of growing up [28]. Other studies found a correlation between parents’ educational degrees and family upbringing styles: parents with higher degrees tend to respect and understand their children [29]. Many pieces of the literature point out that university students’ capabilities, such as personality, emotional management, character, and adaptability, are also influenced by family capital [30].

Due to the social attribute of people, interpersonal communication skills are closely related to the degree of participation in social activities. People with stronger interpersonal communication skills tend to be more active participants and are therefore more likely to acquire social capital [31]. Other authors also pointed out in their research that high interpersonal levels can help university students quickly adjust as they enter a different environment, and help them mix into the environment [32,33]. Students with higher-quality interpersonal communication tend to have greater adaptability in society, while those with lower communication levels often struggle to adapt to the workplace after they graduate, which may subsequently affect their development. That is also one of the manifestations of inequality in higher education [14].

To sum up, most of the existing literature posits that family characteristics, whether at the subjective or objective level, impact the interpersonal level of children. Children from families with more capital tend to have higher communication quality. Based on this, this paper proposes the following hypotheses:

Hypothesis 1.

The comprehensive level of family characteristics has a promoting effect on the quality of university students’ interpersonal communication.

Thus, the higher the comprehensive level of a family’s characteristics, the higher the quality of interpersonal communication of university students who grew up in the family. Among them, the comprehensive level of family characteristics includes the external family characteristics represented by income level, family social connection, and parental education level, as well as the internal family characteristics represented by parent–child communication frequency.

2.3. Family Characteristics and Students’ Performance at University

According to the existing literature, university students’ performance is mainly demonstrated in three dimensions: academic performance; participation in student unions; and internships and part-time employment. Different scholars have had different results when it comes to the influence of family characteristics on students’ performance at university. In terms of academic performance, some researchers believe that it is linked with the level of family capital: the intergenerational mobility of family cultural capital refers to the transmission of knowledge, strategies, and behavioral styles from parents to children to some extent, and parents’ ways of communicating with children and family atmosphere have a positive impact on students’ academic performance and degree of participation in student unions, in addition to improving their creativity and enthusiasm [34,35].

Through the family’s internal relationship network, parents deliver their educational aspirations to children [36], supervise and discipline children [37], and guide their study, or pass on obtained useful information about school to their children [38]. The attention, as well as invested time and energy from parents (and other adults), toward children’s growth serves as pivotal social capital. Empirical research finds that, among various patterns of parents’ participation, the way of discussing school affairs with children has the most remarkable impact on improving children’s academic performance. The more frequently parents interact with their children, the more likely they will be to deliver educational aspirations to their children, which will in turn encourage children to achieve success in their success [14].

Based on the “relative risk aversion mechanism (RRA)”, Goldthorpe proposes that different social classes have different strategies for avoiding class downgrade [39]. The upper class tends to increase investment to secure their children’s advantages in studies, while the lower class is more likely to invest only if their children perform exceptionally well academically. Family conditions are often positively correlated with university students’ academic performance [40]. Some researchers find that family characteristics such as family atmosphere, family–university connection, and a family’s expectations on children’s studies all impact children’s academic performance [41,42]. According to some studies, family conditions may apparently influence university students’ performance at their early stage of admission, but the gap will gradually narrow as they continue to study at university [43]. In terms of joining student unions, university students with better family conditions are more likely to join unions and become student leaders at school, and are more likely to be awarded [44]. Students from an advantaged family background are more active participants in student unions, and they are more likely to assume senior student positions. Scholars, however, have pointed out that students from urban and rural areas display a small difference in their ability to organize and coordinate, and in their ability to serve as student leaders [43]. At present, there are few studies on how family characteristics impact internships and part-time employment. In addition, Mahfud T. finds through empirical research that psychological capital, such as interpersonal communication, and social capital are important factors affecting university students’ internship experiences and even entrepreneurship experiences [45].

To sum up, the impact of family characteristics on university students’ experiences at university can be summarized in two dimensions (see Figure 1): On one hand, family characteristics directly affect students’ experiences at university; on the other, family characteristics affect their interpersonal communication skills, which in turn affect their performance at university. Based on the above analysis, this paper proposes the following hypotheses:

Hypothesis 2a.

The comprehensive level of family characteristics has a positive influence on university students’ experience in school.

That is, the higher the comprehensive level of family characteristics, the richer the school experience university students who grow up in the family will have (such as better academic performance, more participation in student organizations, more internship and part-time work experience, etc.). Among them, the comprehensive level of family characteristics includes the external family characteristics represented by income level, family social connection, and parental education level, as well as the internal family characteristics represented by parent–child communication frequency.

