Structural Model of Deliquent Behavior Influenced by Media Violence in Korea

Kim, Hun Soo; Kim, Hyun sil

Abstract

OBJECTIVE
The purpose of this study was to test theoretical model and to build up of final modified model for explaining juvenile delinquency by media violence. METHOD: Data were collected through questionnaire survey over a period of 2 months. The subjects served for this study consisted of 1,943 adolescents including 707 delinquent adolescents and 1,236 student adolescents in Korea, sampled from Korean student population and delinquent adolescent population in juvenile correctional institions, using proportional stratified random sampling method. But, only 537 respondents including 217 delinquent adolescents and 320 student adolescents among 1,943 that replied perfectly to questionnaires were employed for this study. In this study, exogeneous variable was family dynamic environment, and endogeneous variables were character of adolescents including need satisfaction/frustration, sociability, antisocial personality tendency, complaints of psychosomatic symptoms and depressive trend, juvenile delinquent behavior and media violence themes including the extent of interest in and exposure and modelling impulsiveness and modelling to media violence themes. A total of 18 instruments were used to operationalized concepts in this model. Statistical method employed was covariance structural modelling.
RESULTS
For build up of the best fit model, the authors proceed by saturating the model, that is, including an arrow for all the possible relationships among all the variables. Afterward, a progressive elimination of the nonsignificant relationships, based on the betas that did not show a p<0.05 levels, was performed, this level was chosen for reasons of parsimony since the size of sample implied many statistically significance. The results of this analysis were as follows : In order to investigate the fitness of theoretical model to our data, Chi-square and goodness of fit index were identified. These results of chi-square value and fitness index score did not to be fit to the theoretical model which hypothesized the causal effect among various variables established in this study. Therefore the authors removed the endogeneous variable, sociability among character of adolescent because of being not significantly influenced on delinquent behavior for constructing the best fit model. The parsimonious best fit model included the various constructs defining family dynamic environment, need frustration, antisocial personality pattern, complaints of psychosomatic symptoms, depressive trend and media violence themes as direct paths to delinquent behavior. This final modified model yielded a very definitely improved goodness of fit indices. According to final modified model, it was revealed that the effects of media violence, complaints of psychosomatic symptoms, antisocial personality tendency on juvenile delinquency were significant positively higher. Especially, the most exerting influential variable on delinquent behavior in this model was media violence themes. The family dynamic environment also, as direct and/or indirect paths, was assumed a positive influence on juvenile delinquency as well as media violence themes.
CONCLUSIONS
In brief, it was found that media violence served as the most direct contributor to juvenile delinquency. That is, as the adolescents had the higher extent of interest in and exposure and modelling impulsiveness and modelling to media violence themes, they showed the more frequency of delinquent behavior. The single most powerful contributor by total effect to the development of delinquent behavior identified in this study was the family dynamic environment. That is, as the adolescents had the more unstable family dynamic environment, they became the more frustrated to their psychological need, and revealed the more maladaptive personality pattern, consequently they behaved the higher misconducts such as juvenile delinquency through media violence.