Parental Academic Involvement in Cross-border marriage,Native and Immigrant Families: The Roles of Family Resources and Parental Expectations

Abstract

Migration patterns have become more complicated than before. The increase in cross-border marriage has increased the diversity of the student population. A simple distinction between native and immigrant students overly simplifies their diverse backgrounds. Analyzing data from PISA 2012, we attempt to fill the gap in the literature by comparing parental involvement between cross-border marriage, native, and immigrant families in Hong Kong, with a special focus on the roles of family resources and parental expectations. Our findings show that cross-border marriage families are significantly different from and sit between native families and immigrant families in terms of family resources, parental expectations, and parental involvement. Family resources play a greater role than parental expectations in the differences in parental involvement between cross-border marriage families, native families, and migrant families. This suggests that the disadvantages of cross-border marriage and immigrant families in parental involvement largely stem from a lack of family resources.

Keywords

Parental involvement cross-border marriages immigrant families nativity gap Hong Kong

Introduction

Cross-border marriages are those in which a locally born resident marries a migrant from elsewhere. There has been a dramatic rise in cross-border marriages in many Asian countries over the last few decades (Yeung & Mu, 2019). This has significantly increased the complexity of the population structure in these societies. Not only do a significant proportion of all marriages in these societies now involve a local resident marrying a migrant, but a large share of the children in the population also have cross-border couples as parents (Jones & Shen, 2008). Previous studies have shown that couples in cross-border marriages in East Asian societies are less well educated and more likely to have a lower-class background than couples comprising two local residents (Cheung & Chiu, 2020; Choi & Cheung, 2017; Constable, 2010; Lin & Ma, 2008; Morgan & Hoffmann, 2007). However, some scholars have argued that compared with local parents, immigrants have stronger motivations and higher expectations for their children (Portes & Rumbaut, 2001; Xu & Wu, 2017). However, research has paid insufficient attention to how the structural disadvantages and higher parental expectations said to characterize cross-border marriages may, in turn, affect parental involvement in children’s education. This is an important oversight, because parental academic involvement has profound implications for children’s educational and behavioral outcomes (Cheung & Park, 2016; Ho & Willms, 1996; McNeal, 1999; Park, 2006; Park et al., 2011; Zhang et al., 2019).

In the U.S. context, Ramakrishnan (2004) argued that students who are children of couples in cross-border marriages deserve special attention and should be distinguished from both native and immigrant students because these students are different from the children from native and immigrant families. However, despite the differences, past studies have often grouped the children of cross-border marriage families into either “native” or “immigrant” groups. In some studies, these children have been considered “natives,” which typically refers to individuals with two native parents (Ho, 2006). In other studies, however, the children of cross-border marriage families have been treated as “immigrants” (Xu & Wu, 2017). We argue that the simple distinction between immigrant and native students, which has been commonly used in the literature (Hastedt, 2016; Schwartz & Stiefel, 2006), may not well suit a context in which the proportion of marriages involving cross-border couples has risen substantially (Ramakrishnan, 2004). In particular, it is unclear in the literature whether cross-border families and immigrant families are similarly disadvantaged but highly motivated. There has also been a lack of empirical studies focusing on differences in their parenting practices.

Parental academic involvement includes parental activities across different dimensions depending on activity location and content, such as involvement in children’s education at home and in school. Parental academic involvement at home and in school could be driven by different factors (Park et al., 2011). In this study, we extend the previous literature by comparing the patterns of parental academic involvement at home of cross-border marriage families with those of native and immigrant families in Hong Kong, where the share of cross-border marriages has long been substantial.

Drawing on the resource perspective (Downey, 2002; Ho, 2006; Lareau, 1989) and parental expectations perspective (Duran & Weffer, 1992; Portes & Rumbaut, 2001; Xu & Wu, 2017), the current study examines the relationship between cross-border marriages and parental academic involvement at home in Hong Kong, and its potential mechanisms. We contribute to the literature by examining the following three research questions. (a) How do families of cross-border marriages differ from native and immigrant families in parental involvement at home? (b) How can the differences in parental academic involvement at home among these different groups of families be explained by family resources? (3) How can the differences in parental involvement among these different groups of families be explained by parental education expectations for their children? To answer these questions, we analyze representative survey data from the Programme for International Student Assessment (PISA) 2012 to compare the pattern of parental involvement among native, cross-border, and immigrant families.

Literature and Conceptual Framework

Cross-Border Marriages in Hong Kong

The rise of cross-border marriages is an important demographic feature in Asia (Yeung & Mu, 2019). Cross-border marriages in Asia are very often spatially hypergamous, as they typically involve women from less developed countries marrying men in more developed countries (Constable, 2010). The “marriage squeeze” in the local marriage market caused by increasing educational opportunities for women drives socially disadvantaged men, such as those who are divorced, old, and/or poorly educated, to seek partners from less developed countries (Raymo & Park, 2020). Due to the strengthening of economic ties between mainland China and Hong Kong, the number of marriages registered in Hong Kong that involved Hong Kong men and mainland Chinese women rose by 75% from 1986 to 2006. Such marriages account for approximately two-fifths of all marriages registered in Hong Kong (Hong Kong Census and Statistics Department, 2015; Pong et al., 2014). Children from cross-border marriage families now make up a significant proportion of the student population in Hong Kong. Table 1 shows the trend in the composition of the student population in Hong Kong by nativity and migration background, based on PISA data on 15-year-old students from 2000 to 2018. Among all family types, cross-border marriage families are the only family type whose share rose between 2000 and 2018.

Table 1.

Migration/Nativity Background of Hong Kong Students from the PISA Data 2000–2018.

	2000	2003	2006	2009	2012	2015	2018	Δ2000-2018
Native families	36.2%	38.8%	38.4%	38.8%	40.3%	36.8%	34.6%	−1.6 pct. pts.
Families of cross-border marriage	18.9%	17.6%	17.1%	20.2%	23.6%	27.6%	27.2%	8.3 pct. pts.
1^st generation immigrants	17.5%	20.4%	19.1%	15.5%	14.1%	13.8%	12.8%	−4.7 pct. pts.
2^nd generation immigrants	27.1%	22.9%	24.5%	23.8%	20.4%	21.3%	25.0%	−2.1 pct. pts.
Return migrants	0.3%	0.4%	0.8%	1.6%	1.6%	0.6%	0.4%	0.1 pct. pts.

Source: Authors’ calculation with PISA data 2000 – 2018.

Notes. Native families = locally born students with both parents born locally. Families of cross-border marriage = Students with one locally born parent and one foreign-born parent. 1^st generation immigrants = foreign-born students with both foreign-born parents. 2^nd generation immigrants = locally born students with both foreign-born parents. Return migrants = Foreign-born students with both parents born locally.