Hypothesis 2b.

University students’ interpersonal communication skills have a mesmeric effect when it comes to how the comprehensive level of family characteristics influence their campus experience.

The comprehensive level of family characteristics is defined as above.

3. Materials and Methods

3.1. Sample Selection

The empirical analysis in this paper uses the 2018 data from the China Family Panel Studies (CFPS). The China Family Tracking Survey Database, which tracks data at the individual, family and community levels across the country, covering economic activities, educational outcomes, family relationships and dynamics, population migration and health among other aspects, has been officially available since 2010. The sample covers 25 provinces/municipalities/autonomous regions, with a target sample size of 16,000 households. The objects include all family members in the sample households. Different questionnaires were designed for different groups including communities, families, adults, and children. The survey fully reflects the changes in China’s society, economy, population, education, and health, providing a statistical basis for related research and analysis.

This paper screens the data of CFPS in 2018. Since the research objects are students receiving higher education in China, this paper focuses on undergraduate students. In addition, it removes samples that miss key information or whose information is unavailable. Finally, 1037 samples with relatively complete information were kept for empirical discussion, including 544 students from urban areas and 456 students from rural areas.

3.2. Variable Description and Data Sources

The definition and descriptive statistics of each variable are as follows (see

Table 1. Descriptive statistics of main variables.

Metric			Variable	Average	Value Description
Dependent variables	University students’ quality of interpersonal communication		Coumm	0.8571	1 = relatively high quality of interpersonal communication; 0 = relatively poorer quality of interpersonal communication
	Students’ performance at university	Academic performance	sty	0.3973	1 = relatively strong academic performance; 0 = relatively weak academic performance
		student unions	union_a	0.5005	1 = have experiences of joining student organizations; 0 = no experience of joining student organizations
		internships and part-term employment	internship	0.6358	1 = have internships or part-term employment; 0 = no internships or part-term employment
Independent variables	University students’ family characteristics		C1	3.9199	maximum value = 6.1793, minimum value = −0.8388, standard deviation = 0.9591
			C2	−0.1939	maximum value = 1.9246, minimum value = −1.5188, standard deviation = 0.6749
			C3	8.5232	maximum value = 13.9212, minimum value = −4.5230, standard deviation = 2.2725
Control variables	Gender		gender	0.3761	1 = male; 0 = female
	Type of household registration		household registration	0.5440	1 = urban area; 0 = rural area
	Origin of students		province	0.3086	1 = eastern region; 0 = central and western regions

Data from CFPS 2018.

).

3.2.1. Independent Variables

Taking the family characteristics of university students as the core explanatory variable, this paper, in accordance with the previous theoretical analysis, summarizes the family characteristics of university students into two categories: a family’s internal cultural characteristics and social capital. A family’s internal cultural characteristics are mainly manifested by the children’s frequency of contact with parents, parents’ marital status and the size of the family; a family’s external social capital is mainly demonstrated by parents’ education levels, the family’s external social network, and the financial level. Based on this, 9 questions in the CFPS 2018 questionnaire were selected, namely: A1 “Member Confirmation”, QF706 “Contact Frequency”, QF5 “Relationship”, QEA0 “Current Marital Status”, W01 “Highest Degree”, FINC “Gross Income for the Past 12 Months (CNY/year)”, FP515 “Financial Help to Relatives (CNY/year)”, FU201 “Expenditure for Favors and Gifts (CNY/year)”, FP516 “Financial Help to Others (CNY/year)”. This paper sums up the values of FP515, FU201 and FP516. As there are missing values, the total amount is added by one and then the logarithm is taken; it is defined as the clan of the family, expressed by clan_l; FU201 is treated in the same way, and it is defined as a family’s income, expressed by income_l; A1 is defined as the size of family, expressed by famsize. This paper sorts out the frequency of contact with father/mother respectively from QF706, and takes the average of the two to calculate the frequency of connections with parents, which is expressed by connect_p. The same method is adopted for dealing with QF5 and W01, with the former defined as the relationship with parents, expressed by feel_p, and the latter as the educational level of parents, expressed by edu_p; The parents’ marital status in QEA0 is matched and expressed by married_p.

Because family characteristics contain many variables, and it is difficult for a single indicator to represent the impact of family background on children’s experiences at university and interpersonal communication, this paper adopts the principal component analysis method to standardize and transform the data through the idea of dimension reduction, and converts multiple indicators into three relatively comprehensive measures of family characteristics, which are defined as C1, C2, and C3. This paper first conducts Kaiser-Meyer-Olkin analysis (hereafter referred to as “KMO test”) to check whether the selected variables can be subjected to principal component analysis (see

Table 2. Principal component analysis KMO test results.

