Development of urban adaptation and social identity of migrant children in China: A longitudinal study

International Journal of Intercultural Relations 37 (2013) 354–365

Contents lists available at SciVerse ScienceDirect

International Journal of Intercultural Relations journal homepage: www.elsevier.com/locate/ijintrel

Development of urban adaptation and social identity of migrant children in China: A longitudinal study Xiaojiao Yuan a , Xiaoyi Fang a,∗ , Yang Liu b , Shumeng Hou a , Xiuyun Lin a a b

Institute of Developmental Psychology, Beijing Normal University, Beijing 100875, China Institute of Psychological and Behavior Research, Beihang University, Beijing 100191, China

a r t i c l e

i n f o

Article history: Received 6 January 2012 Received in revised form 7 August 2012 Accepted 28 October 2012 Keywords: Migrant children Urban adaptation Social identity Longitudinal study

a b s t r a c t This study was a longitudinal investigation of the development pattern of urban adaptation, social identity and their dynamic relationship among migrant children within two different types of school settings in China. Six hundred eighty migrant children participated in two assessments over a period of one year. Migrant children in both types of school settings showed better psychological adaptation in the follow-up, but only migrant children in public school improved in sociocultural adaptation. Identification with the culture of origin declined among migrant children in public school and increased among children in migrant children school, while identification with the host culture was inversely affected. The strength of association between urban adaptation and social identity did not differ between the two types of school settings. Identification with the host culture positively predicted subsequent sociocultural and psychological adaptation; however, only sociocultural adaptation positively predicted subsequent identification with the host culture. © 2012 Elsevier Ltd. All rights reserved.

1. Introduction With the expansion of economic reform policy and the acceleration of urbanization in China, an increasing number of rural laborers are migrating into cities, resulting in a rapid increase in population mobility and family migration. Accordingly, the phenomenon of migrant children has emerged in China. According to the Interim Measure of School Education for Temporary Migrant Children (1998) issued by the State Education Committee, “migrant children” refers to children aged 6–14 years (or 7–15) who have temporarily lived as migrants for more than half a year with their parents or guardians. Additionally, the 5th Chinese National Census defined “migrant children” as youth under 18 years of age who have lived in a town/sub-district for more than half a year and whose household register (hukou) is in another town/sub-district (Zhang & Zhao, 2003). Based on the above definitions, this study defined migrant children as youth aged 6–18 who have lived as migrants for more than half a year with their parents or guardians. The number of migrant children in China has been rising sharply over the past twenty years. In Beijing, for example, there were 66,392 migrant children of school age in 1997, and that number increased to 240,000 by the end of 2003 (Xinhua News Agency, 2003) and to 418,000 by 2009 (Beijing News, 2010). In 2010, migrant children accounted for more than 43% of all students enrolled in elementary schools of Beijing (Beijing Medium and Long-term National Educational Reform and Development Plan (2010–2020), 2011). After migrating to the city, there are two different types of school settings for migrant children: public schools and migrant children schools. Public schools are established by the government for the majority of local urban children and have qualified teachers, good environmental facilities and adequate funds (Li et al., 2010). In public schools, migrant children study and live

∗ Corresponding author. Tel.: +86 010 58808232; fax: +86 010 58808230. E-mail address: [email protected] (X. Fang). 0147-1767/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijintrel.2012.10.002

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

355

together with local urban children and experience more contact with the host society (Yuan, Fang, Liu, & Li, 2009). However, without a local urban hukou, migrant children can only temporarily enroll in public schools as transient students. Migrant children schools, on the contrary, are established specifically for migrant children by migrants themselves in response to the increasing demand for schooling among growing numbers of migrant children in the city. These schools are located in rural migrant communities, and the students are typically from families that have moved from rural areas in the various provinces of China (Chen, Wang, & Wang, 2007). However, these migrant children schools are usually under-funded and inadequately staffed compared to public schools (Li et al., 2010). According to statistics provided by the Beijing municipal education commission in 2006, 61% of migrant children in Beijing enrolled in public schools (Xie, Zou, & Li, 2007). It is important to distinguish the two types of school settings that migrant children attend in this study. According to the “interactional perspective,” the particular circumstances of the immigrant group within the new society play a moderating role on immigrant identity and adaptation (Phinney, Horenczyk, Liebkind, & Vedder, 2001). However, few empirical studies have identified distinct circumstances of the immigrant groups within the host culture, and even fewer studies have explored the different development patterns that immigrants in different circumstances may follow. The purpose of this study, therefore, was to use a longitudinal paradigm to explore the adaptation process of migrant children from the perspective of different school settings, which is one of the most salient circumstances facing migrant children in China. Our research is guided by three core issues in the field of immigrant studies: immigrant adaptation outcomes over time, development of social identify during the adaptation process, and the relationship between social identity and adaptation outcomes. As such, this study will explore the development of urban adaptation, social identity, and their dynamic relationship among migrant children within different types of school settings. 1.1. Cultural context of migrant children in China As in most developing countries, large sociocultural differences exist between urban and rural areas in China due to the urban–rural dualistic structure. In the economic domain, the urban economy predominantly comprises industrial manufacturing. Infrastructure such as public roads, communication systems, sanitation and education are significantly developed; in contrast, the rural economy is typically a small peasant economy, and the infrastructure there develops more slowly. There is also a major gap in income and consumption between urban and rural people (Liu, 2008). As for the societies themselves, social relations in rural areas are based on individual interpersonal bonds such as friendship and kinship. Individuals within a rural community know each other well and interact frequently and intensely, and they share common values and traditional customs that maintain their relationships. However, in urban areas, social relations are more official, contractual, impersonalized and professional. Additionally, social structures in rural areas consist mainly of kin clans and neighborhoods, while social structures are far more complex in urban locales and take the form of community, political and economic organizations, associations, and other intermediary or unofficial organizations (Jiang, 2004). In the realm of cultural values, rural areas accumulate and retain traditional Chinese culture, and traditional customs and lifestyles are better protected. On the contrary, urban areas mostly represent the characteristics of modern industrial manufacturing (Zeng & Zhu, 2002). Moving from rural to urban areas, Chinese migrant children share experiences and challenges that are similar to those of international immigrants. They experience various acculturation processes, including conforming to requirements of the new urban environment, integrating diverse values between urban and rural areas, and encountering prejudice and discrimination from local people (Chen et al., 2007). For instance, migrant children are required to study in mandarin rather than their regional dialect; familiarize themselves with urban organizations, facilities and transportation they never encountered in rural areas; and change their kinship and friendship-based mode of life to a more official and impersonal one. As for the development of social values, migrant children are primarily raised and socialized within rural areas where traditional beliefs and practices, such as encouraging self sacrifice for family and filial piety, are emphasized (Chen et al., 2007). Once they move into the city, migrant children need to adjust to the modern value system, which places more emphasis on individual development. However, migrant children in China also differ from international immigrants in some ways. Contrary to the multiculturalism of Western societies (Berry, 1984), Chinese society is characterized by the opposing urban–rural dualistic structure. In the hukou system, residents are classified as having either agricultural status (almost all rural residents) or non-agricultural status (urban residents) (Li et al., 2010). Thus, it is possible that migrant children in China migrate in the context of more conflicting identity statuses compared to international immigrants. As for the institutional aspect, the hukou system, like a fort, divides urban and rural residents into two distinct social spheres. Rural people are not entitled to the same status and rights as their urban counterparts (Zhang, 2011). In the field of education, for example, migrant children are only allowed to finish elementary and junior middle school in the city and must to return to their hometowns to enroll in high school. They are also barred from taking college entrance examinations in the city to which they migrated (Chen et al., 2007). Culturally, rural people are often considered unsophisticated and are viewed as irrational and traditional, whereas urban citizens tend to be viewed as a source of culture and as proponents of social progress (Li, 2006). Thus, some migrant children in China might suffer more from institutional and cultural discrimination than international immigrants due to the unbalanced urban–rural dualistic structure. In addition, Chinese migrant children’s minority status is not as visible as that of international immigrants because of migrating within the same country. It is difficult to identify an individual as having migrant status if they are well assimilated into urban society, especially among migrant children within public schools, where migrant children study and live together with local urban children.

