قرار گرفتن مصرف الکل و بزهکاری در زمینه توسعه و اجتماعی: مورد جوانان کره ای

Abstract In this study involving self-report questionnaire data from 955 tenth-grade students in three locations within Korea, we address the meanings of alcohol use and delinquency for Korean youth. Findings (a) supported a facilitative role for alcohol, but not delinquency, with respect to perceived peer social competence; (b) indicated negative associations of both alcohol use and delinquency with parental relations, valuing academic achievement, and collectivistic values, and positive associations with friends’ risk behaviours; (c) showed no relationship of these behaviours with self-esteem, coping, parental permissiveness or individualistic values; and (d) revealed that perceived benefits of alcohol use and delinquency include not only social facilitation but also exploration and assertion of independence, suggesting a potential connection, as in Western societies, between risk taking, and identity exploration.

Introduction In this study, we explore the meanings of alcohol use and delinquency for Korean adolescents. The Korean context is of interest because, despite societal values and practices that have traditionally been viewed as protective for adolescent risk taking, there have been marked increases in Korean adolescent risk behaviours in recent years. In what follows, we first highlight some of the key findings from Western studies regarding the role of risk taking in adolescent development. Then, following a brief description of Korean society, we discuss the specific goals and research questions for the present study. Throughout, we use the terms risk taking and risk behaviours interchangeably to refer to volitional actions on the part of adolescents that have potential negative health or developmental outcomes. Risk taking and development As pointed out many years ago by Baumrind (1987) and Baumrind (1991), if we are to gain an understanding of adolescent risk behaviours, we must consider the developmental context and functions of these behaviours. While the use of alcohol and other drugs may lead to negative consequences, for many adolescents, experimentation does not signal problematic development. It may in fact be indicative of adaptive identity exploration and play a role in facilitating peer relations (Maggs, 1997; Shedler & Block, 1990; Silbereisen & Noack, 1988; Silbereisen & Reitzle, 1991). In her developmental analysis of adolescent risk taking, Lightfoot (1997) emphasizes the role of risk taking in self-construction and interpersonal relationships. The teens in her study talk of risk taking as pushing the limits, exploring new territory, and in the process, learning about themselves in these new situations. There is a sense of accomplishment that comes from gaining control in unpredictable situations, and “getting away with” something disapproved by adults. They talk of risk as creating excitement and also as serving to reduce tension. Taking risks with peers can provide shared memories, and serve to enhance a group identity, cohesion, and feelings of closeness among group members. However, the picture is complex: the consequences of risk taking vary depending upon many factors including the nature of the behaviour and the type of developmental competencies or relationships being considered. For example, Maggs and Hurrelmann (1998), in a longitudinal study of German adolescents (spanning grades 7–10), found support for their predictions that substance use, in contrast to delinquency, would be positively related prospectively to several indicators of peer relationships. They reasoned that delinquency does not play a facilitative role in peer relations to the same extent as does substance use because delinquency is less normative, cannot be viewed as healthy exploration, and is viewed more negatively by society. Among a sample of seventh- and eighth-grade students, Allen, Weissberg, and Hawkins (1989) found that conforming to adult norms regarding risk behaviours related negatively to peer sociometric ratings and self-reported social competence, and positively to teacher ratings of class behaviour, GPA, and self-reported cognitive competence. Similarly, Maggs, Almeida, and Galambos (1995) reported, in their third-wave, 14-year-old sample, positive associations between risk behaviours and peer involvement and perceived peer acceptance, but negative associations between risk behaviours and self-image. Bell, Forthun, and Sun (2000) studied risk behaviours (alcohol and marijuana, problematic or non-problematic use) in relation to parental attachment and several adolescent competencies, predicting different