Variables	KMO	C1	C2	C3
edu_p	0.5154	0.3779	0.5948	−0.0594
clan_l	0.5169	0.1593	0.0116	0.5175
connect_p	0.5030	0.5655	−0.3906	−0.1560
feel_p	0.5018	0.5547	−0.4050	−0.1732
married_p	0.5157	0.3851	0.5715	−0.1712
income_l	0.5344	0.1912	−0.0354	0.6472
famsize	0.5150	−0.1398	−0.0400	−0.4756
Full sample	0.5146	-	-	-

Data from CFPS 2018.

). The results of the KMO test are positive. According to the results of the principal component analysis in

Table 3. The results of analyzing principal components of family characteristic indicators.

Principal Components	Eigenvalue	Proportion	Cumulative Contribution Rate
C1	1.84747	0.2639	0.2639
C2	1.60867	0.2298	0.4937
C3	1.25942	0.1799	0.6737
C4	0.95674	0.1367	0.8103
C5	0.74345	0.1062	0.9165
C6	0.32575	0.0465	0.9631
C7	0.25651	0.0369	1.0000

Data from CFPS 2018.

, the horizontal line where the eigenvalue is equal to 1 in the gravel diagram (see Figure 2) is the dividing point for retaining the principal components, and the first three principal components are selected to measure the family characteristics of university students.

$$\begin{array}{l}{\mathrm{C}}_1=0.3779\times \mathrm{edu}\_\mathrm{p}+0.1593\times \mathrm{clan}\_\mathrm{l}+0.5655\times \mathrm{connect}\_\mathrm{p}\\ +0.5547\times \mathrm{feel}\_\mathrm{p}+0.3851\times \mathrm{married}\_\mathrm{p}+0.1912\times \mathrm{income}\_\mathrm{l}\\ -0.1398\times \mathrm{famsize}\end{array}$$

(1)

$$\begin{array}{l}{\mathrm{C}}_2=0.5948\times \mathrm{edu}\_\mathrm{p}+0.0116\times \mathrm{clan}\_\mathrm{l}-0.3906\times \mathrm{connect}\_\mathrm{p}\\ -0.4050\times \mathrm{feel}\_\mathrm{p}+0.5715\times \mathrm{married}\_\mathrm{p}-0.0354\times \mathrm{income}\_\mathrm{l}\\ -0.0400\times \mathrm{famsize}\end{array}$$

(2)

$$\begin{array}{l}{\mathrm{C}}_3=-0.0594\times \mathrm{edu}\_\mathrm{p}+0.5175\times \mathrm{clan}\_\mathrm{l}-0.1560\times \mathrm{connect}\_\mathrm{p}\\ -0.1732\times \mathrm{feel}\_\mathrm{p}-0.1712\times \mathrm{married}\_\mathrm{p}+0.6472\times \mathrm{income}\_\mathrm{l}\\ -0.4756\times \mathrm{famsize}\end{array}$$

(3)

3.2.2. Dependent Variables

In this paper, there are two types of dependent variables, which are explored separately: one is university students’ quality of interpersonal communication, and the other is their performance at university.

• Quality of interpersonal communication of university students

According to Hypothesis 1, put forward in the previous theoretical analysis, this paper takes the interpersonal communication skills of university students as the explained variable, expressed by coumm. The logical variable is developed according to QM2011 in the CFPS 2018 questionnaire, “How strong is one’s personal relationship (score)?”, with 11 options, ranging from 0 to 10. This paper assigns a score of 6 and above as 1, which is defined as having high-quality interpersonal communication. The rest are assigned a value of 0, which is defined as having relatively poor-quality interpersonal communication.

• Students’ performance at university

The performance of students at university is one of the variables explained in this paper. According to Hypothesis 2, put forward in the theoretical analysis section, and with reference to the classification of university students’ abilities in the previous literature [13], this paper measures the performance of university students in the following dimensions: academic performance and participation in student unions, as well as part-time employment and internships. For their academic performance, this paper examines QS1011 “non-weekend study time (hours/day)” in the CFPS 2018 questionnaire for measurement. First, the average study time of all individuals is calculated. Individuals whose scores are higher than the average value of 6.67 are assigned a value of 1, while the rest are assigned a value of 0 and represented by sty; for the degree of participation in student unions, this paper examines QS1 “do you serve as a student leader?” and QS2 “have you joined student unions?” in the CFPS questionnaire. Individuals that answer “yes” to both questions are assigned a value of 1, while the rest are assigned a value of 0, which is represented by union_a; for the part-time employment and internships of university students, this paper explores the question QGA101 “internship and part-time experiences” in the CFPS questionnaire, and individuals who say “have had an internship or a part-time job in the past 12 months” are assigned a value of 1, while the rest are assigned a value of 0, which is defined as internship.