356

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

1.2. Urban adaptation of migrant children in China Guided by studies on immigrant acculturation (e.g., Searle & Ward, 1990), we define the urban adaptation of migrant children as the process during which migrant children, based on frequent interactions with urban society, adjust their behaviors and values to achieve a harmonious integration with urban culture. Accordingly, there are two distinct adaptation outcomes: sociocultural adaptation and psychological adaptation (Ward & Rana-Deuba, 1999). Sociocultural adaptation refers to an individual’s social competence in managing daily life in an intercultural setting, while psychological adaptation refers to maintaining good mental health during this process (Berry, Phinney, Sam, & Vedder, 2006; Ward & Rana-Deuba, 1999). Many researchers have explored the status of migrant children’s urban adaptation, and most have become pessimistic about the children’s situation. Migrant children’s social competence in managing their daily lives in an intercultural setting was significantly lower than that of local urban children (Yuan et al., 2009). Moreover, the frequency of problem behaviors among migrant children was much higher than that among urban children (Wang, Wu, & Shen, 2005). Migrant children also suffered from mental stressors such as social anxiety, loneliness, depression (Lin, Fang, Liu, & Lan, 2009) and perceptions of discrimination (Li, Zou, Zhang, & Yang, 2008). In addition, studies have shown that migrant children within public schools tend to adapt more effectively than their counterparts within migrant children schools. For instance, despite the assumption that migrant children might feel more discrimination in public schools, children within migrant children schools were found to be more often worried about discrimination (Zou, Qu, & Zhang, 2004), reported lower school satisfaction (Qu, Zou, & Li, 2008), and achieved lower levels of sociocultural and psychological adaptation, even when controlling for family socioeconomic status (Yuan et al., 2009). Unfortunately, relatively little has been learned about the development of migrant children’s adaptation. Whereas a lot of longitudinal studies have focused on international immigrants’ adaptation process, only two studies have longitudinally explored the development of psychological adaptation (loneliness and depression) among migrant children in China. Findings from these studies indicated that although loneliness and depression among migrant children are higher than among local children, migrant children’s loneliness and depression tends to ameliorate over time (Hou, Yuan, Liu, Fang, & Lin, 2011; Zhou, 2010). However, these two studies only focused on psychological adaptation, without also exploring migrant children’s sociocultural adaptation. As noted by international immigrant studies, sociocultural and psychological adaptation may follow different patterns of fluctuation over time (Jasinskaja-Lahti, 2008; Ward, Okura, Kennedy, & Kojima, 1998). Moreover, studies have yet to uncover the different development patterns for migrant children within different types of school settings. 1.3. Social identity of migrant children in China Social identity refers to the portion of an individual’s self-concept that is derived from perceived membership in social groups and the emotional significance and value attached to this membership (Tajfel, 1972). Moving from the countryside to the city, migrant children may identify with both their culture of origin and the new host culture. Zhou and Zhang (2003) explored migrant children’s identification with their hometown and then Beijing via a single-item questionnaire. The result showed that most migrant children still maintained a strong social identity tied to their culture of origin, and few had yet to strongly identify with the host culture. However, these findings contrast with those from another survey that indicated that only 18.9% of migrant children identified themselves as rural people while 69.9% failed to clearly show a social identity (Liu & Fang, 2011). This discrepancy may derive in part from the inclusion of migrant children within different types of school settings. Bosma and Kunnen (2001) suggested that an individual’s social identity is likely to evolve as he or she grows to recognize that the surrounding society’s beliefs, values, and norms are discordant with their own. Thus, we hypothesized that migrant children within different types of school settings may vary in the extent to which they identify with their culture of origin and host culture due to their differing surrounding circumstances. In addition, the discrepancy regarding social identity may also derive from the length of migrant children’s residence in the city because immigrant identities develop and change over time (Kim, 2006). Thus, more attention needs to be paid to the dynamic developmental process. However, previous studies on the change in immigrant identities over time suggest varying developmental trajectories. For example, a longitudinal study on Mexican-heritage youth in the United States showed that their ethnic identity maintained or strengthened over time (Matsunaga, Hecht, Elek, & Ndiaye, 2010), while Kim (2006) suggested a unidirectional move toward assimilation, meaning an overall increase in identification with the host culture. Distinguishing the particular circumstances of immigrants in the host culture may help explain the discrepancy. In terms of migrant children in China, no longitudinal study has examined identity development. Additionally, due to the unbalanced dualistic culture in the Chinese context and the different circumstances of migrant children within two types of school settings have been exposed to, we cannot directly apply results of international immigrants in other contexts to migrant children in China. 1.4. The relationship between urban adaptation and social identity Social identity is closely related to immigrant adaptation outcomes. Some researchers have indicated that social identity plays a significant role in immigrants’ adaptation to the new culture. Those who more strongly identified with the culture of origin suffered fewer psychological problems, while those who more strongly identified with the host culture had fewer sociocultural adaptation problems (Ward & Kennedy, 1994; Ward & Rana-Deuba, 1999). Some adaptation outcomes, such

