نفوذ طولی از روابط پدر و مادر کودک و افسردگی در بزهکاری سایبری نوجوانان در کره جنوبی: مدل منحنی رشد نهفته

Abstract The purpose of this study is to longitudinally verify the influence of parent–child relationships and depression on adolescent cyber delinquency. Analyses were conducted on panel data of elementary school students from the Youth and Children Data Archive from 2004 to 2008. The results first indicated that the number of adolescents who commit cyber delinquency increases rapidly as their grade rises, with a peak increase occurring in the second grade in middle school (15 years old). In addition, the analysis of the intercept and ratio of change of cyber delinquency confirmed a difference between individuals in the experience on cyber delinquency. Third, the analysis of the factors potentially influencing the developmental process of cyber delinquency indicated that the intercept of the parent–child relationship and depression showed a significant influence. It was further shown that among variables related to the ratio of change in cyber delinquency, the ratio of change for the parent–child relationship and for depression both had a significant influence

. Introduction In accordance with the rapid development of information technology among OECD member countries, the Internet access rate per family in Korea has reached 80% or more, in line with nations such as The Netherlands, Denmark, Sweden, and Norway (OECD, 2010). Moreover, the Internet use rate for teenagers is approximately 99.9%, suggesting that almost all adolescents in South Korea are using the Internet (The Korea Internet and Security Agency, 2010). This is important because the Internet provides adolescents an open channel through which they can vent pressures related to school, family, and friends (Suler, 2000). Further, by not exposing age, sex, social status, or physical, psychological, and social identities, cyberspace provides an opportunity for adolescent activities that maintain anonymity (Cheon, 2000 and Hong and Kim, 2011). Accordingly, cyberspace has become the medium for a new teenage culture. However, the influence of the Internet can be not only positive but also negative, depending on users' intentions; in particular, it can be a platform for escapism and cyber delinquency (Jonson & Miller, 1998, June). This is particularly the case in Korea, where the incidence of juvenile delinquency in cyberspace is rapidly increasing every year. According to Korea's Cyber Police Agency, cyber delinquency by adolescents accounts for 19.5% (23,966 cases) of cyber delinquency in 2010. Despite the seriousness of the problem, there is a lack of empirical studies in cyberspace. Existing studies on cyber delinquency also focus mainly on middle or high school students despite the increasing need to focus on elementary school cyber delinquency. Keeping in mind the point that cyber-delinquency possesses super-national characteristics, its definition must have a rather inclusive and general meaning that can be applied across borders. Understanding of the concept of delinquency should therefore precede definition of cyber-delinquency. In general, delinquency, as a super-ordinate concept in the broad context, is a comprehensive criminal terminology used to describe the crimes committed by adolescents. In fact, criminologists especially use the term delinquency instead of crime when considering the misdeeds conducted by adolescents (Lee & Ahn, 2005). Delinquency, in general, includes the norms violated by adolescents, and it also covers diverse misconduct handled by the Children's Welfare Law. It can also be seen as the broad concept that includes the following five concepts in the field of criminology: crime, deviance, offense, sin, and vice. From this perspective, cyber-delinquency can be seen as cyber-crime committed by adolescents who violate the laws of cyberspace, as well as a variety of problematic behaviors, acts of deviation, and improper conduct unfit for the adolescents' status within the context of a multi-dimensional concept. Henceforth, cyber-delinquency separates itself from cyber-crime, which stands as a standard of important conduct, and works as the super-ordinate concept containing it. If a person who committed crime on cyberspace is an adult, the term cyber-crime is used. In contrast, cyber-delinquency refers to the criminal acts as well as acts of delinquency committed by adolescents. Treating cyber-delinquency as Internet-crime may be allowable. However due to the fact that Internet-crime does not include the crime that does not utilize the Internet as the medium, the term cyber-delinquency is more typically used in the broad context. However, the term cyber-delinquency, which refers to the acts of delinquency on cyberspace, as primarily mentioned in this research, is not used universally. This is due to the fact that the problem associated with delinquent acts of adolescents has been rapidly growing with the help of the development of electronic information technology, and also that research into cyber-delinquency has not been frequently conducted. Nevertheless, taking all of the information covered in the study into account, cyber-delinquency indicates all types of delinquent acts or crimes committed by adolescents on cyber space and can also be defined to include many problematic behaviors and acts of deviance in addition to cyber-crime that go against the law. Cyber delinquency activities include criminal acts violating laws such as hacking, virus infringement, and use of false resident registration numbers as well as an act generally recognized as delinquent such as searching for pornography, obscene and violent chatting, and illegal copying of software (Han, 2001, Hong and Kim, 2011 and Jo and Yang, 2001). Therefore, by applying this concept in a broad sense, adolescent cyber-delinquency is defined as all associated acts of crime and