3.2.3. Control Variables

For control variables in this paper, three variables at the individual and regional levels are selected, which are the individual ascribed characteristics of the surveyed university students (i.e., gender), the type of household registration (urban or rural), and the origin of students (eastern, central, or western regions). The descriptive statistics of each variable are available in

Table 4. Sample structures of surveyed university students.

Gender	Male	37.61%
	Female	62.39%
Type of household registration	Urban area	54.40%
	Rural area	45.60%
Student source area	Eastern region	30.86%
	Central and western regions	69.14%
Educational level of parents	Below high school	85.14%
	High school	10.89%
	High school or above	3.59%
Parents’ marital status	Married	54.97%
	Divorce, widowhood, remarriage, etc.	45.03%

Data from CFPS 2018.

.

3.3. Model Setting and Research Strategy

Logistic regression is a nonlinear regression, which is widely used in research and theoretical analysis in various fields. It is one of the basic statistical analysis methods and can be used to simulate the probability of an event. Since the dependent variable (university students’ interpersonal communication skills) and the three dimensions of university students’ performance at university in this paper are all virtually binary variables, the benchmark analysis of this paper selects the binary regression Logistic model for empirical analysis. The model is constructed as follows:

$$\mathrm{logit}\left(\mathrm{P}\right)=\ln \frac{\mathrm{P}}{1-\mathrm{P}}=\mathsf{\alpha }+{\displaystyle \sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathsf{\beta }}_{\mathrm{i}}}{\mathrm{X}}_{\mathrm{i}}$$

(4)

4. Results and Analysis

4.1. Sample Characteristics and Descriptive Statistics

The sample of this study is from the CFPS 2018 questionnaire survey, which covers 1037 university students nationwide. The sample structure is as follows (see Table 4).

4.2. Benchmark Regression Results

4.2.1. Analysis of the Factors Influencing University Students’ Interpersonal Relationships

The interpersonal relationship benchmark regression results are presented in

Table 5. Regression results of interpersonal relationship benchmarks.

Variables		Coumm
		(1)	(2)
Independent variables	C1	1.943 *** (3.40)	1.798 ** (2.91)
	C2	0.558 *** (−3.52)	0.584 ** (−3.16)
	C3	0.821 ** (−2.64)	0.849 ** (−2.13)
Control variables	Gender	-	1.495 * (0.54)
	Household registration	-	0.937 (−0.31)
	Province	-	1.130 (0.54)
N		1037	1025
Adj R-square		0.0239	0.0265

Note: Data from CFPS 2018. The values within the brackets are z values; * indicates significance above the 10% level, ** indicates significance above the 5% level, and *** indicates significance above the 1% level.

. The regression results shown in column (1) only regress the core explanatory variables of university students’ family characteristics C1, C2, and C3; column (2) shows the regression results obtained after all variables are controlled; it is found that the probability ratios through estimation of dependent variables in column (1) and (2) are in the same direction and have similar values, which indicates a small possibility of variables being missed. Hypothesis 1 can be confirmed by the basic regression results of university students’ interpersonal relationships, that is, family characteristics have a positive impact on university students’ interpersonal relationships. On the one hand, it shows that a nurturing family atmosphere affects children’s characters by strengthening their interpersonal communication skills; on the other hand, families with better financial conditions can give more opportunities to children for honing their communication skills. At the same time, it can be seen from the benchmark regression results that university students’ gender impacts their interpersonal communication skills. This indicates that there are gender differences regarding university students’ personal communication. It is consistent with the conclusion that male university students are more inclined to take the initiative to make friends [46]. In terms of the type of household registration and students’ origin, it can be found that regional background factors do not significantly impact interpersonal communication skills. No matter whether university students are from urban or rural areas; or eastern, central, or western regions, there is not much difference in their interpersonal communication skills.

4.2.2. Analysis of Factors Influencing Students’ Performance at University

The benchmark regression results of the three dimensions of students’ performance at university are presented in

Table 6. Regression results of benchmarks for students’ performance at university.