As of 2012, locally born students with two locally born parents are still the largest group (40.3%) in the student population, but they do not constitute the overwhelming majority. As Hong Kong is an immigrant city, students with migrant backgrounds make up the majority of the student population. Among them, a significant proportion of students are from cross-border marriage families; they comprise a quarter of the total student population in Hong Kong. Among the 65 countries and regions participating in PISA 2012, Hong Kong ranked second in terms of the percentage of 15-year-old students in cross-border marriage families. Nonetheless, Hong Kong is not set apart from other highly developed economies in the region. There is a similarly high percentage of students in cross-border families in some other Asian cities, including Singapore (19.9%) and Macao (18.7%).

Despite the rising significance of cross-border marriage families in the student population, many studies have continued to adopt a dichotomous (native versus immigrants) or trichotomous (native versus first-generation and second-generation immigrants) categorization approach to comparing families. Previous research has tended to make simple distinctions between natives and immigrants without paying attention to the children of cross-border marriages (e.g., Cattaneo & Wolter, 2015; Hastedt, 2016; Pong, 2009; Schwartz & Stiefel, 2006). For example, Ho (2006) compared the parental involvement of native families with that of immigrant families. She found a lower level of parental involvement for immigrant students, but students with one parent born outside of Hong Kong were grouped among the “native” families. In another study, Xu and Wu (2017) compared education expectations and academic achievement between native students and second-generation immigrant students. In contrast with Ho (2006), Xu and Wu (2017) classified students with one parent born outside Hong Kong as second-generation immigrants, and found that this group had higher expectations and better academic achievement than their native peers.

These studies made the implicit assumption that the major characteristics of cross-border marriage families are similar to those of either native or immigrant families. This assumption was also made in the original PISA report (OECD, 2014), in which all children with one locally born parent were categorized as native students. Using PISA data, some scholars have compared the performance of native students with first-generation and second-generation immigrants (E.g., Mantovani & Gasperoni, 2018; Rangvid, 2007). As these studies adopted the definition used by the PISA, cross-border marriage families were treated as native families. However, this assumption is misleading if cross-border marriage families are substantively different from native families. Indeed, this assumption has already been challenged in the United States (Ramakrishnan, 2004).

As a form of “global hypergamy” (Constable, 2010), cross-border marriages in Hong Kong mostly involve socioeconomically disadvantaged local men who experience difficulty in finding local partners and are therefore motivated to look for marital partners in mainland China (Choi & Cheung, 2017; Lin & Ma, 2008; Pong et al., 2014). Compared with men who marry local women, men who marry women from mainland China generally have lower levels of education and income, and are more likely to be unskilled workers (Hong Kong Census and Statistics Department, 1999). In terms of socioeconomic background, cross-border marriage families are in general disadvantaged compared with native families in which the parents are both local.

In Hong Kong, as in many other contexts, immigrant families in which both parents were foreign born are also socially disadvantaged (Zhang & Wu, 2011). The parents in these families tend to have lower levels of educational attainment, less cultural and social capital, and lower-status occupations (Garcia Coll et al., 2002; Zhang & Ye, 2018). It remains unclear whether cross-border marriage families are as structurally disadvantaged as these immigrant families. In particular, we have little knowledge of how cross-border families are different from first- and second-generation immigrant families.

Immigrants are a self-selected group who often have a strong motivation for a better life and high expectations for their offspring (Borjas, 1987; Kao & Tienda, 1998; Zhang et al., 2019). Past studies have found that students and parents from immigrant families often have higher educational and occupational expectations than their native counterparts (Zhang et al., 2019). It is doubtful that parents in cross-border marriages have the same high level of expectations as immigrant families, because the locally born parent in a cross-border marriage may not have the same “immigrant drive” seen among immigrant families (Borjas, 1987). We also have little knowledge of whether cross-border families are a self-selected group of highly aspirational parents.

In short, despite their significance in the population composition, we still have limited knowledge of how cross-border marriage families fare in comparison with native and immigrant families, respectively. In particular, little research has compared parental academic involvement at home in cross-border marriage families with that in native and other immigrant families. Parental academic involvement at home has an important influence on children’s educational outcomes, and therefore has profound implications for the inter-generational social mobility of immigrant families (McNeal, 1999; Park, 2006; Park et al., 2011; Hamilton, Roksa & Nielson, 2018).

Parental Academic Involvement: The Role of Family Resources and Parental Education Expectations

One of the major perspectives on the social disparities in parental academic involvement is the resource perspective (Lareau 1989, 2002). Parental academic involvement requires access to resources that enable parents to be involved in their children’s education. This perspective suggests that economic and cultural resources are positively associated with parental involvement. Families with more resources tend to exhibit more active and frequent parental involvement (Astone & McLanahan, 1991; McNeal, 1999; Ho, 2006; Hamilton et al., 2018). Although they also want to be supportive, working-class parents are more likely than middle-class parents to be constrained from being involved in their children’s education by disadvantaged socioeconomic conditions and a relative lack of cultural capital (Ho, 2006; Lareau, 1989). Family resources include the “know-how” crucial for parental involvement. Parents with more resources have more cultural capital, and thus have more knowledge of what to expect when they become involved in their children’s education. They hold more information about the educational system and are more capable of providing support than parents with fewer resources (Lareau, 2002). Empirical studies have confirmed that middle-class parents are more involved in children’s school and daily activities than working-class parents in many contexts (Ho, 2006; Lareau, 2002).

The parental expectations perspective also helps to explain the social disparities in parental involvement. This perspective suggests that the educational expectations that parents have for their children are an influential factor in parental involvement. Parental academic involvement requires effort on the part of parents. Motivated parents are more willing to put effort into their children’s education. Net of the effect of family resources, parents with higher expectations for their children’s educational attainment have a higher level of involvement in their children’s education (Englund et al., 2004). Parents with higher expectations attach greater value to their involvement in their children’s education and are more motivated to invest time in it (Yamamoto & Holloway, 2010). A positive relationship between expectations and involvement has been found in many Western societies (Englund et al., 2004, Sy et al., 2005). However, this relationship is not always consistent across groups of families. For example, Sy et al. (2007) found a positive correlation between parental education expectations and involvement among European Americans, but no direct relationship among Asian Americans. Using another sample, Sy and Schulenberg (2005) even found that low parental expectations triggered more parental involvement. Hence, the relationship between parental expectations and parental involvement may be culturally and contextually dependent.

Parental Academic Involvement in Native, Cross-Border Marriage, and Immigrant Families

Based on the aforementioned literature, we expect that family resources and parental education expectations are both positively associated with parental involvement. We also expect native families to have more resources than cross-border marriage and immigrant families, and immigrant families to have higher expectations than cross-border marriage and native families. The resource levels and parental expectations of cross-border marriage families sit between those of native and immigrant families. Hence, we further expect that differences in family resources and parental educational expectations among cross-border marriage, immigrant, and native families may at least partly account for the different levels of parental involvement among the three groups.