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

357

Time 2

Time 1 Sociocultural adaptaon

Sociocultural adaptaon

Psychological adaptaon

Psychological adaptaon

Idenficaon with culture of origin

the

Idenficaon with culture of origin

the

Idenficaon host culture

the

Idenficaon host culture

the

with

with

Fig. 1. Hypothesized model of the relationship between urban adaptation and social identity.

as language use, may also affect one’s perception of identity (Matsunaga et al., 2010). Chan (2001) proposed a reciprocal relationship between adaptation outcomes and immigrant identities. On the one hand, many factors might lead to different social identities among immigrants, which may affect urban adaptation. On the other hand, different levels of acculturation could in turn affect social identities. However, cross-sectional studies do not allow us to determine the exact relationship between urban adaptation and social identity because they only reveal the correlation between the two variables during a certain period of time. Longitudinal research is needed to fully investigate the dynamic relationship between the two. Moreover, from an interactional perspective, the relationship between adaptation outcomes and immigrant social identity is also affected by the receiving society and the particular circumstances of the immigrants (Phinney et al., 2001). Under the unbalanced dualistic cultural context in China and the differing circumstances of the two types of school settings, we predict that the relationship between adaptation outcomes and identity among migrant children in China may present different patterns compared to previous international immigrant studies. 1.5. The present study Based on the above discussion, this study uses a longitudinal paradigm to explore the adaptation process of migrant children within two different types of school settings in China, including the development of their urban adaptation and social identity and the dynamic relationship between them. Accordingly, we have made the following hypotheses: (1) based on previous longitudinal studies on international immigrants (Jasinskaja-Lahti, 2008) and migrant children in China (Zhou, 2010), we hypothesized that migrant children in general will display enhanced sociocultural and psychological adaptation over a one-year period, but migrant children in public school will make better progress than their counterparts in migrant children school; (2) based on the different views of identity development in immigrants proposed by Matsunaga et al. (2010) and Kim (2006), we hypothesized that for migrant children in public school, identification with the culture of origin will decline and identification with the host culture will increase, while children within migrant children school will develop in the opposite way; (3) according to Chan’s (2001) theoretical model and the cross-lagged SEM approach, we hypothesized a longitudinal reciprocal relationship between urban adaptation and social identity, as shown in Fig. 1. Specifically, the level of urban adaptation and social identity at Time 1 will each affect their own subsequent levels and the subsequent levels of the other at Time 2; moreover, based on the interactive perspective (Phinney et al., 2001), we also hypothesized that for migrant children in public school, identification with the host culture will play a more important role in the adaptation process, while for children in migrant children school, identification with culture of origin will be more important. 2. Method 2.1. Participants At Time 1 (September 2007), 1164 migrant children were selected via cluster sampling from grade four of elementary school to grade three of junior middle school from six schools in Beijing (five public schools and one migrant children school). The sample comprised 842 (72.3%) migrant children from public schools and 322 (27.7%) from a migrant children school (including both elementary and junior middle schools). Of these, 633 (54.4%) participants were boys, 499 (42.9%) girls, and 32 (2.7%) participants left the gender item blank. Half of the participants (582) attended middle school, and the other half attended elementary school. There was no significant difference in gender and educational phase (elementary school vs. junior middle school) between the migrant children in the public school and those in the migrant children school. The average age of the participants was 12 years (SD = 2), and their average length of residency in Beijing was 6 years (SD = 4).