Variables		sty		union_a		Internship
		(1)	(2)	(1)	(2)	(1)	(2)
Independent variables	C1	2.423 *** (7.34)	2.431 *** (7.10)	3.616 *** (10.44)	3.464 *** (9.79)	1.410 ** (2.05)	1.378 * (1.90)
	C2	0.479 *** (−6.71)	0.484 *** (−6.49)	0.425 *** (−7.70)	0.435 *** (−7.33)	0.747 ** (−2.05)	0.735 ** (−2.14)
	C3	0.721 *** (−6.70)	0.718 *** (−6.63)	0.592 *** (−10.02)	0.596 *** (−9.71)	0.853 ** (−2.50)	0.853 ** (−2.46)
Control variables	Gender	-	0.942 (−0.42)	-	1.209 (1.31)	-	1.109 (0.67)
	Household registration	-	1.020 (0.14)	-	1.101 (0.66)	-	1.084(0.51)
	Province	-	0.759 * (−1.83)	-	1.015 (0.10)	-	1.381 * (1.82)
N		1011	1000	1011	1000	757	754
Adj R-square		0.0657	0.0661	0.1231	0.1244	0.0080	0.0132

Note: Data from CFPS 2018. The values within the brackets are z values; * indicates significance above the 10% level, ** indicates significance above the 5% level, and *** indicates significance above the 1% level.

. Column (1) in the sty, union_a, and internship models all show the regression results when only the core explanatory variables C1, C2, and C3 are considered. Column (2) illustrates the effect of the independent variable (family characteristics) on academic performance (sty), participation in student activities (union_a), and internships and part-time employment after the control variable is added. It can be seen from the regression results that no matter whether the control variables are added or not, family characteristics would have a positive impact on sty, union_a, and internships. This result is consistent with Hypothesis 2a, put forward in the theoretical analysis section, i.e., family characteristics have a positive impact on students’ experiences at university. By comparing the three dimensions of students’ performance at university, it is found that students’ internships and part-time experiences are less affected by family characteristics than students’ academic performance (sty) and student activity participation (union_a). This indicates that compared with sty and union_a, the internship and part-time experiences of university students are more determined by their personal characteristics. In addition, contrary to the regression results obtained from examining university students’ interpersonal communication skills, students’ regions of origin have a generally significant impact on their academic performance and involvement in internships. This indicates that a region’s education and economy affect students’ access to educational resources and their beliefs and awareness. This will have a certain impact on their subsequent academic performance and internships.

4.3. Analysis of the Impact Mechanism: Test of the Mediation Effect

The previous analysis shows that family characteristics are positively related to students’ performance at university. In accordance with Hypothesis 2a and Hypothesis 2b, put forth in the theoretical basis section, this paper posits the belief that family characteristics impact students’ performance at university primarily in two ways: first, students’ family characteristics directly influence their academic performance and participation in student unions, as well as part-time employment and internships; second, family characteristics build children’s characters throughout their growth process, impacting their interpersonal communication skills when they become university students, which in turn affects their academic performance. In order to test the above two mechanisms, the paper uses the mediation effect to check whether the above two paths are valid. Based on this, this paper constructs the following model:

$${\mathrm{coumm}}_{\mathrm{i}}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_{1,\mathrm{j}}{\displaystyle \sum _{\mathrm{j}=1}^3{\mathrm{C}}_{\mathrm{j},\mathrm{i}}}+{\mathsf{\alpha }}_2{\mathrm{Control}}_{\mathrm{i}}+{\mathsf{\epsilon }}_{\mathrm{i}}$$

(5)

$${\mathrm{sty}}_{\mathrm{i}}{/\mathrm{union}\_\mathrm{a}}_{\mathrm{i}}{/\mathrm{intership}}_{\mathrm{i}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{coumm}}_{\mathrm{i}}+{\mathsf{\beta }}_2{\mathrm{Control}}_{\mathrm{i}}+{\mathsf{\epsilon }}_{\mathrm{i}}$$

(6)

$${\mathrm{sty}}_{\mathrm{i}}{/\mathrm{union}\_\mathrm{a}}_{\mathrm{i}}{/\mathrm{intership}}_{\mathrm{i}}={\mathsf{\gamma }}_0+{\mathsf{\gamma }}_{1,\mathrm{j}}{\displaystyle \sum _{\mathrm{j}=1}^3{\mathrm{C}}_{\mathrm{j},\mathrm{i}}}+{\mathsf{\gamma }}_2{\mathrm{coumm}}_{\mathrm{i}}+{\mathsf{\gamma }}_3{\mathrm{Control}}_{\mathrm{i}}+{\mathsf{\epsilon }}_{\mathrm{i}}$$