Regarding the difference between native and cross-border marriage families, we hypothesize family resources as a reinforcing mechanism between nativity and parental involvement. We expect that the difference in parental involvement between cross-border marriage and native families will be reduced after controlling for family resources. In contrast, parental expectations may be a suppressing mechanism in the relationship between nativity and parental involvement. We expect the difference in parental involvement between cross-border marriage families and native families to be larger after controlling for parental expectations. The differences between cross-border marriage and immigrant families are a rarely studied topic. As there is one local parent in each cross-border marriage family, we expect parents from cross-border marriage families to have more resources but less motivation than their counterparts from immigrant families. We, therefore, hypothesize that the same reinforcing and suppressing mechanisms may influence the difference in parental involvement between cross-border marriage families and immigrant families.

The actual difference in parental involvement among family types depends on the validity as well as the relative strength of the two mechanisms. If family resources play a major role in shaping parental involvement, we expect native families to have a higher level of involvement. In contrast, if parental expectations play a more significant role in shaping parental involvement, then cross-border marriage families and immigrant families may have a higher level of parental involvement than native families.

Besides, the parents from cross-border marriage and immigrant families may not be as capable of involvement in their children’s education as native parents, who are more familiar with the local education system and educational materials. We hypothesize that parents from cross-border marriage families and immigrant families are less capable of being involved in their children’s education, and hence spend less time talking to their children about school-related matters, after controlling for resources and expectations. Although there is one parent from cross-border marriage families growing up in Hong Kong, often the mothers in cross-border marriage families are the immigrants who are less familiar with the local education system. Since mothers are more typically in charge of children’s education than fathers, we expect parents from cross-border marriage families and immigrant families are similarly disadvantaged in parental academic involvement after controlling resources and expectations.

Research Methods

Data

In this study, we analyzed data from PISA Hong Kong. Administered by the Organization for Economic Cooperation and Development (OECD), PISA is a triennial international survey measuring the literacy skills of 15-year-old secondary school students in the participating countries and regions. In total, seven waves of PISA data are publicly available with the most updated version collected in 2018. Some important variables such as family structure, parents’ employment statuses, and living with grandparents, which are likely related to both immigrant status and parental involvement, are not available in PISA 2015 and PISA 2018. Also, information about parental education expectations is not available in PISA 2015. Therefore, we analyze data from PISA 2012 in this study. To check the robustness of our findings, we have also conducted additional analyses with PISA 2018 data, replicating our regression analyses without controlling some background variables that were available only in PISA 2012 but not in PISA 2018 (results not shown in the tables). We confirm that our main findings remain the same in both PISA 2012 and PISA 2018 data.

The PISA survey uses a two-stage stratified sampling procedure to obtain a representative sample. For PISA Hong Kong 2012, 4,670 students were sampled from 148 schools. The survey consists of both student and parent questionnaires. In this study, information on parental involvement with and expectations for children’s education were taken from the parent questionnaire, with all other information taken from the student questionnaire (HKPISA Centre, 2013).

Dependent Variables

Parental academic involvement at home

The dependent variables in this study were self-reported items related to parental academic involvement at home. In PISA, parents were asked how frequently they provided support for their children’s learning. In the survey, seven aspects of parental involvement at home were collected: discussing school well-being with their children, helping their children with mathematic homework, discussing how their children perform in mathematics classes, obtaining mathematics materials for their children, discussing with their children how mathematics can be applied in everyday life, eating main meal, and spending time talking. Each of these is an ordinal variable indicating the frequency of the specific item, chosen from never or hardly ever (1), once or twice a year (2), once or twice a month (3), once or twice a week (4), and every day or almost every day (5). The original PISA data has provided a validated scale on parental support based on Item-Response Theory (IRT) analysis of seven items. However, two of the items included in the original scale refer to parent-child interactions that are not relevant to academic involvement—“eating main meal” and “spending time talking.” Hence, we followed the PISA’s procedure in constructing the scale of parental academic involvement at home based on IRT analysis of the first five items, which are only relevant to parental academic involvement at home. The constructed scale is a predicted score from the generalized partial credit model with a mean of 0 and a standard deviation of 1 (The detailed procedure in forming the IRT scale could be found in PISA technical report). Cronbach’s alpha for the scale was 0.82, indicating a high level of internal reliability, and close to the original PISA scale (α = 0.80).

Independent Variables

Nativity status

Based on the birthplaces of the students and their parents, we classified the students into five migration status groups. In this study, “native families” refer to those in which students and both parents were born in Hong Kong; first-generation immigrant families are those in which students and both parents were born outside of Hong Kong; and second-generation immigrant families are those in which students were born locally and both parents were born outside of Hong Kong.

Locally born or foreign-born students with one locally born parent and one parent born outside Hong Kong were classified as cross-border marriage families. In the Hong Kong context, due to its unique migration policy, students born locally or born elsewhere with one foreign-born parent are not considered very different and can thus be regarded as a single category of students. We, therefore, did not further separate locally born students with one foreign-born parent and foreign-born students with one foreign-born parent into two separate groups of cross-border marriage families. We conducted additional analyses and confirmed that differences in parents’ background (such as education level), parental expectations, and parental academic involvement between locally born and foreign-born students from cross-border marriage families were not statistically significant. In the Hong Kong context, the foreign-born parent in a cross-border marriage is typically the mother. In our data, 75% of the foreign-born parents in cross-border families were mothers.

Lastly, foreign-born students with two locally born parents were categorized as return migrants. In the Hong Kong context, the parents of some children are temporary migrant workers who are well-educated working professionals (Fong, 2011). This category is small in number, but it is inappropriate to group them with other immigrants, who are often disadvantaged. Hence, we included this group of families as a standalone category in the analysis, although they are not the focus of this study.

Family resources

We used the PISA index of economic, social, and cultural status (ESCS), which measures the family resource level of the sampled students. This validated scale has been widely used in educational research based on PISA data. The ESCS index is a composite score based on parents’ years of schooling, their occupations’ social status (ISEI), and the possession at home of 16 items indicating the level of family wealth (such as car and a room for the student), home educational resources (such as a desk to study at, a quiet place to study, and a computer for schoolwork), and cultural resources (such as books and musical instruments). The alpha reliability of the scale is 0.76. For details of the construction of the scale, please consult PISA’s technical report (OECD, 2014). We rescaled the index into percentiles for ease of interpretation.

Parental tertiary education expectations

Another intermediate variable in this study was the parents’ educational expectations of their children. The parents were asked to indicate the highest level of education that they expected their children to complete, with the answers ranging from 1 (junior secondary, ISCED level 2) to 6 (Bachelor’s degree or above, ISCED level 5A or 6). Since the frequency distribution of these categories is unbalanced with few responses from level 3 and below, we regrouped the categories into a dichotomous variable (1 = complete tertiary education (ISCED Level 5 or above), 0 = otherwise).