patterns of associations depending upon type of substance and extent of use. Supporting expectations, peer competency was positively related to both alcohol (women only) and marijuana use, but was not related to substance use problems. Parental attachment was not related to alcohol use, but was negatively related to both marijuana use and substance use problems. These studies speak to the conflicting contextual demands facing adolescents as they strive for competency in the overlapping yet distinct worlds of peers, parents, and school. They explore the variability in meanings, or consequences, of risk behaviours depending upon particular behaviours within a particular context. Substance use appears to mean something different for adolescents than does delinquency. Risk behaviours seemingly have certain implications for peer social competency and identity exploration, but different implications in terms of parental relations and academic achievement. Level of involvement in risk behaviours also matters—experimentation must be differentiated from developmentally dysfunctional involvement. Korean context Korea has been characterized as a collectivistic society in which harmony, cooperation, and consensus are encouraged. One implication of collectivism for adolescent development, proposed by Chun and MacDermid (1997), is that “intergenerational fusion” may be more adaptive for adolescents in Korean society than is individuation, which is thought to be an indicator of healthy development in Western societies. Another implication is the suggestion that the interdependence of selves in collectivistic compared to individualistic societies tends to encourage suppression of behaviour that is socially disapproved (Greenberger, Chen, Beam, Whang, & Dong, 2000), resulting in lower levels of adolescent risk behaviours. Because of concerns in recent years about the loss of national identity due to Western influences, there has been a re-assertion of traditional values, called the “new Confucian ethics” (Oh, 1997). Even more than in other Asian societies, the family is of central importance (Lee, 1977). This “familism” is characterized by parental authority and filial piety; parents are granted absolute authority, and children are expected to be obedient and respectful to parent and to make sacrifices for the family (Chung & Yoo, 2000; Hurh, 1998). Another important aspect of Korean society is its emphasis on education. This focus on education stems from the Confucian emphasis on the importance of self-improvement and discipline—values of wisdom and competence that can be gained through education. Education also is viewed as essential for financial security and as an indicator of family status. Hence, there is a great deal of pressure on children to succeed academically (Oh, 1997; Yang, 1988). Korean society has undergone and continues to experience substantial economic, political, and social changes (Greenberger et al., 2000). Korea is in transition to a more global economy, accompanied by a greater diversity of cultural input. Influenced by Western music, movies, and life styles, Korean adolescents are more exposed to Western cultural values and beliefs than was true in the past. Several recent studies, reviewed by Yoon (2004), indicate that adolescents hold less traditional attitudes than do their parents regarding topics such as filial piety, gender roles, and child rearing. Adolescent alcohol use also has been increasing in recent years. Based on a sample of 13,000 middle- and high-school students, the percentage of students who reported abstaining from alcohol during the past month dropped from 72.3% in 1999 to 59.8% in 2002 (Commission on Youth Protection, 2002). Cross-cultural research on adolescent risk behaviours Cross-cultural studies have indicated that European and American adolescents are more likely to engage in risk behaviours than are Asian and African adolescents (Eide, Acuda, Khan, Aaroe, & Loeb, 1997; Feldman, Rosenthal, Mont-Reynaud, Leung, & Lau, 1991; Greenberger et al., 2000; see Chen, Greenberger, Lester, Dong & Guo, 1998, for an exception). Greenberger et al. (2000) found that American adolescents engaged in more misconduct and perceived more tolerance from others of these behaviours than did Korean adolescents, and Korean adolescents exceeded Chinese adolescents in these respects. There were some group differences in correlates, but the most important correlates of misconduct for all groups (although this varied to some extent by gender) were close friends’ perceived attitudes and behaviours regarding misconduct. These societal differences in the incidence of adolescent risk behaviours