(7)

Model (2) is used to describe the estimated coefficient of the mediating variable interpersonal communication (coumm), i.e., the estimated influence of the independent variable—university students’ family characteristics—on interpersonal communication; Model (3) represents the influence of the mediating variable on the explained variable, i.e., the estimated value of how coumm influence students’ performance at university (sty, union_a, internship); Model (4) focuses on the influence of independent variables (family characteristics)—C1, C2, and C3—on the dependent variable—university students’ performance at university—after the mediating variable coumm is added to the regression equation. The regression results of the binary logistic model are as follows (see

Table 7. Mediating effect test.

Dependent Variables	sty	union_a	Internship
${\alpha }_{11}$	1.7975 *** (2.91)	3.4644 *** (9.79)	1.3784 * (1.90)
${\alpha }_{12}$	0.5837 *** (−3.16)	0.4351 *** (−7.33)	0.7345 ** (−2.14)
${\alpha }_{13}$	0.8487 *** (−2.13)	0.5959 *** (−9.71)	0.8531 *** (−2.46)
${\beta }_1$	1.2246 (0.99)	1.4282 * (1.70)	1.3910 * (1.90)
${\gamma }_{11}$	1.0873 (0.54)	1.5094 ** (2.63)	1.4254 ** (2.08)
${\gamma }_{12}$	0.8004 * (−1.75)	0.7582 ** (−2.15)	0.7202 ** (−2.26)
${\gamma }_{13}$	0.9240 (−1.35)	0.7598 *** (−4.39)	0.8425 ** (−2.62)
N	846	846	846

Note: Data from CFPS 2018. The values within the brackets are z values; * indicates significance above the 10% level, ** indicates significance above the 5% level, and *** indicates significance above the 1% level.

). For the dependent variable—participation in students unions (union_a)—the mediation effect is tested. If the coefficient is significantly below the 1% level, it indicates that the dependent variable—family characteristics—has a positive impact on interpersonal communication; if the coefficient ${\gamma }_{1j}$ is significantly below the 10% level, it indicates that university students’ interpersonal communication ability (coumm) will positively affect students’ participation in student unions; if the coefficient is significantly below the 5% level, it indicates that family characteristics would motivate students’ participation in students unions by exerting a positive influence on interpersonal communication, during which interpersonal communication plays a mediation role between family characteristics and students’ participation in student unions. Similarly, for part-time internships, when ${\alpha }_{1j}$ is significantly below 10%, 5%, and 1%, ${\beta }_1$ is significantly below the 10% level, and ${\gamma }_{1j}$ is significantly below the 5% level. This also indicates that interpersonal communication plays a mediating role between the dependent variable—family characteristics—and the explained variable—internships and part-time employment. It is worth noting that for academic performance sty, coefficients ${\beta }_1$ and ${\gamma }_{1j}$ are both practically insignificant. This proves that interpersonal communication has a minimal impact on the academic performance of university students, be it a separate or a mediating variable.

4.4. Heterogeneity Analysis

People’s innate characteristics (gender) often affect their personality, habits, and thinking patterns, etc., resulting in differences between gender groups, which thus affects students’ interpersonal communication skills and performance at university. Based on that, this paper conducts a heterogeneity analysis of the gender of university students. The results are shown in

Table 8. Heterogeneity analysis results.

Variables	Coumm		sty		union_a		Internship
	Male	Female	Male	Female	Male	Female	Male	Female
C1	2.519 ** (2.84)	1.495 (1.56)	1.126 (0.50)	3.169 *** (7.56)	2.079 ** (3.10)	4.231 *** (9.30)	2.096 ** (2.81)	1.020 (0.09)
C2	0.472 ** (−2.64)	0.669 * (−1.87)	0.764 (−1.42)	0.419 *** (−6.01)	0.537 ** (−3.20)	0.414 *** (−6.10)	0.829 (−0.84)	0.679 ** (−2.02)
C3	0.795 * (−1.87)	0.875 (−1.35)	0.871 (−1.54)	0.683 *** (−6.16)	0.718 *** (−3.53)	0.552 *** (−8.93)	0.705 *** (−3.25)	0.955 (−0.55)
Other control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	383	463	383	463	383	463	383	463
Adj R-square	0.0470	0.0121	0.0203	0.1127	0.0356	0.1682	0.0369	0.0136

Note: Data from CFPS 2018. In the brackets are z values; * indicates significance above the 10% level, ** indicates significance above the 5% level, and *** indicates significance above the 1% level.