Control variables

In order to examine the net association between nativity status and parental involvement at home, we controlled for several background variables that are thought to be correlated with the nativity status and parental involvement. First, student characteristics such as gender (female = 0; male = 1) amd school grade (below Grade 10 = 0; Grade 10 or above = 1) are controlled. School grade is controlled because past studies found that parents of immigrants students in Hong Kong tend to put their children into a lower grade, known as academic “redshirting,” which could also be related to family resources, parental expectations, and parental academic involvement (Pong, 2009). Family characteristics such as family structure (two-parent family = 1; single-mother family = 2; single-father family = 3; others = 4), co-residence with sibling (no = 0; yes = 1), and co-residence with grandparents (no = 0; yes = 1) are also controlled. Family structure is highly correlated with cross-border marriages. Past studies found that cross-border married couples are more likely to experience marital conflict and disruption (Cheung & Chiu, 2020; Choi & Cheung, 2017). The differences in parental involvement among family types could be due to the higher proportion of single-parent families in immigrant and cross-border marriage families (Cheung & Park, 2016). Besides, more children could make parental involvement more difficult and depress parent’s educational expectations for a given child (Downey, 1995; Wei et al., 2016). Meanwhile, the expected fertility is lower among native parents in Hong Kong, as compared to immigrants. Grandparents can affect parenting and its influences on children (Akhtar, Malik, & Begeer, 2017; Bruce-Edditing, 2020). PISA 2012 does not have information on the degree and quality of interactions with grandparents but only whether the student lives with a grandparent. Note that immigrant children are less likely to live with grandparents because the grandparents may not be living in Hong Kong.

Lastly, parental involvement may be encouraged more by some schools than by others (Ho, 2006). Hence, we also controlled for school fixed effects by including dummy variables for the sampled schools. By controlling for the school fixed effects, it allows us to examine whether the differences in parental involvement among family types exist within schools while ignoring the between-school differences.

Analytical Strategy

In the first step of our multivariate analysis, we ran a linear regression model and logistic regression models, respectively, to compare the family resources and parental expectations of cross-border marriage, native, and immigrant families, controlling for the background variables. Since the logit coefficients could not be compared across models and the interpretation of logit coefficients could be tricky and non-intuitive (Mood, 2010), we choose to present the average marginal effects (AMEs) derived from the logistic regression models for comparing parental expectations between cross-border marriage, native, and immigrant families. AMEs are the estimated changes in the probability of expecting their children to complete tertiary education when the independent variable increases by one unit (for continuous variables), or the difference in the probability between the group of interest and the reference category (for nominal variables), controlling for other covariates. Hence, the AMEs could be interpreted in a way very similar to how results of linear regression models are interpreted.

In the second step, we estimated five nested regression models to determine how various factors accounted for the differences in parental academic involvement for students of different family types. In the baseline model (Model 1), we estimated the association between cross-border marriage status and parental involvement, controlling for the background variables only. In Model 2, we also controlled for family resources. In Model 3, we controlled for parental expectations and the background variables, but not for family resources. In Model 4, we controlled for both family resources and parental expectations, along with the background variables. Since different schools may have used different strategies and spent various amounts of resources in encouraging parental academic involvement at home, we additionally controlled the school fixed-effects in the final model to examine the within-school differences in parental involvement for students of different family types.

Taking the complex sample design of PISA into account, all of the analyses were weighted, and all of the variances were cluster-adjusted using balanced repeated replication. There were some cases with missing values in the sample, mostly for co-residence status with grandparents (6.4%). We used listwise deletion to handle these cases. Approximately 13.5% of cases were excluded from the analytical sample, which consisted of 4,038 students.

Results

Descriptive Statistics

Table 2 shows the descriptive statistics of the key variables in this study. The return migrants had the most family resources, measured as the PISA index of ESCS (in percentiles), of all of the groups in the student population (85.75). As expected, the native families (63.72) had more resources than families of cross-border marriages (46.15). Compared with the cross-border marriage families, the second-generation immigrants (40.06) and first-generation immigrants (33.21) had even lower levels of family resources. In other words, the resource level of cross-border marriage families seems to be lower than that of native families and higher than that of immigrant families. The differences between the cross-border marriage families and other types of families were all statistically significant at the .001 level. A more detailed comparison of parents’ education level, which was included as an element of the PISA index of ESCS, with nativity status confirms that cross-border marriages in Hong Kong typically involve disadvantaged men marrying women who are also with a disadvantaged background. Only 14.64% and 8.40% of fathers and mothers in cross-border marriage families completed tertiary education, as compared to fathers (23.30%) and mothers (14.84%) in native families. Still, families of cross-border marriages are not as disadvantaged as first-generation (6.42% and 4.87%) and second-generation immigrant families (6.81% and 5.21%).

Table 2.

Descriptive Statistics and Bivariate Analysis: Nativity and Family Characteristics.

	Families of CBM	Natives	First-Gen Immigrants	Second-Gen Immigrants	Return Migrants
Parental involvement (M=0, SD=1)	−0.04	0.16***	−0.08	−0.18**	0.27***
Family resources (1-100 percentiles)	46.15	63.72***	33.21***	40.06***	85.75***
Parental education expectations (Tertiary =1)	76.34%	81.33%**	74.17%	81.16%***	100.00%***
Student characteristics
Grade
Below grade 10	33.94%	23.68%	68.03%	25.89%	20.04%
Grade 10 or above	66.06%	76.32%***	31.97%***	74.11%***	79.96%*
Gender
Female	49.86%	45.25%	49.15%	48.07%	40.25%
Male	50.14%	54.75%*	50.85%	51.93%	59.75%
Family characteristics
Family structure
Two-parent families	81.97%	87.29%	80.33%	87.51%	89.62%
Single-mother families	11.68%	8.52%*	14.22%	9.56%	6.23%
Single-father families	4.09%	2.14%**	3.07%	1.09%***	4.15%
Others	2.26%	2.05%	2.37%	1.83%	0.00%***
Father employed
No	15.53%	8.32%	22.41%	13.30%	9.47%
Yes	84.47%	91.68%***	77.59%***	86.70%	90.53%
Mother employed
No	40.72%	34.04%	40.54%	40.36%	35.86%
Yes	59.28%	65.96%*	59.46%	59.64%	64.14%
Living with siblings
No	32.59%	33.89%	32.06%	16.20%	37.49%
Yes	67.41%	66.11%	67.94%	83.80%	62.51%***
Living with grandparents
No	85.31%	82.92%	53.36%	87.52%	88.16%
Yes	14.69%	17.08%	16.64%	12.48%	11.84%
N	906	1546	549	785	59

Notes: ***p < .001, **p < .01, *p < .05.