have been interpreted in terms of variations in cultural values and socialization practices (e.g. Arnett, 1992; Feldman et al., 1991), for example, to differences in values regarding personal freedom, conformity, collective welfare, obedience, and the role of family (Bond, 1988). Greenberger et al. (2000) suggested that social and economic change modifies normative expectations for adolescents. Accompanying these changes is a reduction in consensus about values for adults as well as for adolescents. Thus, there is a greater diversity of options for adolescents in societies with more individualistic orientations and more global contacts. Our purpose in this study is not to test the above speculations about social and economic change within Korea and its impact upon norms of adolescent conduct. Rather, we are interested in the adolescents’ perspectives on risk taking. Despite the increasing incidence of adolescent risk behaviours in Korea, research has not addressed the meanings of these behaviours for Korean adolescents. As argued by Baumrind (1987), Lightfoot (1997), and many others, we cannot adequately understand adolescent risk taking apart from its developmental context, i.e. the way it relates to the developmental “tasks” of adolescence, or its developmental meanings for the adolescent. And these developmental expectations may vary across cultural contexts. For example, although Western investigators have identified peer relations as a domain in which risk taking may be developmentally facilitative, these findings may not extend to societies where it is not an expectation that adolescents increasingly engage with peers and begin to assert independence from parents. In this study, we target the specific risk behaviours of alcohol use and delinquency, and we assess meanings of these behaviours in two ways, first, via their correlations with other variables, and second, by obtaining adolescents’ perspectives of the costs and benefits of each. Research questions Our primary research question was whether we would find, as in the Western studies described earlier (e.g. Maggs & Hurrelmann, 1998), a positive relationship between risk behaviours and peer social competence and, if so, whether alcohol use differs from delinquency in this respect. Second, because Korea is characterized as a collectivistic society, we were interested in whether delinquency, which often infringes upon the rights of others, has more negative implications for Korean adolescents than does alcohol use. Specifically, this question concerns differential associations of the two risk behaviours with measures of parental relationships, several adolescent competencies, and friends’ risk behaviours. Our third question was whether individualistic in contrast to collectivistic values would relate positively to alcohol use and delinquency. Korea is characterized as a predominantly collectivistic society, yet this characterization tends to minimize within-society variation in values, which may be substantial. Thus, we focus here on individual differences among Korean adolescents in these values. Finally, our fourth question concerned perceived benefits and costs of alcohol use and delinquency. We included this assessment in order to gain insight into reasons for engaging (or not engaging) in the behaviours from the Korean adolescent's perspective.

Results Preliminary analyses Alcohol use and delinquency Nearly half (48.3%) of the students reported using alcohol in the past three months. The amount of use was 1–2 times for 22.6% of the sample, 3–5 times for 11.1%, 6–9 times for 4%, and 10 times or more for 10.6%. Asked whether they had had three or more drinks on one occasion, 34% reported that they had (9.1% said 10 times or more). And 20.8% of students reported getting drunk (5% 10 times or more). For the delinquency items, only a small percentage of the sample reported that they had stolen something worth less than $50 (6.6%), gotten into trouble with the police (7.4%), or skipped school for a day (6.9%). Larger percentages of students reported having cut class (14.8%), shoplifted from a store (17.1%), destroyed school or other public property (19.8%), cheated on a test (22.4%), or gotten into a fight at school (29.4%). Alcohol use and delinquency were related to one another, chi square (2)=152.44, p<.001p<.001. Of those who reported no alcohol use, 33.5% reported one or more incidences of delinquent behaviour. For those reporting some alcohol use (1–2 times), 55.9% reported one or more incidences of delinquent behaviour, and for those in the highest alcohol use category, 83.8% reported delinquent behaviour. Demographic variables Alcohol