delinquency committed by adolescents in and around cyber space. Juvenile delinquents are those whose adaptation to conflicts and situational stresses arising in a transitional developmental process from childhood to adulthood results in deep conflicts with parents, psychological problems such as anxiety and depression, and behavioral problems, including drug use or antisocial actions (Larson et al., 1998 and Lyons, 2004). The present study aims to approach adolescent cyber delinquency by focusing on these universal characteristics of juvenile delinquents. Theories of delinquency that reflect the characteristics of juvenile delinquents include strain theory (Agnew, 1992) and social bonding theory (Hirschi, 1969). Agnew (1992) suggests in his strain theory that all strains that adolescents experience in daily life can become the cause of their delinquency. The typical strains are thought to result from marital discord between parents and conflicts with friends. Moreover, as such strains in daily life generate negative emotions such as depression, they can cause delinquency. The social bonding theory proposed by Hirschi (1969) further suggests that social bonding relations of adolescents–that is, relationships with parents and a peer group–are a critical factor in delinquency. This approach specifically suggests that those who have close ties with parents fear that delinquent acts will disappoint them. Accordingly, adolescents who are not attached to their parents are more likely to commit delinquent acts, including violation (Cemkovich & Giordano, 1987). Unlike traditional offline delinquency, many different situations in cyberspace afford adolescent anonymity; consequently, they feel free to violate social norms, unlike in the real world. Given the limited applicability of the above theories, which aim to explain delinquency in the real world, rather than the cyber world, theories of cyber delinquency are comparatively unconstrained and require empirical verification. In particular, existing studies report a close correlation between parent–child relationships and depression with cyber delinquency. First, parent–child relationships are closely related to adolescent delinquency, thus potentially explaining cyber delinquency. In cases where parent–child relationships are associated with negative qualities, adolescents may become depressed and commit delinquent acts in cyberspace (Kwon, 2005, Han, 2001 and Kim, 2009). In addition, Satir suggests that in a family where parents communicate with their children in an unclear and doubly tying communication form, parents may have a low level of self-worth and use children for their own values. In such cases, adolescents are predisposed to adjustment disorders, including delinquency or mental health problems (Barnes and Olson, 1985 and Choi and Hong, 1997). Importantly, though depression and delinquency are commonly found adjustment disorders with a high comorbidity (Wolff & Ollendick, 2006), Beyers and Loeber (2003), examining the developmental relation between depression and delinquency, showed that depression predicts an increase in delinquency, while delinquency did not predict an increase in depression. This suggests that depression precedes delinquency. In the light of these results, this study provides a longitudinal evaluation of the kinds of influences parent–child relationships and individually internalized problems (including depression) have on adolescent cyber delinquency. This work builds on the results of many earlier studies, including those on the strain theory and social bonding theory. This study specifically aims to intensively examine (1) how cyber delinquency changes over time, (2) the kinds of influences parent–child relationships have on changes in cyber delinquency over time, (3) the kinds of influences parent–child relationships have on changes in depression over time, and (4) the kinds of influences depression has on changes in cyber delinquency over time.

Results 3.1. Unconditional model This study explored how adolescent cyber delinquency, parent–child relationships, and depression changed over time. Furthermore, the study considered the causality of cyber delinquency, parent–child relationships, and depression, varying the measuring points of time as shown in Table 4. Table 4. Mean and standard deviation by each variable. Variable First survey Second survey Third survey Fourth survey Fifth survey M SD M SD M SD M SD M SD Parent–child relationship 3.579 .732 3.613 .745 3.634 .708 3.535 .741 3.503 .750 Depression 2.186 .907 2.185 .920 2.273 .908 2.356 .893 2.389 .912 Cyber delinquency .067 .144 .097 .172 .104 .187 .109 .182 .102 .170 Table options In order to figure out the change of variables measured at each time point, the model calculated the average of each variable first. Analyses of the experience of cyber delinquency over time showed that the frequency of cyber delinquency increased over a period of 4 years from the fourth grade in elementary school (11 years old) to the first grade in middle school (15 years old), and then decreased once adolescents reached the second grade in middle school (16 years old). This result corresponds to that of existing studies on delinquency in which the frequency of cyber delinquency reaches its peak in middle adolescence (15–16 years old) and then decreases thereafter (Hoeve et al., 2008, Jung, 2010, Landsheer and van Dijkum, 2005 and Thornberry, 1987). Before applying the latent growth curve model, which estimates relations based on the level of change of cyber delinquency, parent–child relationships, and depression, an appropriate model with an appropriate change function for each variable was selected. Based on the average change at three points in time suggested in Table 4, both nonlinear and linear changes were applied to each variable. After reviewing the average change at each point in time, it assessed the