Similarly, the return migrants had the highest level of parental expectations, with 100% of them expecting their children to complete tertiary education. The native families (81.33%) had higher parental expectations than the families of cross-border marriages (76.34%) (p < .01). Notably, the families of second-generation immigrants (81.16%) also had higher parental expectations than cross-border marriage families (p < .001). The first-generation immigrant families (74.17%) had a similar level of parental expectations to the cross-border marriage families.

As expected, with more resources and higher expectations, the return migrant families (0.27) and native families (0.16) had a higher level of parental involvement than the cross-border marriage families (-0.04) (p < .001). The second-generation immigrant families had fewer resources but higher expectations than cross-border families. The level of parental involvement of the second-generation immigrant families was significantly lower than that of the cross-border marriage families (p < .01). Meanwhile, the first-generation immigrant families had fewer resources than, but a similar level of parental expectations to, the cross-border marriage families. The difference in parental involvement between the two groups of families was not statistically significant.

The pattern for the control variables was not as clear. Most 15-year-old students in Hong Kong are studying in Grade 10 (senior secondary level). However, the percentage of students from cross-border marriage families studying in Grade 10 is lower than that in the overall sample. Most of the first-generation immigrants were studying at Grade 9 or below, probably due to academic redshirting. Regarding parents’ employment statuses, cross-border marriage families, first-generation and second-generation immigrant families are more disadvantaged: lower proportions of fathers and mothers were employed, as compared to parents from native and return migrant families. The students from native, second-generation immigrants, and return migrant families were more likely to be from two-parent families than the students from cross-border marriage and first-generation immigrant families, who were more likely to be from single-parent families. The students from return migrant families were less likely to have siblings living with them than the students from other family types.

In sum, cross-border marriage families are statistically different from native and second-generation immigrant families in terms of parental involvement, family resources, and parental expectations, and different from first-generation immigrant families in terms of family resources. There are also many differences in other characteristics between cross-border marriage families and other types of family.

Regression Analysis

We further examined the differences in family resources and parental expectations between cross-border marriage families and other family types when controlling for the background variables. Table 3 shows the results of linear regression in predicting family resources (ESCS index) and logistic regression in predicting parental expectations.

Table 3.

Linear and Logistic Regression Models: Coefficients and Average Marginal Effects of Nativity, Resources and Parental Tertiary Education Expectations (N = 4,038).

	OLS		Logit
	Family Resources		Parental Expectations		Parental Expectations
	Coef.	S.E.	AME^a	S.E.	AME^a	S.E.
Nativity status
Families of CBM	Ref.		Ref.		Ref.
Native families	16.22***	(1.40)	0.04*	(0.02)	−0.04*	(0.02)
First-gen. immigrants	−11.50***	(1.40)	0.01	(0.02)	0.05**	(0.02)
Second-gen. immigrants	−6.03***	(1.30)	0.05*	(0.02)	0.06***	(0.02)
Return migrants	38.06***	(1.87)	0.23***	(0.01)	0.22***	(0.01)
Family Resources	–	–	–	–	0.004***	(0.000)
Control variables
Grade
Grade 10 or above (=1)	2.93**	(1.04)	0.07***	(0.01)	0.06***	(0.01)
Gender
Male (=1)	1.68	(1.21)	0.01	(0.02)	0.00	(0.01)
Family structure
Two-parent families	Ref.		Ref.		Ref.
Single-mother families	−3.90**	(1.40)	−0.00	(0.02)	0.02	(0.02)
Single-father families	−4.92	(2.63)	−0.13*	(0.05)	−0.11*	(0.05)
Others	−7.38*	(3.27)	−0.07	(0.04)	−0.04	(0.04)
Father employed
Yes (=1)	6.55***	(1.23)	0.03	(0.02)	0.01	(0.02)
Mother employed
Yes (=1)	1.90	(0.98)	−0.01	(0.01)	−0.02	(0.01)
Living with siblings
Yes (=1)	−4.07***	(0.84)	−0.06***	(0.01)	−0.04**	(0.01)
Living with grandparents
Yes (=1)	3.22**	(1.13)	−0.02	(0.01)	−0.03*	(0.01)
Constant	39.81***	(2.02)	0.73***	(0.03)	0.50***	(0.04)
R² / Pseudo-R²	0.21		0.03		0.10

Notes: ***p < .001, **p < .01, *p < .05.

Average marginal effects are reported instead of logit coefficients because average marginal effects could be interpreted more intuitively and coefficients across nonlinear models could not be compared directly (Mood, 2010).

We first discuss the results of OLS model predicting family resources. After controlling for the background covariates, the return migrants (B = 38.06, p < .001) and native families (B = 16.22, p < .001) were significantly more resource-rich than the cross-border marriage families. The first-generation (B=-11.50, p < .001) and second-generation (B = -6.03, p < .001) immigrant families were significantly less resource-rich than the cross-border marriage families. Moving to the results of logistic regression predicting whether parents expected tertiary education for their children, the first model controlled for the background covariates but not family resources. Note that the numbers presented for the analysis of parent’s educational expectations are AMEs from logistic regression. The cross-border families were not significantly different from the first-generation immigrant families in terms of parental expectations while parents from native families (AME = 0.04, p < .05), second-generation immigrant (AME = 0.05, p < .05) and return migrant families (AME = 0.23, p < .001) have higher expectations than parents from cross-border marriage families by four, five and twenty-three percentage points, respectively.

The next model, in which family resources were included in the analysis, shows a positive association of family resources with parental expectations (AME = 0.004, p < .001). For one percentile increase in the family resources, the parents were more likely to expect their children to complete tertiary education by 0.4 percentage points. After we additionally controlled for family resources, the native families actually had a lower level of parental expectations than the cross-border marriage families by four percentage points (AME = -0.04, p < 0.05). However, both the first-generation (AME = 0.05, p < 0.01) and second-generation (AME = 0.06, p < 0.01) immigrant families had higher parental expectations than the cross-border marriage families by five and six percentage points, respectively. This finding confirms the existence of the “immigrant drive” suggested in the literature, even though it is partially suppressed by a lack of family resources. The cross-border marriage families also had higher expectations than native families, but the level of expectations was lower than that of the immigrant families in which both parents were migrants. In summary, cross-border marriage families represent a middle category between native and immigrant families in terms of both family resources and parental expectations. The R-squared values are relatively low for both models for family resources (R² = 0.21) and parental expectations (McFadden’s Pseudo-R² = 0.10), suggesting that there are high proportions of unexplained variance for both variables.

Table 4 shows the regression results comparing the level of parental involvement between cross-border marriage families and other types of families. Model 1 compared the level of parental involvement among different family types while controlling for control variables only. The return migrant (B = 0.30, p < .01) and native families (B = 0.19, p < .001) had higher levels of parental involvement than the cross-border marriage families. The second-generation immigrant families had a lower level of parental involvement than the cross-border marriage families (B = -0.12, p < .05) and represented the least involved type of family. The difference between cross-border marriage families and first-generation immigrant families was not statistically significant.