use was significantly related only to age and geographic location (Table 2). Use occurred more frequently for older adolescents, and for the mid-size location (Yeongdeok). Delinquency, however, was significantly related to nearly all demographic variables, occurring more frequently among males and older adolescents, for the mid-size location (Yeongdeok), and for those whose parents had completed fewer years of education and whose fathers worked in less-skilled occupations. Correlations among predictor variables Among the parenting variables, strong positive associations were found for mother and father attachment, for attachment and parental permissiveness, and for attachment and parental monitoring (Table 3). Parental monitoring was unrelated to permissiveness. The strongest correlations among the adolescent competency variables occurred for self-esteem in relation to the other variables—coping, valuing achievement, social competence, and also for coping in relation to social competence. Particularly noteworthy in these correlations is that individualism and collectivism, as measured here, are not opposing constructs (they are positively correlated with one another), but their patterns of correlations with other variables are somewhat different. While individualism is strongly correlated only with self-esteem and valuing achievement, collectivism is more related to parental attachment, parental monitoring, social competence, and self-esteem. Table 3. Correlations among predictor variables MAT FAT MP FP PM CO S-E VA SC FR IND COL MAT FAT .46⁎⁎ MP .60⁎⁎ .24⁎⁎ FP .29⁎⁎ .50⁎⁎ .49⁎⁎ PM .34⁎⁎ .32⁎⁎ −.02 .01 CO .08⁎ .16⁎⁎ .05 .13⁎⁎ .01 S-E .31⁎⁎ .31⁎⁎ .20⁎⁎ .18⁎⁎ .31⁎⁎ .24⁎⁎ VA .19⁎⁎ .09⁎⁎ .12⁎⁎ .07⁎ .19⁎⁎ −.13⁎⁎ .22⁎⁎ SC .23⁎⁎ .22⁎⁎ .12⁎⁎ .12⁎⁎ .15⁎⁎ .30⁎⁎ .36⁎⁎ .04 FR −.08⁎⁎ −.11⁎⁎ −.02 −.07⁎ −.12⁎⁎ −.08⁎ .01 −.09⁎⁎ .06 IND .16⁎⁎ .14⁎⁎ .16⁎⁎ .11⁎⁎ .18⁎⁎ −.06 .42⁎⁎ .65⁎⁎ .09⁎⁎ .01 COL .37⁎⁎ .36⁎⁎ .18⁎⁎ .13⁎⁎ .30⁎⁎ .04 .43⁎⁎ .24⁎⁎ .32⁎⁎ −.03 .38⁎⁎ Note: MAT: mother attachment; FAT: father attachment; MP: mother permissiveness; FP: father permissiveness; PM: parental monitoring; CO: coping; S-E: self-esteem; VA: value achievement; SC: social competence; FR: friends’ risk behaviours; IND: individualism; COL: collectivism. ⁎ p<.05. ⁎⁎ p<.01. Table options Predictors of alcohol use and delinquency Based upon discriminant analyses which included all of the predictor variables (as listed in Table 4), one significant discriminant function was found for both alcohol use and delinquency, p's<.001. The canonical correlations for the discriminant functions were .50 and .30, respectively, indicating a somewhat stronger relationship between the predictor and criterion variables for alcohol use than for delinquency. This could be due, in part, to the greater variability of alcohol use compared with delinquency. Table 4. Discriminant analyses for alcohol use and delinquency Structure coefficients Alcohol use Delinquency Friends’ risk behaviours .91 .90 Social competence .25 .14 Parental monitoring −.23 −.40 Father attachment −.16 −.25 Mother attachment −.14 −.18 Value achievement −.14 −.21 Individualism −.04 −.13 Collectivism −.09 −.19 Self-esteem −.02 −.18 Coping .02 .02 Father permissiveness .02 −.05 Mother permissiveness −.02 −.06 Table options The importance of each of the variables in the discriminant functions is indicated by structure coefficients (Table 4). For alcohol use, tukey tests (Table 5) indicated that the four variables with the highest structure coefficients significantly differentiated among alcohol use groups, and p values for two additional variables were close to the .05 criterion. For three of these variables (parental monitoring, mother attachment, valuing achievement), the distinction was between the abstaining group and the two use groups which did not differ from one another. For the other three variables (friends’ risk behaviours, social competence, father attachment), group means were ordered such that the low-use group fell midway between abstainers and the higher use group. In terms of the direction of group differences, scores on parental monitoring, father attachment, mother attachment, and valuing achievement were negatively related to alcohol use, while scores on social competence and friends’ risk behaviours were positively associated with use. Table 5. Alcohol use: group means and Tukey tests None One to two times Three or more Friends’ risk behaviours⁎ 1.62 1.92 2.57 Social competence⁎ 3.38a 3.52ab 3.64b Parental monitoring⁎ 3.40a 3.31b 3.08b Father attachment⁎ 3.03a 2.96ab 2.83b Mother attachment+ 3.65a 3.50a 3.48a Value achievement+ 3.65a 3.46a 3.44a Individualism 3.35a 3.33a 3.31a