본 연구는 우리나라 중·고등학교 여자청소년의 폭력행위 실태를 살펴보고 그 폭력행위 예측요인으로써 가족관계와 가족구조, 교우관계, 학교관계가 여자청소년의 폭력행위에 어떤 영향을 미치는지 살펴보는데 그 목적이 있다. 2006년 5월 8일부터 2006월 5월 30일까지 한국청소년정책연구원에서 약 한 달 동안 서울시내 25개구에서 총 22개 중·고등학교를 대상으로 1,122명으로 조사한 데이터를 2차 자료로 활용하였다. 이 중 여자청소년은 550명으로 전체의 49%에 해당하였다. 통계분석 결과, 비록 t-test에서는 여자청소년 폭력집단과 여자청소년 비폭력집단이 부모관계와 학교관계 정도에서 통계적으로 유의미한 집단 차이를 보였으나 로지스틱 회귀분석 모델에서는 가족형태와 가족관계 두 요인만 여자청소년 폭력행위에 영향을 미치는 유의미한 변수로 확인되었다. 남자청소년의 폭력행위 예측모델을 분석하여 여자청소년 예측모델과 비교한 결과, 남자청소년에서는 가족구조 외에 교우관계와 학교관계가 폭력행위에 영향을 미치는 중요한 변수인 것으로 나타났다. 흥미로운 것은 남자청소년의 경우, 교우관계가 긍정적일수록 폭력행위가 증가할 가능성이 큰 것으로 나타났는바 청소년 범죄의 학습 및 차별적 접촉이론이 남자청소년에게 어느 정도 설명력을 갖는 것으로 보였다. 이하 본 연구에서 통계분석을 근거로 여자청소년폭력행위 예방을 위한 정책적 대안을 제시한다.

The current study explores the effect of family relationship on violent female juveniles. Although strong correlations have been shown between aversive family relationship and violence behavior, there is paucity of studies that focus on unique characteristics of violent female youth. It is unclear whether or not the impact of dysfuctional family relationship have more serious impact on young females than young males, and if so, to what extent. This study employs Korea Social Science Data (KOSSDA) in order to measure how aversive family relationships are associated with girls' violence. Specifically, this study examines the relationships between young female's violence and score of relationship with their parents. In addition, the present study uses scores of relationship with juveniles' friends and school teachers. The sample population consists of 1,122 students who are in middle school or high school in Seoul. The survey was conducted in 2006 for approximately a year by using cluster sample techniques. In logistic regression analysis, only two factors were identified as a significant predictors: parentless family and aversive family relationship. When types of school, income, relationship with friends, relationship with school teachers were controlled, on average the odds of committing violence behaviors for girls who lived with biological father and mother are 49% times lower than for those who lived with one-parent family (Exp(B)=0.51, p<0.05). Furthermore, on average, and holding the other variables in the model constant, the odds of committing violent acts for girls are 6% times are decreased as one unit of relationship score with parents are increased. The crucial issues that have implications for further juvenile delinquency policy on gender-specific correction will be discussed.

The current study explores the effect of family relationship on violent female juveniles. Although strong correlations have been shown between aversive family relationship and violence behavior, there is paucity of studies that focus on unique characteristics of violent female youth. It is unclear whether or not the impact of dysfuctional family relationship have more serious impact on young females than young males, and if so, to what extent. This study employs Korea Social Science Data (KOSSDA) in order to measure how aversive family relationships are associated with girls' violence. Specifically, this study examines the relationships between young female's violence and score of relationship with their parents. In addition, the present study uses scores of relationship with juveniles' friends and school teachers. The sample population consists of 1,122 students who are in middle school or high school in Seoul. The survey was conducted in 2006 for approximately a year by using cluster sample techniques. In logistic regression analysis, only two factors were identified as a significant predictors: parentless family and aversive family relationship. When types of school, income, relationship with friends, relationship with school teachers were controlled, on average the odds of committing violence behaviors for girls who lived with biological father and mother are 49% times lower than for those who lived with one-parent family (Exp(B)=0.51, p<0.05). Furthermore, on average, and holding the other variables in the model constant, the odds of committing violent acts for girls are 6% times are decreased as one unit of relationship score with parents are increased. The crucial issues that have implications for further juvenile delinquency policy on gender-specific correction will be discussed.

Female juveniles, violent behavior, Prediction model, family relationship, relational theory