358

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

The follow-up survey (Time 2, September 2008) included data obtained from 680 participants, 568 (83.5%) of whom were from public schools and 112 (16.5%) were from the migrant children school. Of these, 364 (53.5%) were boys, 296 (43.5%) girls, and 20 (2.9%) left the gender item blank; 368 (54.1%) of the participants attended middle school, and 312 (45.9%) attended elementary school, and they had an average age of 13 years (SD = 2). The average length of residency in Beijing was 7 years (SD = 5). Because of the high mobility of migrants, 484 of the original participants were missed in the follow-up assessment. Of those, 253 were missing because of graduation and 231 had dropped out or returned to their native hometowns. To examine possible selection bias due to sample attrition, a 2 -test and MANOVAs on relevant demographic factors (gender, length of residence in Beijing, school setting) and original level of adaptation and identity measures were performed. The results revealed no differences in terms of gender or length of residence between the migrant children who were missing and those who were retained in the Time 2 assessment. The proportion of migrant children within public schools participating in both assessments was larger (67.5%) than that of children within the migrant children school (34.8%). However, further analysis showed no difference in the original level of urban adaptation (sociocultural and psychological adaptation) and social identity (identification with the culture of origin and the host culture) between the participants who were retained or missing in Time 2 assessment for both types of school settings. 2.2. Measures 2.2.1. Sociocultural adaptation Sociocultural adaptation was assessed using the Sociocultural Adaptation Scale (SCAS) (Ward & Kennedy, 1999). The first version of the SCAS (16 items) was originally developed by Searle and Ward (1990) as an assessment of intercultural competence, with an emphasis on behavioral domains. The most recent version of the SCAS (29 items) includes more cognitive domains (Ward & Kennedy, 1999). The SCAS requires participants to indicate the amount of difficulty experienced in a number of areas along a five-point scale (from “no difficulty” to “extreme difficulty”). It can be easily modified according to the characteristics of the immigrant sample and has been used with various populations (Ward & Kennedy, 1999). In this study, we revised the scale to make it fit the sample of migrant children in China. For example, “understand the differences across cultures” was revised to “understand the differences between Beijing and native hometowns,” and one item concerning religion was deleted to better reflect the Chinese context. The final version of the scale includes 28 items; higher scores indicate better sociocultural adaptation. The SCAS has been widely used in previous studies and has demonstrated internal consistency reliability, with Cronbach’s alpha coefficients ranging from 0.75 to 0.91 (Ward & Kennedy, 1999). In this study, Cronbach’s alpha coefficient was 0.92 at both Time 1 and Time 2. 2.2.2. Psychological adaptation In line with previous research, psychological adaptation was assessed by targeting loneliness and depression in this study (Ward & Rana-Deuba, 1999; Zhou, 2010), with lower levels of loneliness and depression representing better psychological adaptation. Loneliness was measured by the Loneliness Questionnaire (Zou, 2003), which was first revised from Asher, Hyntel, and Renshaw’s loneliness questionnaire (1984). It contains 21 items and adopts a 4-point scale from “1: not at all” to “4: perfectly right.” It measures four aspects of loneliness, including pure feelings of loneliness, the perception of competence in social interactions, the assessment of current peers and the perception of dissatisfaction with important relationships. Higher scores indicate stronger feelings of loneliness. This instrument has been used with Chinese adolescents in the previous study and has demonstrated internal consistency reliability, with a Cronbach’s alpha coefficient of 0.92 (Li & Zou, 2007). In this study, Cronbach’s alpha coefficients of Time1 and Time 2 assessment were 0.91 and 0.92, respectively. Depression was measured by a Chinese version of the Center for Epidemiologic Studies Depression (CES-D) Scale designed by Radloff (1977). It comprises 20 questions and assesses participants along a 4-point scale from “1: never” to “4: always.” Higher scores indicate stronger feeling of depression. The CES-D has been tested in different ethnic contexts and has widely proven reliable and valid, with Cronbach’s alpha coefficients ranging from 0.83 to 0.88 (Husaini, Neff, Harrington, Hughes, & Stone, 1980; Roberts, 1980). In this study, Cronbach’s alpha coefficients of Time1 and Time 2 assessments were 0.91 and 0.88, respectively. 2.2.3. Social identity Social identity was assessed via a revised Chinese version of the Acculturation Index (AI), designed by Ward and Kennedy (1994). The original scale consists of two dimensions, identification with the culture of origin and identification with the host culture, and includes 21 cognitive and behavioral items such as “language” and “in-group and out group perception”. Participants are asked to rate the similarity of their lives with both a representation of the culture of origin and with the host culture on each item using a 7-point scale from “1: the least similar” to “7: extremely similar.” Based on the Chinese context and the characteristics of migrant children, four items were removed from the Chinese version of the scale, including “religious belief,” “cultural activity,” “employment activity” and “political ideology.” Higher scores indicate stronger identification with the culture being considered. This instrument has been used with international immigrants in the previous study and both dimensions have demonstrated internal consistency reliability, with Cronbach’s alpha coefficients ranging from 0.89 to 0.92 (Ward & Rana-Deuba, 1999). In this study, the Cronbach’s alpha coefficients for identification

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

359

with the culture of origin of Time1 and Time 2 assessments were 0.90 and 0.93, respectively. Likewise, the Cronbach’s alpha coefficients for identification with the host culture of Time1 and Time 2 assessments were 0.90 and 0.93, respectively. 2.3. Procedure All of the English questionnaires were first translated into Chinese by the research group, then back-translated by bilingual Chinese natives using standard back-translation techniques, and checked for preservation of meaning and cultural appropriateness by two professors in the department of psychology. Next, a pilot study was administered to 77 selected migrant children to test whether the research instruments were reliable and fully comprehensible by migrant children. The Time 1 survey was conducted in September 2007 among 1164 migrant children. All of the investigators received training in guidance and the content of the questionnaires before implementation of the survey. The survey was conducted by taking a class as a unit and assigning a single investigator to each class. The schoolteachers were asked to leave the classroom during the process. All of the questionnaires were checked and collected on site. The Time 2 survey was conducted in September 2008 using the same research instruments. After the survey, the investigators checked with the teachers in charge of each class to record the missing participants from the Time 1 survey and the reasons they were missing. School leaders and teachers approved all of the assessments in advance. Students who did not want to participate in the survey could choose to study in their seats. Each participant received a pen as a gift after each assessment. 2.4. Data analysis The data analyzed in this article are from the participants who were included in both the Time 1 and Time 2 assessments. Data analysis employed SPSS 19.0 and AMOS 17.0 software to conduct reliability testing, MANOVA, correlation analysis and structural equation modeling. A repeated measures MANOVA was conducted to examine changes in urban adaptation and social identity of migrant children within different types of school settings over one year (to test hypothesis 1 and 2). Structural equation modeling (multiple-group analysis) was conducted to examine the dynamic relationships between urban adaptation and social identity of migrant children within different types of school settings over time (to test hypothesis 3). 3. Results 3.1. Changes in urban adaptation and social identity of migrant children The means and standard deviations for urban adaptation (sociocultural adaptation and psychological adaptation) and social identity (identification with the culture of origin and the host culture) of the migrant children within public schools and the migrant children school in both assessments are shown in Table 1. A 2 (Time) × 2 (School setting) multivariate repeated measures analysis was conducted to examine changes in urban adaptation and social identity of migrant children within different types of school settings over one year. In previous studies (Jasinskaja-Lahti, 2008; Ward et al., 1998), the development of adaptation outcomes followed different time courses; thus, we used the length of residency in Beijing as a covariate to control for its effect. Multivariate tests revealed a significant main effect for time [F(5,619) = 9.73, p < 0.001; 2p = 0.07], school setting [F(5,619) = 22.41, p < 0.001; 2p = 0.15], and a Time × School setting interaction [F(5,619) = 3.40, p < 0.01; 2p = 0.03]. Specifically, the significant interaction effect of time and school setting was found for sociocultural adaptation [F(1,623) = 5.26, p < 0.05; 2p = 0.01], identification with the culture of origin [F(1,623) = 6.02, p < 0.05; 2p = 0.01] and identification with the host culture [F(1,623) = 9.53, p < 0.01; 2p = 0.02]. Further tests indicated that migrant children within public schools enhanced their level of sociocultural Table 1 Urban adaptation and social identity among migrant children at two assessments. Measures

Sociocultural adaptation

Urban adaptation

Psychological adaptation Loneliness Depression Identification with the culture of origin

Social identity

a b

Identification with the host culture

MCS = migrant children school. PS = public school.

