زمینه مدرسه، روابط دوستی و بهداشت روانی نوجوانان: تجزیه و تحلیل چند سطحی از بررسی پانل جوانان کره ای (KYPS)

Research on the social determinants of health suggests that interpersonal networks play a critical role in facilitating individual mental and physical well-being. Prior studies also indicate that ecological or contextual factors contribute to positive health outcomes. This study extends prior research by examining the factors associated with adolescent health in an Asian context. Based on the multilevel analysis of the Korean Youth Panel Survey (2006 & 2007), a longitudinal project funded by the Korean government, it investigates some of the key variables related to the mental health of Korean students. Much of previous research focuses on the functions of social capital. This study contributes to the social epidemiology literature by investigating the possible downside of network ties. Specifically, it asks whether having delinquent friends is associated with negative mental health experiences. In addition, little research has been conducted concerning the associations between adolescent health outcomes and school characteristics. This study moves in that direction by examining the relationship between mental well-being of students and a variety of school related variables (e.g., subjective attitude toward school and quality of relationship with peers and teachers). Hierarchical linear modeling shows that, among the social capital control variables, being properly integrated into the family and frequent peer interaction significantly add to mental health. At the individual (student) level only, ties to delinquent friends are negatively associated with mental health, while at both individual and contextual levels, school characteristics are positively related to adolescent subjective well-being.

Contextual factors are critical in maintaining and promoting mental and physical well-being. Much evidence indicates that health outcomes are significantly shaped by factors that lie above and beyond individual actors. In particular, social ties and related social support are shown to be valuable in facilitating subjective and physical health (Berkman et al., 2000, Ferlander, 2007, Kawachi et al., 2008, Smith and Christakis, 2008 and Umberson and Montez, 2010). People who are embedded in better social relations are happier and more satisfied with life generally (Haller and Hadler, 2006, Kroll, 2011 and van der Horstm and Coffe, 2011). Those with greater social capital (e.g., interpersonal trust, number of friends, frequency of interaction) are also more likely to report themselves as being physically healthier (Cornwell and Waite, 2009, Ferlander and Makinen, 2009, Fujiwara and Kawachi, 2008, Giordano et al., 2012, Song and Lin, 2009 and Verhaeghe and Tampubolon, 2012). There are ongoing debates concerning the precise mechanisms linking social factors and health outcomes such as social influence, social control, information transfer, instrumental assistance, and emotional belongingness (Thoits, 2011 and Umberson and Montez, 2010). Despite the conceptual and causal uncertainty surrounding the role of social context, a plethora of studies support the claim that physical and mental health is strongly associated with the quantity and quality of social connectedness and other contextual factors. In addition to social or network embeddedness, previous research focuses on the impact of broader environmental factors, in particular, the role of neighborhood, on health outcomes (Browning and Cagney, 2002, Haines et al., 2011, van Hooijdonk et al., 2008, Mohnen et al., 2011 and Moore et al., 2011). In making this connection, scholars conjure up the notion of macro-level social capital or collective efficacy. The former is the resource accessible through one's membership in a higher-level group (Kawachi et al., 2008, Poortinga, 2006a and Poortinga, 2006b). For example, Mohnen et al. (2011) demonstrate that neighborhood social capital is significantly related to self-reported health based on a national sample. Verhaeghe and Tampubolon (2012) also find that individual-level social capital (general trust, social participation, network sources) mediates the deleterious effects of neighborhood disadvantage on health. The concept of collective efficacy, on the other hand, derives from a sense of social cohesion and informal social control typically embedded at the neighborhood level (Maimon et al., 2010 and Sampson et al., 1999). According to one study, neighborhood structural characteristics, social organization and culture are strong predictors of self-reported physical well-being (Browning and Cagney, 2002). In another study, neighborhood effects (via informal social control) and network social capital are found to interact to influence depressive symptoms among a representative sample of urban residents (Haines et al., 2011). The common denominator underlying these two related lines of inquiry is that the environmental context, however defined, constitutes a significant determinant of psychological and physiological health. 1.1. Aim of this study This study seeks to contribute to the social epidemiology literature in following ways. First, in examining the role of social capital, previous studies have conceptualized social network in largely functional terms (see Coleman, 1988). With the assumption that social ties produce health benefits, research efforts have been mostly directed at establishing a positive relationship between, for example, network size and self-rated health (e.g., Cornwell and Waite, 2009; Song, 2011; van der Horstm and Coffe, 2011). This study, however, proposes that network connectedness in and of itself may or may not lead to better health outcomes. The key is the quality of the social relationship, a topic which the existing literature does not adequately address. Depending on the quality of the tie, it is hypothesized that certain relationships can have an adverse effect on mental and psychological health of individuals involved. Second, most of the past research on social determinants of health is generally based on data from Western countries ( Kumar et al., 2012 and Sujarwoto and Tampubolon, 2013) and the studies on health and neighborhood effects in particular almost exclusively deal with adult populations ( Aminzadeh et al., 2013). Understanding the correlates of adolescent health is critical since health advantages and disadvantages of social capital is cumulative ( Umberson and Montez, 2010). The current study thus endeavors to bridge the empirical gap by focusing on a national youth sample in an East Asian context. School is one of the most important contextual factors influencing health behaviors and outcomes of adolescents (Bearman and Moody, 2004 and Maiman and Kuhl, 2008; for a review, see Whitlock et al., 2014). Despite the wide recognition of its critical role, however, the linkage between school connectedness and adolescent mental health has not been adequately examined in the literature (Shocet et al., 2006 and Oberle et al., 2011). This research moves in that direction by measuring multiple school-related characteristics and analyzing their associations with subjective well-being of students. Lastly, much of existing research relies on cross-sectional data and thus encounters the thorny problem of endogeneity. This study minimizes this methodological issue by utilizing panel data, which allows for a better temporal connection between health outcome and independent variables (Giordano et al., 2012).