.

For the interpersonal communication skills of university students, the positive effects of family characteristics C1, C2, and C3 are more evident in male university students, while they are not evident in female university students. The reason is probably that female students tend to engage in interpersonal communication for the purpose of interaction and self-improvement. For students’ performance at university, gender difference has a significant impact on academic performance sty, as the academic performance of female university students is significantly affected by their family characteristics; for students’ participation in student unions, family characteristics have a significant impact on the dependent variable union_a among both male and female university students; and in terms of part-time employment and internships, family characteristics are more positively significant for male university students than for female university students.

4.5. Robustness Test

To verify that the results of the empirical part of the main body are universal rather than caused by specific variables or selected values, this paper conducts a robustness test for the aforementioned results.

This paper mainly tests the robustness of the results by replacing the core explanatory variables.Descriptive statistics for substitution variables are shown in

Table 9. Statistical characteristics of the robustness test of family characteristics where explanatory variables are replaced.

Variables			Sample Size	Average	Standard Deviation	Minimum Value	Maximum Value
Explanatory variable	Frequency of contact with parents	connect_p	1037	0.7015	0.4138	0	1
	Logarithm of annual household income	income_l	1037	10.2141	2.6297	0	14.2855

Data from CFPS 2018.

. The robustness test results are shown in

Table 10. Robustness test results.

Variable	Coumm	sty	union_a	Internship
connect_p	3.2353 ** (4.13)	5.5185 *** (8.74)	7.1599 *** (10.38)	1.5723 * (1.76)
income_l	0.8918 * (−1.72)	0.9204 ** (−2.50)	0.8395 *** (−4.39)	0.8847 ** (−2.54)
other control variables	Yes	Yes	Yes	Yes
N	846	1000	1000	754
Adj R-square	0.0350	0.0739	0.1162	0.0151

Note: Data from CFPS 2018. The values within the brackets are z values; * indicates significance above the 10% level, ** indicates significance above the 5% level, and *** indicates significance above the 1% level.

. The robustness of the family characteristics of university students is tested in this way: Drawing on the existing literature and practices, and using the survey data of CFPS in 2018, this paper selects a representative variable from both a family’s internal factors and external capital, which are “frequency of connection with parents in the past month (connect_p)” and “annual family income log value (income_l)”, to replace the original core explanatory variables C1, C2, and C3. The test is conducted by repeating the aforementioned regression steps through the use of substitution variables while retaining other conditions.

Robustness test results are shown in Table 10. By replacing core explanatory variable, connect_p and income_l, representative university students’ family characteristics all have a significantly positive impact on coumm, sty, union_a, and internship. In other words, families with good communication atmospheres and high-income levels facilitate university students’ quality of interpersonal communication, academic performance, degree of participation in student unions, and internship and part-time experiences. The robustness test results further support the conclusion in the empirical part of the main body. The average marginal effect of different groups based on robustness test results is shown in Figure 3.

5. Discussions

Family upbringing is an important mechanism of intergenerational reproduction. Based on the Bourdieu–Coleman Theory and from the perspective of social capital and intergenerational mobility, this paper discusses the influence of family characteristics on the quality of university students’ interpersonal communication and school behavior, and its mechanism. The results show that:

First, the comprehensive level of family characteristics has a significant positive impact on the quality of interpersonal communication in university students. This verifies the conjecture of Hypothesis 1and further proves that the parenting style of Chinese families presents remarkable class differences. Families with a higher comprehensive level of characteristics pay more attention to the cultivation of emotional expression and comprehensive quality of university students in the growth process, and often have a closer parent–child relationship. At present, the family background of university students in China is in a state of differentiation. Factors such as the educational background and occupation of the parents decide the family living standard to a certain extent, and then affect the richness of family cultural activities and family cultural resources. From the material level, families of different income also have diversified intensity of cultural investment in the growth of university students, and the “cultural reproduction” mechanism still plays an important role. Family external capital will be transmitted between parents and children through education investment and training opportunities, affecting children’s learning ability, communication ability and organization and coordination abilities, as well as maintaining the existing social class [10,47].

Second, family background will have an impact on university students’ academic performance and participation in various activities, which is consistent with the previous Hypothesis 2a. In an emotionally warm family, social capital has far more influence on the development of university students in the aspects of its significance, universality, and effectiveness than family economic capital and social capital outside of the family. The positive effects of the latter two kinds of family capital on university students’ ability development are not as significant as those brought about by family communication. In the emotional warmth parenting style, on the one hand, parents effectively convey the mainstream values and expectations of society to their children through educational expectations; on the other hand, emotional encouragement and support can help children form excellent reflective and creative abilities. This sort of social interaction within the family, based on personality equality, has good compatibility with the basic concept of modern society. The cultural training children received at home gives them a natural “advantage” compared with others in social organizations.