School setting

MCS PSb

a

MCS PS MCS PS MCS PS MCS PS

Time 1

Time 2

M

SD

M

SD

3.51 4.03

0.54 0.58

3.59 4.21

0.57 0.52

2.13 1.87 2.12 1.89 3.97 3.75 4.41 5.02

0.54 0.61 0.44 0.54 1.24 1.44 1.22 1.32

2.00 1.71 2.04 1.84 4.16 3.55 4.07 5.11

0.50 0.57 0.47 0.53 1.21 1.48 1.30 1.27

360

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

Fig. 2. Changes in urban adaptation and social identity among migrant children within different school settings.

adaptation [t(567) = −7.99, p < 0.001, d = −0.33], while children within the migrant children school displayed no significant improvement in sociocultural adaptation [t(109) = −1.38, p = 0.169, d = 0.13]. Migrant children within public schools showed a significant decline in identification with the culture of origin [t(538) = 2.79, p < 0.01, d = 0.13] and marginally significantly enhanced identification with the host culture [t(555) = −1.80, p < 0.1, d = −0.08], while children within the migrant children school showed results in the opposite direction, with identification with host culture significantly declining [t(108) = 2.37, p < 0.05, d = 0.26] and identification with culture of origin marginally significantly increasing [t(110) = −1.67, p < 0.1, d = −0.19]. There was no interaction effect on psychological adaptation for either loneliness [F(1,623) = 0.38, p = 0.537] or depression [F(1,623) = 0.12, p = 0.733]. Univariate tests indicated that migrant children in general showed less loneliness [t(667) = 6.22, p < 0.001, d = 0.25] and depression [t(667) = 2.33, p < 0.05, d = 0.10] in the follow-up, as shown in Fig. 2. 3.2. The relationship between urban adaptation and social identity Correlation analyses were conducted to examine the interrelationship among the study variables at the two assessments among migrant children within different types of school settings separately, as shown in Table 2. The proposed model was then tested with a structural equation modeling approach using the Amos 17.0 software package. As both the sociocultural adaptation and social identity scales contain many items, we first used the item-to-construct balance procedure, parceling the items to obtain more than one indicator for a latent variable (Little, Cunningham, Shahar, & Widaman, 2002). We obtained three indicators for sociocultural adaptation, identification with the culture of origin and the host culture. Psychological adaptation is specified by the indicators of loneliness and depression in the model. Multiple-group analysis was used to test whether the relationships between urban adaptation (sociocultural adaptation and psychological adaptation) and social identity (identification with the culture of origin and identification with the host culture) varied among migrant children within public schools and migrant children schools. We compared two models with and without the regression weights equivalently constrained between the two types of school settings. The results showed that the compact model would not worsen the model fit, 2Diff (12) = 7.23, p = 0.842. Pairwise parameter comparison also showed no significant difference in any of the regression weights between urban adaptation and social identity among migrant children within different types of school settings, with Z scores ranging from 0.067 to 1.154. Therefore, Fig. 3 presents the results of an overall model including migrant children within both types of school settings. Although the 2 value reached

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

361

Table 2 Correlations among study variables at two assessments.

1. Sociocultural adaptation T1 2. Loneliness T1 3. Depression T1 4. Identification with the culture of origin T1 5. Identification with the host culture T1 6. Sociocultural adaptation T2 7. Loneliness T2 8. Depression T2 9. Identification with the culture of origin T2 10. Identification with the host culture T2

1

2

3

4

5

6

7

8

9

10

– −0.43** −0.35** −0.28** 0.42** 0.51** −0.27** −0.20** −0.17** 0.31**

−0.32** – 0.64** 0.11** −0.29** −0.35** 0.44** 0.30** 0.10* −0.20**

−0.32** 0.68** – 0.05 −0.23** −0.29** 0.37** 0.38** 0.06 −0.19**

−0.02 0.08 0.06 – −0.41** −0.21** 0.06 0.06 0.38** −0.26**

0.31** −0.07 −0.13 −0.50** – 0.40** −0.23** −0.17** −0.26** 0.47**

0.50** −0.24* −0.28** −0.12 0.25** – −0.39** −0.34** −0.22** 0.45**

−0.32** 0.28** 0.17 0.07 −0.21* −0.37** – 0.68** 0.06 −0.33**

−0.31** 0.27** 0.17 0.06 −0.06 −0.31** 0.59** – 0.07 −0.28**

−0.03 0.13 0.18 0.33** −0.18* −0.15 −0.09 −0.09 – −0.33**

0.32** −0.05 −0.18* −0.16 0.35** 0.38** −0.21** −0.11 −0.38** –

Note. T1 means the assessment at Time 1, T2 means the assessment at Time 2. Lower triangular matrix presents for the sample of migrant children within public schools (N = 568), and higher triangular matrix presents for the sample of migrant children within the migrant children school (N = 112). * p < 0.05. ** p < 0.01.

significance due to the large sample size [2 (185) = 396.35, 2 /df = 2.14, p < 0.001, CFI = 0.981, RMSEA = 0.041], these indices indicated an acceptable model fit. Only statistically significant parameters in the measurement and structural models are shown in the figure to enhance clarity, and all of the reported estimates are standardized. As shown in Fig. 3, sociocultural adaptation, psychological adaptation and identification with both cultures were correlated with each other at Time 1. The better migrant children adapted socioculturally and psychologically, the more strongly they identified with the host culture and the weaker they identified with the culture of origin. All the correlations remained significant at Time 2, except for the relationship between psychological adaptation and identification with the culture of origin. As expected, for all of the variables, the values measured at Time 1 significantly predicted the value at Time 2. Regarding the relationships between urban adaptation and social identity over time, an initially stronger identification with the host culture at Time 1 predicted both better sociocultural adaptation and psychological adaptation at Time 2, while in turn, only

Time 2

Time 1 SA 1 SA 2

0.90 0.89

0.54

Sociocultural

Sociocultural

adaptaon

0.86 -0.33**

-0.27**

Depression 0.75

ICO 3 IHC 1

0.11**

0.88

Idenficaon with the culture of origin

0.87 0.88

IHC 3

0.88 0.86

-0.14

SA 3

**

-0.45**

Psychological maladaptaon

0.19

0.73 Depression 0.35**

0.21**

**

0.89 ICO 1

0.39**

-0.34** -0.14**

Idenficaon with the culture of origin

-0.27**

-0.29**

0.91 ICO 2 0.91 ICO 3 0.91 IHC 1

Idenficaon with

IHC 2

0.49**

maladaptaon

0.47**

0.88

SA 2

0.91 Loneliness Psychological

ICO 2

0.89 0.88

-0.54**

ICO 1

SA 1

adaptaon

SA 3

Loneliness 0.86

0.91 **

the host culture

0.40**

Idenficaon with the host culture

0.90

IHC 2

0.89 IHC3

Fig. 3. Cross-lagged SEM analysis of urban adaptation and social identity. Note: **p < 0.01, *p < 0.05. SA = sociocultural adaptation; ICO = identification with the culture of origin; IHC = identification with the host culture.