Collectivism 3.62a 3.60a 3.54a Self-esteem 3.17a 3.13a 3.15a Coping 2.99a 3.06a 3.00a Father permissiveness 2.87a 2.88a 2.89a Mother permissiveness 2.98a 2.85a 2.97a Note: Group labels refer to average scale scores for the three alchol use items. Time frame was past 3 months. Means with different superscripts are significantly different. ⁎ p=.05. + p<.07. Table options Delinquency group means (Table 6) indicated that, for the most part, the same variables important for differentiating alcohol use groups also differentiated delinquency groups: friends risk behaviours, parental monitoring, father attachment, and valuing achievement. However, unlike alcohol use, social competence was not significantly different for the delinquency groups, nor was mother attachment. Table 6. Delinquency: group means No involvement Some involvement Friends’ risk behaviours⁎ 1.69 2.11 Parenting monitoring⁎ 3.43 3.19 Father attachment⁎ 3.03 2.90 Value achievment⁎ 3.62 3.48 Collectivism+ 3.64 3.57 Mother attachment 3.62 3.52 Mother permissiveness 2.96 2.93 Father permissiveness 2.89 2.87 Coping 3.01 3.01 Self-esteem 3.19 3.12 Social competence 3.44 3.51 Individualism 3.36 3.31 Significant group differences. ⁎ p<.05. + p<.07. Table options Finally, we conducted MANOVAs to determine whether the above findings were modified by the demographic variables of location, age, and gender. Predictor variables in the MANOVAs were alcohol use category (or delinquency category), the three demographic variables, and first-order interactions between alcohol use (or delinquency) and these demographic variables. Criterion variables were all of the variables listed in Table 4. In the delinquency analysis, there were no significant multivariate interactions between delinquency and the demographic variables. In the alcohol use analysis, there was one significant multivariate interaction, and this was between alcohol use and age, Wilk's λ F(24,1760)=1.97F(24,1760)=1.97, p=.003p=.003. Examination of individual variables indicated that this multivariate interaction occurred because of an interaction between use and age for friends’ risk behaviours, F(2,890)=9.03F(2,890)=9.03, p<.001p<.001. A plot of this interaction showed only minor variations by age. Despite these minor differences, there was a clear increase in reports of friends’ risk behaviours across alcohol use groups for both ages. Perceived benefits and costs of substance use and delinquency In discriminant analyses, one significant discriminant function was found for perceived benefits and costs as predictors of alcohol use, p <.001. The canonical correlation was .40, and structure coefficients were .80 and −.68 for benefits and costs, respectively. Based on one-way ANOVAs, both perceived benefits and costs (criterion variables) were significantly related to alcohol use, F(2,952)=56.95F(2,952)=56.95, p<.001p<.001; and F(2,952)=40.47F(2,952)=40.47, p<.001p<.001, respectively. Tukey tests indicated significant differences among all alcohol use groups for perceived benefits (none: M=2.06M=2.06; one–two times: M=2.32M=2.32; three or more: M=2.84M=2.84) and also for perceived costs (none: M=3.60M=3.60; one-two times: M=3.39M=3.39; three or more: M=2.97M=2.97). In a MANOVA which included alcohol use, gender, age, and location as predictor variables, no significant interactions of these demographic variables with alcohol use were found. The three most important benefits perceived by students for substance use by teens were: (a) It helps them forget their problems (M=2.93M=2.93); (b) They want to see what it's like (M=2.87M=2.87); (c) It helps them have more fun with their friends (M=2.59M=2.59). The three most important costs were: (a) They don’t want to mess up their bodies (M=3.95M=3.95); (b) Their parent(s)/guardian(s) won’t approve (M=3.81M=3.81); (c) Their school, athletic or work performance would suffer (M=3.58M=3.58). The importance of these benefits and costs did not vary by group: they were the top rated reasons for all three alcohol use groups. Unlike the findings for alcohol use, perceived costs and benefits of delinquency did not significantly discriminate delinquency groups, p=.09, canonical correlation=.07. Overall, the three most important perceived benefits of delinquency were: (a) They want to fit in with their friends (M=2.71); (b) They want to show their parents they can do what they want (M=2.69); (c) It helps them forget their problems (M=2.66). The most important costs were: (a) Their parents(s)/guardians won’t approve (M=3.89); (b) Their school, athletic or work performance would suffer (M=3.72); (c) It might mess up their future plans for college or a career (M=3.60).