Table 4.

Linear Regression Models: Coefficients of Nativity on Parental Academic Involvement at Home (N = 4,038).

	Model 1		Model 2		Model 3		Model 4		Model 5
	Coef.	S.E.	Coef.	S.E.	Coef.	Coef.	Coef.	S.E.	Coef.	S.E.
Nativity status
Families of CBM	Ref.		Ref.		Ref.		Ref.		Ref.
Native families	0.19***	(0.03)	0.08**	(0.03)	0.19***	(0.03)	0.09**	(0.03)	0.10**	(0.03)
First-gen. immigrants	−0.06	(0.05)	0.02	(0.05)	−0.06	(0.05)	0.01	(0.05)	0.00	(0.05)
Second-gen. immigrants	−0.12*	(0.05)	−0.08	(0.05)	−0.13**	(0.05)	−0.09	(0.05)	−0.09	(0.05)
Return migrants	0.30**	(0.09)	0.04	(0.09)	0.24**	(0.09)	0.03	(0.09)	0.06	(0.09)
Family Resources			0.01***	(0.00)			0.01***	(0.00)	0.01***	(0.00)
Parental Education Expectations
Tertiary (=1)					0.25***	(0.05)	0.14**	(0.05)	0.16**	(0.05)
Control variables
Grade
Grade 10 or above (=1)	−0.11**	(0.03)	−0.13***	(0.03)	−0.13***	(0.03)	−0.14***	(0.03)	−0.14***	(0.03)
Gender
Male (=1)	0.02	(0.03)	0.00	(0.03)	0.01	(0.03)	0.00	(0.03)	−0.01	(0.03)
Family structure
Two-parent families	Ref.		Ref.		Ref.		Ref.		Ref.
Single-mother families	−0.17**	(0.05)	−0.14**	(0.05)	−0.17**	(0.05)	−0.15**	(0.05)	−0.16***	(0.05)
Single-father families	0.03	(0.13)	0.06	(0.13)	0.06	(0.13)	0.08	(0.13)	0.03	(0.13)
Others	−0.39**	(0.12)	−0.34**	(0.11)	−0.38**	(0.12)	−0.34**	(0.11)	−0.32**	(0.12)
Father employed
Yes (=1)	0.17***	(0.05)	0.12*	(0.05)	0.16**	(0.05)	0.12*	(0.05)	0.13**	(0.05)
Mother employed
Yes (=1)	−0.00	(0.03)	−0.02	(0.03)	−0.00	(0.03)	−0.01	(0.03)	−0.04	(0.03)
Living with siblings
Yes (=1)	−0.12***	(0.03)	−0.09**	(0.03)	−0.10**	(0.03)	−0.09**	(0.03)	−0.09**	(0.03)
Living with grandparents
Yes (=1)	0.07	(0.04)	0.04	(0.04)	0.07	(0.04)	0.05	(0.04)	0.05	(0.04)
Constant	−0.02	(0.07)	−0.29***	(0.06)	−0.20**	(0.07)	−0.37***	(0.07)	−0.29***	(0.06)
School fixed-effects	No		No		No		No		Controlled
R²	0.04		0.08		0.05		0.08		0.13

Notes: ***p < .001, **p < .01, *p < .05.

In Model 2, we additionally included family resources in the analysis. Consistent with the resource perspective, family resources were found to be positively associated with parental involvement (B = 0.01, p < .001). After controlling for family resources, the return migrants and second-generation immigrants were no longer statistically different from the cross-border marriages in terms of parental involvement. In other words, the advantage of return migrants and the disadvantage of second-generation immigrants in parental involvement could be explained by the difference in family resources. In Model 2, the native families still had a higher level of parental involvement than the cross-border marriage families (B = 0.08, p < .01), although the difference was now substantially reduced, from 0.19 to 0.08. In other words, family resources explained around 58% of the difference in parental involvement between the cross-border marriage families and native families.

In Model 3, we controlled for the background variables and parental expectations, but not for family resources. Consistent with the parental expectations perspective, parental expectations were found to be positively associated with parental involvement (B = 0.25, p < .001). For the coefficients of nativity status, the pattern was similar to that of Model 1. The return migrants (B = 0.24, p < .01) and native families (B = 0.19, p < .001) had higher levels of parental involvement. The second-generation immigrant families had lower levels of parental involvement than the cross-border marriage families (B = -0.13, p < .01). Again, the cross-border marriage families and first-generation immigrant families were not statistically different in terms of parental involvement. The sizes of the coefficients were also similar to those in Model 1, suggesting that parental expectations do not have a strong suppression effect on the relationship between nativity and parental involvement.

In Model 4, we controlled for both family resources and parental expectations, along with the background variables. Again, family resources (B = 0.01, p < .001) and parental expectations (B = 0.14, p < .01) were significantly and positively associated with parental involvement. The coefficients of Model 4 were similar to those of Model 2. Parental involvement in the cross-border families was not significantly different from that in the first-generation immigrant, second-generation immigrant, and return migrant families. The native families had a higher level of parental involvement than cross-border marriage families (B = 0.09, p < .001). In the final model, we additionally controlled for school fixed effects. The coefficients of family types remain similar as in Model 4, showing that the within-school differences in parental involvement across family types are similar.

In summary, the results of the regression models show evidence of family resources as a reinforcing mechanism between nativity and parental involvement. While parental expectations are related to parental involvement, parental expectations do not suppress the disadvantage of immigrants in parental involvement. Besides, the R-squared values for the models are also low (from 0.04 to 0.13), suggesting that there is a high proportion of unexplained variance in parental involvement. Although cross-border marriage families are significantly more resource-rich than first-generation immigrant families, the difference in parental involvement between cross-border marriage families and first-generation immigration families is not significant in all models. This result may suggest other hidden mechanisms at work in suppressing the differences between the two groups.

In addition to the analysis presented in Table 4, we also performed additional analyses to examine the relationships between the independent variables and the different indicators of parental involvement separately. The patterns are very similar between the independent variables and different indicators of parental involvement. Due to space constraints, we do not show the results here.

Discussion

To summarize our findings, cross-border marriage families are significantly different from and sit between native families and immigrant families in terms of family resources, parental expectations, and parental involvement. Cross-border marriage families are more resource-rich than first- and second-generation immigrant families, but have fewer resources than native families. This is consistent with earlier studies’ finding that couples involved in cross-border marriages in Hong Kong are more likely to be socioeconomically disadvantaged (Choi & Cheung, 2017; Lin & Ma, 2008; Pong et al., 2014). It also expands on the literature showing that cross-border marriage families are less disadvantaged as immigrant families in terms of resources. Our findings confirm that immigrants have higher parental expectations for their children’s education when family resources are controlled for. Our study also adds to the literature that families of cross-border marriages have higher parental expectations than native families, although they are less aspirational than first- and second-generation immigrants. In terms of parental involvement, cross-border marriage families are significantly different from native and immigrant families.