362

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

better sociocultural adaptation at Time 1 predicted stronger identification with the host culture. Identification with the culture of origin showed no significant longitudinal relationship with either sociocultural or psychological adaptation. 4. Discussion 4.1. Changes in urban adaptation and social identity of migrant children As assumed in hypothesis 1, over a one-year interval, positive changes were found in the average level of urban adaptation among migrant children in China, which supports Zhou’s (2010) longitudinal study on the change in migrant children’s loneness and depression. This general positive trend is also in accordance with some other international immigrant studies. As noted in a one-year follow-up study by Ward et al. (1998) among Japanese students in New Zealand, adjustment problems were greatest at the entry point and decreased over time. Additionally, in a ten-year longitudinal study, Hou and Beiser (2006) found that language problems and rates of depression declined among Southeast Asian refugees over time. In addition, we found some differences among migrant children within different types of school settings, which partially supports hypothesis 1. Migrant children in public schools showed better progress in sociocultural adaptation, while there was no group difference with regard to improvement in psychological adaptation. The discrepancy can be explained by the distinct framework between sociocultural and psychological adaptation and by the different circumstances that the two groups of migrant children had been exposed to. Sociocultural adaptation refers to an individual’s social competence in managing his or her daily life in the intercultural setting, which is understood within the social learning paradigm, while psychological adaptation refers to retaining good mental health during the process, which is best understood within the stress and coping framework (Berry et al., 2006; Ward et al., 1998). Thus, sociocultural adaptation is much more related to circumstances, such as the amount of contact with host nationals and cultural knowledge, than is psychological adaptation. In the case of migrant children in China, despite their transient student status, living and studying together with local children in public school can provide them with more opportunities to experience the host culture than simply studying with a cohort group in a segregated migrant children school. In terms of social identity development, migrant children within different types of school settings followed opposite development patterns, rather than strengthening ethnic identity (Matsunaga et al., 2010) or making a unidirectional move toward assimilation (Kim, 2006). The polarization of social identities sharpened over the one-year interval: migrant children within public schools, who identified more strongly with the host culture and more weakly with culture of origin than did children within the migrant children school at Time 1, further strengthened their identification with the host culture and identified less with the culture of origin, while their counterparts within the migrant children school strengthened their identification with the culture of origin and declined identification with the host culture, which supported hypothesis 2. The results have provided longitudinal empirical evidence to support Phinney et al. (2001) theoretical perspective that the particular circumstances of the immigrant group within the new society have a significant moderating effect on their identity development. According to Phinney et al. (2001), identification with the culture of origin is likely to be strong where immigrants have a strong desire to retain their identities and pluralism is accepted, while identification with the host culture is likely to be strong where there is pressure toward assimilation. In the case of migrant children in China, particular circumstances differed for migrant children within the two types of school settings. For migrant children within public schools who were living and studying together with local urban children, being exposed to and experiencing the host culture was quite influential toward assimilation. Moreover, due to the unbalanced dualistic cultural structure, rural people suffer a lot of discriminations from local people (Li, 2006). Migrant children within public schools are less likely to want to retain their original identities due to their invisible rural status and interactions with local urban children. On the contrary, children in migrant children schools who primarily only have contact with people from their hometowns and some other rural areas speak and behave in a way consistent with their origins, forming a “rural area in the city,” (Yuan et al., 2009) and leading to enhanced identification with the culture of origin and a weakened identification with the host culture. 4.2. Longitudinal relationships between urban adaptation and social identity According to Chan’s (2001) theoretical model and the cross-lagged SEM approach, we assumed longitudinal reciprocal relations between both identity and adaptation outcomes in the hypothesized model, and we found a reciprocal relation between identification with the host culture and sociocultural adaptation, a cross-lagged relation between identification with the host culture and psychological adaptation, and no longitudinal relation between identification with the culture of origin and adaptation outcomes. In terms of the relationship between identification with the host culture and sociocultural adaptation, a longitudinal reciprocal relation was supported; this relation suggests that stronger identification with one’s host culture predicts better sociocultural adaptation in the future, whereas better sociocultural adaptation could further promote the growth of identification with the host culture. The result has provided longitudinal empirical support to Ward and Kennedy (1999) conclusion that host culture identity has a promotional effect on sociocultural adaptation. This finding can be explained by social identity theory, which holds that when an individual’s identification with a group becomes salient, he or she will behave in ways that magnify the difference with the out-group and minimize differences with in-group members in important ways (Howard, 1998). Thus, when migrant children identify themselves more as urban citizens, they will try harder to