Thirdly, the quality of university students’ interpersonal communication will have a mesmeric effect on their school performance, which verifies the conjecture of Hypothesis 2b. According to the results, university students with higher interpersonal quality tend to be more outgoing and enthusiastic about participating in student organizations and internships. With the increase of interaction frequency between the level of external and internal family capital, there will be positive impact on the quality of university students’ interpersonal communication. Resultingly, higher-quality interpersonal communication usually provide students with more experiences of participation in organizations, internships, and part-time work. This finding has certain significance for us in understanding the intergenerational mobility and personality shaping of university students in China.

6. Conclusions and Recommendations

From the above analysis, it can be concluded that, compared with the family living standard, the significance of family cultural capital in the socialization process of university students is highlighted. This responds to the research controversy in academic circles: the gap between factors such as family living standard and family socioeconomic status with stronger ascribed significance is being compensated by cultural capital factors, such as family cultural activities and family cultural resources, and the decisive role of the family living standard is being weakened. In other words, through the accumulation of family cultural capital, university students’ professional knowledge literacy will be effectively promoted and improved. From the perspective of administrators, this study further emphasizes the importance of “home-school co-construction” and provides support for talent training.

Based on the analysis above, the paper further gives the following recommendations from three dimensions: individual, family, and government:

At the personal level, we should encourage university students to get rid of the view of “hierarchy solidification”, that family factors determine the development opportunities and social status of the younger generation, and it is difficult for individuals to change their social classes through their own efforts. University students should pay attention to improving their quality of interpersonal communication during school, actively participate in various student unions, and improve their abilities in all aspects.

At the family level, the core and crux of the phenomenon of “poor families having difficulty cultivating outstanding children” lies in the lack of social interaction within the family based on understanding, warmth, trust, encouragement, respect, and rules, which parents of poor families lack for various subjective and objective reasons compared to their better-off counterparts [48]. A good family environment (such as warm family communication style and family economic foundation) affects children’s ability development through the internal social capital shaped by an emotionally warm parenting style [49]. Although it is difficult for parents of disadvantaged families to change their own educational level, occupation, and family economic situation in short order, they can still create a good parent–child interaction atmosphere within the family by changing their parenting style, which can also make up for the adverse effects of other families’ lack of capital on their children to a certain extent. In this way, children will have enough “family foundation” to gradually accumulate advantages in each stage of education, and eventually achieve upward mobility. Therefore, families should not only pay attention to physical capital when investing in their children’s education, but also enhance the frequency of communication within the family with emotional warmth toward their children, so that they can have a subtle influence on their children through a good family atmosphere.

At the government and administration level, in order to enable poor families to cultivate outstanding children, the state and social macro system should also pay more attention to the poor families and care for them while valuing the motivation of the poor parents themselves. Promoting intergenerational mobility requires both internal and external efforts to better address the intergenerational mobility of the underprivileged in society. The government needs to focus on family education and increase the investment in it, especially for disadvantaged families. The Internet channel and community knowledge popularization can be used to expand publicity, so that more families at the bottom of society realize that the basis for the good development of children comes from the emotional warmth, encouragement, and support of parents within the family. At the same time, it is necessary to provide easier access to learning resources, training platforms, and opportunities for these families; publicize the concept of reasonable education; and provide operational education methods. From the perspective of business and management, a reduction the interference of family capital on the labor market in talent recruitment and selection, and an enhancement of the transparency and justice of personnel policies and practices are recommended. More businesses would be improved by providing jobs to poor graduate students, helping them solve economic difficulties, and preventing them from giving up better development opportunities due to family economic pressure.

There are still limitations in this study. For one thing, the study emphasizes the influence of family characteristics on university students, not negating the influence of other factors on the shaping of university students, such as the system and cultural shaping transmitted by the school itself [50] and university students’ own subjective initiative differences, etc. [51]. Due to limited research samples, these mechanisms have not been investigated in depth, remaining open to further study. Furthermore, the official CFPS database as a whole was most recently updated in the year 2018. The personal library for the year 2020 was launched, instead of a family database and other datasets. Samples can be extended to various regions, demographics, and even to other countries as well. Limited access to research data hinders the study from using the latest data for further research, which remains open to continuous study after the perfection of more official data sets.