Besides, our findings suggest that family resources and parental expectations are both positively associated with parental involvement. However, we only find evidence for the role of family resources in explaining the relationship between nativity and parental involvement. Cross-border marriage families have a lower level of parental involvement than native families but a higher level than immigrant families, mainly because of the relative positions of each group in terms of family resources. However, even after controlling for family resources, we find a higher level of parental involvement in native families, suggesting some other hidden barriers to parental involvement for cross-border marriage families. The difference in parental expectations between different family types is statistically significant but not substantial. Most parents in Hong Kong, regardless of their nativity status, expect tertiary education for their children. Hence, even though parental expectations is a positive factor in parental involvement, the variation in the level of parental expectations is not strong enough as a suppressing mechanism to compensate for the disadvantages of cross-border marriage families and immigrant families in parental involvement. As the reinforcing mechanism of family resources is stronger than the suppressing mechanism of parental expectations in explaining parental involvement, we see a higher level of parental involvement among native families although they have lower expectations than cross-border marriage and immigrant families.

Our study contributes to the literature on migration and parental involvement in two meaningful ways. First, the findings echo Ramakrishnan’s argument (2004) that students with one foreign-born parent deserve special attention. Grouping cross-border marriage families either with “native” or “immigrant” families, as done by many past studies (e.g., Hastedt, 2016; Pong, 2009; Schwartz & Stiefel, 2006), is problematic, at least in the Hong Kong context. Our results indicate that cross-border marriage families from native families and immigrant families are indeed different in terms of the level of family resources, parental expectation and parental involvement. Since there are significant differences among these groups, grouping cross-border marriage families with either native families or immigrant families would underestimate the native-immigrant gap in parental involvement and other related outcomes. Hence, the lack of the native-immigrant gap in past studies, which used simple categorization of native and immigrant families, might be a result of underestimation of the gap. Furthermore, the proportion of cross-border marriage is increasing sharply over the years. The problem of underestimating the native-immigrant gap would become even more severe if the proportion of cross-border marriage families increases further. This problem is not exclusive to Hong Kong but also applies in other societies with a high proportion of cross-border marriage families. Future research on the immigrant population should not oversimplify the student population by ignoring the significance of cross-border marriage families. International comparative studies (e.g., Hastedt, 2016; Schwartz & Stiefel, 2006) comparing native-immigrant gaps should also pay special attention to the migration context for countries with various sources of migrants because the relative positions of cross-border marriage families and immigrant families vary in different migration contexts.

Second, our findings indicate that family resources play a greater role than parental expectations in the differences in parental involvement between cross-border marriage, native, and immigrant families. This suggests that the disadvantages of cross-border marriage and immigrant families in parental involvement largely stem from a lack of family resources. Parents from cross-border marriage and immigrant families are more motivated to get involved in their children’s education but face barriers to such involvement. Providing resources for parents from lower socioeconomic backgrounds may be effective in helping those disadvantaged groups, including cross-border and immigrant families. Public policy should aim to provide help at the family level on top of school-level support. Policy-makers and practitioners should pay attention to the lack of knowledge and resources as barriers to the involvement of parents of migrant families. Designated programs could be designed to help those less resource-rich parents to get involved in their children’s education.

We recognize some limitations of our study. First, we compared the parental academic involvement of native and immigrant families without examining the effect of the number of years the students and their parents have lived in Hong Kong. Past studies have shown that age at arrival and the number of years residing in Hong Kong are both related to social mobility (Zhang & Wu, 2011; Zhang & Ye, 2018). Further studies could explore the importance of the assimilation process to parental academic involvement in immigrant and cross-border families.

Second, we used one composite index to measure family resources in this study because the indicators of family resources are highly correlated. Yet, it would be interesting to look at how different aspects of family resources are related to parental involvement. We conducted additional analysis to examine the roles of parents’ schooling, cultural possession at home, and family wealth on parental involvement. We found that the roles of parents’ schooling and cultural possession at home are significantly related to parental involvement, while family wealth is not. Further studies could explore the detailed relationships between more aspects of family resources and parental involvement.

Third, there could be other mechanisms explaining the relationship between cross-border marriage and parental involvement. There are many distinctive features of cross-border marriage. For example, past studies found that the age gap between husbands and wives from cross-border marriage tend to be larger (Choi & Cheung, 2017). These features may also be related to parental involvement, but are not examined in this study. Further investigations are needed to examine other possible mechanisms explaining the relationship between cross-border marriage and parental involvement.

Lastly, while Hong Kong provides an interesting context for sociologists to explore the increasing complexity of the student body in East Asia, its migration context is different from that of many other places. It is important to examine the issues raised in this study in other contexts.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Adam Cheung acknowledges support from the Early Career Scheme Grant funded by the Research Grants Council of Hong Kong (ECS: HKBU/28603817) and Start-Up Grant funded by Hong Kong Baptist University. Hyunjoon Park acknowledges support from the National Research Foundation of Korea Grant (NRF-2016S1A3A2924944).

ORCID iD

Adam Ka-Lok Cheung

References

Akhtar

Malik

J. A.

Begeer

(2017). The grandparents’ influence: Parenting styles and social competence among children of joint families. Journal of Child and Family studies, 26(2), 603–611.

Astone

N. M.

McLanahan

(1991). Family structure, parental practices and high school completion. American Sociological Review, 56(3), 309–320.

Borjas

G. J.

(1987). Self-selection and the earnings of immigrants. American Economic Review, 77(4), 531–553.

Bruce-Eddings

2020. Dealing with interfering grandparents: How to navigate a challenging relationship. New York Times, April 17. https://www.nytimes.com/2020/04/17/parenting/interfering-grandparents.html

Cattaneo

M. A.

Wolter

S. C.

(2015). Better migrants, better PISA results: Findings from a natural experiment. IZA Journal of Migration, 4(1), 1–19.

Cheung

A. K. L.

Park

(2016). Single parenthood, parental involvement and students’ educational outcomes in hong kong. Marriage and Family Review, 52(1–2), 15–10.

Cheung

A. K. L.

Chiu

J. T. Y.

(2020). Husband-to-wife sexual coercion in cross-border marriage: A relationship power perspective. Current Sociology, 69(1), 77–98.

Choi

S. Y. P.

Cheung

A. K. L.

(2017). Dissimilar and disadvantaged age discrepancy, financial stress, and marital conflict in cross-border marriages. Journal of Family Issues, 38(18), 2521–2544.

Constable

(2010). Cross-border marriages: Gender and mobility in transnational Asia. University of Pennsylvania Press.

10.