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

363

live and behave in the same way that local urban children behave, which will improve their subsequent level of sociocultural adaptation. Additionally, the result also supported the view that some adaptation outcomes may affect one’s perception of identity (Matsunaga et al., 2010). According to self-categorization theory (Turner, 1985), the process of categorization of the self and others into in-group and out-group defines an individual’s social identity. Migrant children who have better social competence in managing daily life in the city are more likely to categorize themselves into an urban citizen group. As for the cross-lagged relation between identification with the host culture and psychological adaptation, stronger identification with the host culture at Time 1 predicted better psychological adaptation at Time 2, while the initial level of psychological adaptation did not predict later identification with host culture. This result suggests a quasi-causality of identification with the host culture on psychological adaptation. Stronger identification with the host culture could reduce migrant children’s psychological stress because it leads to better social competence in managing daily life in the future. Moreover, due to the unbalanced dualistic structure of urban and rural areas in China, rural people are discriminated against and are considered to be less cultured than urban people (Li, 2006). Therefore, migrant children would perceive less psychological pressure if they identified themselves more as urban citizens. On the contrary, the formation and development of migrant children’s identification with the host culture is a process of categorization, which mainly depends on their perceptions of their own competence in managing life in the city rather than how well they deal with psychological stress. However, in contrast to our hypothesis and previous cross-sectional findings that strong identification with one’s culture of origin predicted enhanced psychological well being (Ward & Kennedy, 1994; Ward & Rana-Deuba, 1999), no longitudinal relationship between original cultural identity and adaptation outcomes (either sociocultural adaptation or psychological adaptation) was found among migrant children in China. This finding might be related to the cultural differences between international immigrants and migrant children in China. For international immigrants, multiculturalism and pluralism are accepted (Berry, 1984). Thus, immigrants could strongly identify with both their culture of origin and the host culture, which played different roles in the adaptation process. However, for migrant children in China, due to the opposing dualistic structure in China, identifying oneself as urban or from a rural population conflict with each other. Additionally, identification with urban culture plays a more important role than identification with the culture of origin due to the unbalanced cultural structure. Although stronger identification with the culture of origin provides a better sense of belonging, it also impairs migrant children’s identification with the host culture, which is important for integrating into urban society. There is also some empirical evidence to support this explanation. We found significant negative correlations between identification with the culture of origin and identification with the host culture among migrant children in both assessments (r ranging from −0.33 to −0.50). However, Ward and Rana-Deuba (1999) reported no significant correlations (r = −0.04) between these two identities in their study of international immigrants. Moreover, in contrast to our interactive hypothesis, the role of the two types of identification in predicting adaptation outcomes was not moderated by the different school settings. Instead, it is stronger identification with the host culture that predicts better adaptation outcomes for migrant children within both types of school settings. As noted above, identification with the culture of origin and the host culture do not have separate effects among migrant children in China. Additionally, due to the unbalanced cultural structure, assimilation usually leads to more chances to integrate into the host society and fewer chances of being excluded and discriminated against. Thus, even for migrant children within migrant children schools, identification with the host culture plays a more important role in their urban adaptation. 4.3. Limitations The current study is limited by sample attrition, limited number of assessment times and not using a multiple-method. Although we conducted some preliminary analyses to limit the possible attrition effect and to ensure that participants who were retained and those who were missing in the follow-up assessment experienced the same level of adaptation at Time 1, it may also be the case that participants who do not develop and adapt well during the process are more likely to go back to their home town and drop out from our study, leaving a sample of migrant children with better adaptation processes. We should also be cautious before drawing conclusions about the developmental trajectory of migrant children in China because we only collected data at two time points. Based on our pretest and posttest data, we could only compare the level of migrant children’s urban adaptation and social identity over the interval of one year, so we cannot conduct a nonlinear analysis such as testing the U-shaped curve hypothesis. Future studies should make more effort to collect data at additional time points to better determine the exact developmental trajectory of migrant children in China. In addition, only self-reports were used in the current study, and that same method variance might explain some of the explained variance. Multiple-method assessment such as parent and teacher evaluations should be used in future studies. 4.4. Implications This study emphasized the importance of the cultural context and particular circumstances in migrant children studies. Due to various cultural differences, series studies in the future should explore the adaptation process of migrant children under the context of the specific dualistic cultural structure in China rather than directly borrowing from international immigrant study results in Western societies. The present study also highlighted the significant role of the particular surrounding circumstances in which migrant children live within the host society. Migrant children living within different circumstances may experience a different adaptation process. Future studies should be more specific and targeted; for example, they should