suitability of the models to select the appropriate change model for each factor. Table 5 shows the mean and variance of the suitability of the non-linear model for the intercept and each variable along with the mean and variance of the slope. As suggested in Table 5, when applying the variables of parent–child relationships, depression, and cyber delinquency within a nonlinear change model, the result model is not suitable. On the other hand, application of these variables within a linear change model yielded a satisfactory result. This indicates that parent–child relationships, depression, and cyber delinquency change significantly over time. The significant variance for the intercept and slope of these variables also shows that there are differences among individuals. Table 5. Results of each variable's model suitability (N = 2844). x2 (df) TLI CFI RMSEA Intercept Slope Mean Variance Mean Variance Parent–child relationships Nonlinear change 381.314 (13) .885 .900 .100 Linear change 130.532 (10) .951 .967 .065 3.616⁎⁎⁎ −.021⁎⁎⁎ .303⁎⁎⁎ .018⁎⁎⁎ Depression Nonlinear change 364.114 (13) .803 .829 .097 Linear change 59.686 (10) .964 .976 .042 2.167⁎⁎⁎ .057⁎⁎⁎ .350⁎⁎⁎ .027⁎⁎⁎ Cyber delinquency Nonlinear change 453.929 (13) .670 .714 .109 Linear change 180.180 (10) .835 .890 .077 .076⁎⁎⁎ .003⁎⁎ .009⁎⁎⁎ .001⁎⁎⁎ ⁎⁎⁎ p < .001. ⁎⁎ p < .01. Table options 3.2. Conditional model Given the results of the unconditional model in the first stage, we next analyzed the longitudinal influences of parent–child relationships and depression on cyber delinquency in the second stage. The latent growth curve model verified the multi-variate latent growth curve model establishing the relation between parent–child relationships, depression, and cyber delinquency. The suitability of the model was satisfactory, yielding CMIN (df) = 554.044(96), TLI = .927, CFI = .942, and RMSEA = .041. With the suitability of the model verified, the kinds of influences parent–child relationships and depression had on cyber delinquency were next examined. As shown in Table 6, all courses were shown to be significant with the exception of the influence of the parent–child relationship intercept and the depression intercept on cyber delinquency. First, the intercept of parent–child relationships had a significantly negative influence on the intercept of cyber delinquency (β = −.153, p < .001). This means that when parent–child relationships were good in the first year, the level of cyber delinquency in the first year was low. Next, the intercept of parent–child relationships had a significantly negative influence on the intercept of cyber delinquency (β = −.194, p < .001). This means that when parent–child relationships improve rapidly, the frequency of cyber delinquency also decreases rapidly; when parent–child relationships improve slowly, the frequency of cyber delinquency also decreases slowly. Accordingly, it can be shown that when the value of parent–child relationships increases sharply from fourth grade in elementary school to second grade in middle school, the value of cyber delinquency increases only slightly. On the contrary, this means that for those whose parent–child relationships increased, the problem of cyber delinquency increased at a relatively higher rate. Such a result corresponds to the case of elementary school students in which the children of parents reporting laissez-faire child-rearing attitudes exhibited an increased tendency to become addicted to games relative to other groups (Lee & Park, 2004). Table 6. Influence of parent–child relationships and depression on cyber delinquency. B β S.E. Intercept of parent–child relationship → intercept of cyber delinquency −.026⁎⁎⁎ −.153 .006 Intercept of parent–child relationship → slope of cyber delinquency −.003 −.050 .002 Slope of parent–child relationship → cyber delinquency −.044⁎⁎⁎ −.194 .012 Intercept of parent–child relationship → intercept of depression −.388⁎⁎⁎ −.359 .033 Intercept of parent–child relationship → slope of depression .026⁎ .086 .012 Slope of parent–child relationship → slope of depression −.497⁎⁎⁎ −.399 .053 Intercept of depression → intercept of cyber delinquency .044⁎⁎⁎ .286 .006 Intercept of depression → slope of cyber delinquency −.002 −.038 .002 Slope of depression → slope of cyber delinquency .023⁎ .125 .010 ⁎⁎⁎ p < .001. ⁎ p < .05. Table options The intercept of parent–child relationships also had a significantly negative influence on the intercept of depression (β = −.359, p < .001). This means that when the value of the parent–child relationship in the first year is high, depression in the first year is low. In addition, the intercept of the parent–child relationship had a significantly positive influence on the slope of depression (β = .086, p < .001). This means that the higher the parent–child relationship, the slower depression increases. Finally, the intercept of depression was positively and significantly related to the cyber delinquency intercept (β = .286, p < .001). The slope of depression was positively and significantly related to the cyber delinquency slope (β = .023, p < .05). This means that the higher the value of depression in the first year, the higher the cyber delinquency in the first year. In addition, the faster depression increased, the faster cyber delinquency increased. In such a longitudinal structural relation, the intercept and slope of the parent–child relationship showed a significant influence on the intercept and slope of depression. The intercept and slope of depression, in turn, showed a significant influence on the intercept and slope of cyber delinquency. Through such a relation, it can be said that depression plays an intermediate role in the influence of parent–child relationships on cyber delinquency. Whether depression has an influence on delinquency or delinquency has an influence on depression is under discussion (see Lee, 2004).