Downey

D. B.

(1995). Family size, parental resources, and children’s educational performance. American Sociological Review 60(5), 746–761.

11.

Downey

D. B.

(2002). Parental and Family Involvement in Education. In Molnar

(Ed.), School reform proposals: The research evidence (pp. 113–134). Information Age Pub.

12.

Duran

B. J.

Weffer

R. E.

(1992). Immigrants’ aspirations, high school process, and academic outcomes. American Educational Research Journal, 29(1), 163–181.

13.

Englund

M. M.

Luckner

A. E.

Whaley

G. J.

Egeland

(2004). Children’s achievement in early elementary school: Longitudinal effects of parental involvement, expectations, and quality of assistance. Journal of Educational Psychology 96(4), 723–730

14.

Fong

(2011). Return migration in hong kong: Earnings of recent return migrants. Social Transformations in Chinese Societies, 7(1), 93–116.

15.

Garcia Coll

Akiba

Palacios

Bailey

Silver

DiMartino

Chin

. (2002). Parental involvement in children’s education: Lessons from three immigrant groups. Parenting: Science and Practice, 2(3), 303–324.

16.

Hamilton

Roksa

Nielsen

(2018). Providing a ‘‘Leg up’’: Parental involvement and opportunity hoarding in college. Sociology of Education, 91(2), 111–131.

17.

Hastedt

(2016). Immigrant students in TIMSS. In Hastedt

(Ed.), Mathematics achievement of immigrant students (pp. 47–73). Springer International Publishing.

18.

HKPISA Centre. (2013). HKPISA 2012 Preliminary Report. HKPISA Centre, The Chinese University of Hong Kong.

19.

E. S. C.

(2006). Social disparity of family involvement in hong kong: Effect of family resources and family network. School Community Journal, 16, 7–26.

20.

E. S. C.

Willms

J. D.

(1996). Effects of parental involvement on eighth-grade achievement. Sociology of Education, 69(2), 126-141.

21.

Hong Kong Census and Statistics Department. (1999). Hong Kong residents with spouses/children in the mainland of China. Special Topics Report No. 22. Hong Kong SAR Government.

22.

Hong Kong Census and Statistics Department. (2015). Hong Kong monthly digest of statistics: Marriage and divorce trends in Hong Kong, 1991 to 2013. Hong Kong: Hong Kong SAR Government.

23.

Jones

Shen

(2008). International marriage in East and Southeast Asia: trends and research emphases. Citizenship Studies, 12(1), 9–25.

24.

Kao

Tienda

(1998). Educational aspirations of minority youth. American Journal of Education, 106(3), 349–384.

25.

Lareau

(1989). Home advantage: Social class and parental involvement in elementary education. Falmer.

26.

Lareau

(2002). Invisible inequality: Social class and childrearing in black families and white families. American Sociological Review, 67(5), 747–776.

27.

Lin

(2008). Examining cross-border marriages in Hong Kong since its return to China in 1997. Population, Space and Place, 14, 407–18.

28.

Mantovani

Gasperoni

(2018). Native and immigrant parents’ involvement in school-related activities in France and Italy. Italian Journal of Sociology of Education, 10(3), 110–139.

29.

McNeal

R. B.

(1999). Parental involvement as social capital: Differential effectiveness on science achievement, truancy, and dropping out. Social Forces, 78(1), 117–44.

30.

Mood

Carina

. 2010. Logistic regression: Why we cannot do what we think we can do, and what we can do about it. European Sociological Review 26(1), 67–82.

31.

Morgan

C. V.

Hoffman

J. P.

(2007). International marriages in Japan: A prefecture-level Analysis. International Journal of Sociology of the Family, 33, 157–182.

32.

OECD. (2014). PISA 2012 Technical Report. OECD.

33.

Park

(2006). Parental involvement and educational outcomes: The significance of institutional arrangements of educational system. International Perspectives on Education and Society, 7, 187–208.

34.

Park

Byun

Kim

(2011). Parental involvement and students’ cognitive outcomes in Korea focusing on private tutoring. Sociology of Education, 84(1), 3–22.

35.

Pong

S. L.

(2009). Grade level and achievement of immigrants’ children: Academic redshirting in Hong Kong. Educational Research and Evaluation, 15(4), 405–25.

36.

Pong

S.L.

Post

Fok

M.S.Y.

(2014). Blurring boundaries? Immigration and exogamous marriages in Hong Kong. Population and Development Review, 40, 629–652.

37.

Portes

Rumbaut

R. G

. (2001). Legacies: The story of the immigrant second generation. University of California Press.

38.

Rangvid

B. S.

(2007). Sources of immigrants’ underachievement: Results from PISA—Copenhagen. Education Economics, 15(3), 293–326.

39.

Ramakrishnan

K. S.

(2004). Second-Generation Immigrants? The “2.5 Generation” in the United States. Social Science Quarterly, 85, 380–399.

40.

Raymo

J. M.

Park

(2020). Marriage decline in Korea: Changing composition of the domestic marriage market and growth in international marriage. Demography, 57, 171–194.

41.

Schwartz

A. E.

Stiefel

(2006). Is there a nativity gap? New evidence on the academic performance of immigrant students. Education, Finance and Policy, 1(1), 17–49.

42.

S. R.

Schulenberg

J. E.

(2005). Parent beliefs and children’s achievement trajectories during the transition to school in Asian American and European American families. International Journal of Behavioral Development, 29(6), 505–515.

43.

S. R.

Rowley

S. J.

Schulenberg

J. E.

(2007). Predictors of parent involvement across contexts in Asian American and European American families. Journal of Comparative Family Studies, 38(1), 1–28.

44.

Wei

Zhou

Han

Liu

Luo

(2016). The relationship between parental involvement and elementary students’ academic achievement in China.Journal of Comparative Family Studies, 47(4), 483–500.

45.

(2017). The rise of the second generation: aspirations, motivations and academic success of chinese immigrants’ children in Hong Kong. Journal of Ethnic and Migration Studies, 43(7), 1164–1189.

46.

Yamamoto

Holloway

S. D.

(2010). Parental expectations and children’s academic performance in sociocultural context. Educational Psychology Review, 22(3), 189–214.

47.

Yeung

W. J. J.

(2019). Migration and marriage in Asian contexts. Journal of Ethnic and Migration Studies, 46(14),2863–2879.

48.

Zhang

(2011). Social changes, cohort quality, and economic adaptation of Chinese immigrants in Hong Kong, 1991-2006. Asian and Pacific Migration Journal, 20,1–29.

49.

Zhang

(2018). Mode of migration, age at arrival, and occupational attainment of immigrants from mainland China to Hong Kong. Chinese Sociological Review, 50(1), 83–112.

50.

Zhang

Nie

(2019). Family background, parenting practices, and child outcomes: Chinese migrants’ offspring in Hong Kong. Chinese Journal of Sociology, 5(3), 263–282.