364

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

explore migrant children’s developmental pattern within different school, family and community circumstances and pay more attention to findings based at the group or individual levels, rather than solely identifying an average pattern among all migrant children. There are also some practice suggestions that can be drawn from this study. As we found, public school is better for the development of urban adaptation in migrant children in China. Therefore, various measures should be taken by agencies and departments to encourage public schools to enroll more migrant children. In addition, due to the antecedent and key role of identification with the host culture in the adaptation process of migrant children in China, increasing migrant children’s identification with the host culture should become the focus of our future studies and intervention practices. 4.5. Conclusions Overall, we conclude that the level of urban adaptation among migrant children in China is improving, especially for those enrolled in public schools. Identification with the culture of origin and the host culture followed different patterns for migrant children within different types of school settings. Identification with the host culture plays an important role in migrant children’s urban adaptation processes. Acknowledgements This research is supported by a key project of the Key Research Base for Humanities and Social Sciences under the Ministry of Education (05JJDXLX16), project of the 11th Five-Year Plan of Municipal Educational Science of Beijing (BAA08004) and project of the 11th Five-Year Plan for Beijing Philosophy and Social Science Development (09BdSH051), China. References Asher, S. R., Hymel, S., & Renshaw, P. D. (1984). Loneliness in children. Child Development, 55, 1456–1464. Beijing Medium and Long-term National Educational Reform and Development Plan. (2010–2020). Beijing Education Committee, issued on March 24, 2011. Beijing News. (2010). Residents in Beijing will reach at 25 million in 2020. [Electronic version]. Retrieved from http://news.sina.com.cn/ c/2010-07-17/025620697107.shtml Berry, J. W. (1984). Multicultural policy in Canada: A social psychological analysis. Canadian Journal of Behavioral Science, 16(4), 353–370. Berry, J. W., Phinney, J. S., Sam, D. L., & Vedder, P. (2006). Immigrant youth: Acculturation, identity, and adaptation. Applied Psychology, 55(3), 303–332. Bosma, H. A., & Kunnen, E. S. (2001). Determinants and mechanisms in ego identity development: A review and synthesis. Developmental Review, 21, 39–66. Chan, R. (2001). Acculturation of young new arrivals from Mainland China to Hong Kong. Doctoral Dissertation of Chinese University of Hong Kong. Chen, X., Wang, L., & Wang, Z. (2007). Shyness-sensitivity and social, school, and psychological adjustment in rural migrant and urban children in China. Child Development, 80(5), 1499–1513. Hou, F., & Beiser, M. (2006). Learning the language of a new country: A ten-year study of English acquisition by South-East Asian refugees in Canada. International Migration, 44, 135–165. Hou, S., Yuan, X., Liu, Y., Lin, X., & Fang, X. (2011). The effect of social support and perceived discrimination on loneliness among migrant children: A longitudinal study. Psychological Development and Education, 27(4), 401–411. Howard, R. (1998). Being Canadian: Citizenship in Canada. Citizenship Studies, 2, 133–152. Husaini, B. A., Neff, J. A., Harrington, J. B., Hughes, M. D., & Stone, R. H. (1980). Depression in rural communities: Validating the CES-D scale. Journal of Community Psychology, 8(1), 20–27. Interim Measure of School Education for Temporary Migrant Children. (1998). SNC No. [1998]2, issued on March 2, 1998. Jasinskaja-Lahti, I. (2008). Long-term immigrant adaptation: Eight-year follow-up study among immigrants from Russia and Estonia living in Finland. International Journal of Psychology, 43(1), 6–18. Jiang, L. (2004). Urbanism and urban adaptation of migrant workers. Social Science Research, 5, 92–96. Kim, Y. Y. (2006). From ethnic to interethnic: The case for identity adaptation and transformation. Journal of Language and Social Psychology, 25, 283–300. Li, C., & Zou, H. (2007). Relationships between perceptual differences of family functioning and adolescent loneliness. Psychological Science, 30(4), 810–813. Li, X. (2006). The institutional analysis of dualistic urban–rural economy structure problem and relevant counter measure. Journal of Public Management, 3, 87–93. Li, X., Zhang, L., Fang, X., Stanton, B., Xiong, Q., Lin, D., & Mathur, A. (2010). Schooling of migrant children in China: Perspectives of school teachers. Vulnerable Children and Youth Studies, 5, 79–87. Li, X., Zou, H., Zhang, J., & Yang, Y. (2008). Qualitative study on the generating mechanism of migrant children’s perception of discrimination: Form a social comparison’s perspective. Psychological Research, 1(2), 66–70. Lin, X., Fang, X., Liu, Y., & Lan, J. (2009). The effect mechanism of stigma perception on mental health among migrant children in Beijing. Journal of Psychology, 41(10), 967–979. Little, T. D., Cunningham, W. A., Shahar, G., & Widaman, K. F. (2002). To parcel or not to parcel: Exploring the question, weighing the merits. Structural Equation Modeling: A Multidisciplinary Journal, 9(2), 151–173. Liu, Y. (2008). Empirical analysis on the effect of income gap between urban and rural areas on resident consumption. Seeker, 1, 64–66. Liu, Y., & Fang, X. (2011). The status of migrant children’s social identity. Journal of China National School of Administration, 3, 61–67. Matsunaga, M., Hecht, M. L., Elek, E., & Ndiaye, K. (2010). A longitudinal analysis of Mexican-Heritage youth in the Southwest United States. Journal of Cross-Cultural Psychology, 41(3), 410–427. Phinney, J. S., Horenczyk, G., Liebkind, K., & Vedder, P. (2001). Ethnic identity, immigration, and well-being: An interactional perspective. Journal of Social Issues, 57(3), 493–510. Qu, K., Zou, H., & Li, X. (2008). Migrant children’s school satisfaction and its relation with teacher–student relationship and learning behavior. Chinese Journal of Special Education, 7, 50–55. Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. Roberts, R. E. (1980). Reliability of the CES-D scale in different ethnic contexts. Psychiatry Research, 2(2), 125–134. Searle, W., & Ward, C. (1990). The prediction of psychological and sociocultural adjustment during cross-cultural transitions. International Journal of Intercultural Relations, 14(4), 449–464. Tajfel, H. (1972). Social categorization. In S. Moscovici (Ed.), Introduction àa psychologie sociale. Paris: Larousse. Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive theory of group behaviour. In E. J. Lawler (Ed.), Advances in group processes: Theory and research. Greenwich, CT: JAI.

X. Yuan et al. / International Journal of Intercultural Relations 37 (2013) 354–365

365

Wang, F., Wu, X., & Shen, Y. (2005). Behavioral disorders and self-concept of children of farmer-workers in primary schools. Chinese Journal of School Doctor, 19, 345–347. Ward, C., & Kennedy, A. (1994). Acculturation strategies, psychological adjustment, and sociocultural competence during cross-cultural transitions. International Journal of Intercultural Relations, 18, 329–343. Ward, C., & Kennedy, A. (1999). The measurement of sociocultural adaptation. International Journal of Intercultural Relations, 23(4), 659–677. Ward, C., Okura, Y., Kennedy, A., & Kojima, T. (1998). The U-curve on trial: A longitudinal study of psychological and sociocultural adjustment during cross-cultural transition. International Journal of Intercultural Relations, 22, 277–291. Ward, C., & Rana-Deuba, A. (1999). Acculturation and adaptation revisited. Journal of Cross-Cultural Psychology, 30(4), 422–442. Xie, Y., Zou, H., & Li, X. (2007). A comparative research on migrant children’s teacher–student relations between public school and migrant school. Journal of the Chinese Society of Education, 6, 9–12. Xinhua News Agency. (2003). The investigation of status of compulsory education of migrant children in Beijing [Electronic version]. Retrieved from http://news.xinhuanet.com/edu/2003-06/24/content 934603.htm Yuan, X., Fang, X., Liu, Y., & Li, Z. (2009). Educational settings and city adaptation of migrant children. Journal of Beijing Normal University (Social Science), 5, 25–32. Zeng, J., & Zhu, Y. (2002). Urban–rural correlated development and cultural integration. Human Geography, 17(4), 7–11. Zhang, H. (2011). Eliminating double dualistic structure: The foundation of urban–rural integration [Electronic version]. Retrieved from http://www.zgxcfx.com/Article Show.asp?ArticleID=32924 Zhang, L., & Zhao, S. (2003). Let us share sunshine: A survey report of migrant children in nine cities. Beijing: China Children Center of Working Committee for Children and Women under the State Council of China. Zhou, H. (2010). The status and development of migrant children’s mental health: Analysis based on “longitudinal investigation of children’s development”. Population Research, 34(2), 66–75. Zhou, H., & Zhang, N. (2003). Social integration of migrant children. Chinese Journal of Population Science, 4, 69–73. Zou, H. (2003). Developing characters, functions and predictors of peer relationship among adolescence. Beijing: Beijing Normal University Press. Zou, H., Qu, Z., & Zhang, Q. (2004). A survey on status of migrant children’s survival and being protective in nine cities. Youth Studies, 1, 1–7.