Role of family structure and process variables in the explanation of children's peer violence

Socio-ekoloski modeli, uz individualne cimbenike, naglasavaju i važnost kontekstualnih cimbenika za pojavu vrsnjackog nasilja, a među kojima se isticu obitelj i skola. Cilj je istraživanja provjeriti prediktivnu vrijednost nekih strukturalnih i procesnih obiteljskih varijabli te varijabli mezosustava obitelji i skole za objasnjenje tjelesnog i verbalnog vrsnjackog nasilja. U istraživanju je sudjelovalo 880 ucenika od petog do osmog razreda osnovne skole, njihovi roditelji i razrednici. Podaci o nasilnickom ponasanju prikupljeni su samoprocjenama, samonominacijama i nominacijama ucenika iz razreda. Podaci o roditeljskim postupcima prikupljeni su od djece, dok su roditelji dali podatke o socioekonomskom statusu obitelji, poznavanju djetetovih vrsnjaka i ukljucenosti u djetetovo obrazovanje. Razrednici su izvijestili o ucestalosti kontakata roditelja i skole. Podaci su analizirani multivarijantnim viserazinskim modeliranjem. Od demografskih varijabli obje vrste nasilnickog ponasanja objasnjava starija djetetova dob, a tjelesno nasilje jos i muski spol. Od strukturalnih obiteljskih varijabli na individualnoj razini razlike u nasilnickim ponasanjima predviđa necjelovitost obitelji, a na međurazrednoj razini razlika u obiteljskim prihodima ucenika. Od procesnih obiteljskih varijabli na individualnoj razini vrsnjacko nasilje predviđaju ucestalije kažnjavanje djeteta i slabiji roditeljski nadzor. Tjelesno nasilnicko ponasanje na individualnoj razini predviđaju jos i veca roditeljska popustljivost te veca roditeljska ukljucenost u obrazovanje djeteta. Verbalno se nasilnicko ponasanje na međurazrednoj razini može predvidjeti na temelju nejednakosti obiteljskih prihoda unutar razreda, dok znacajnih prediktora za tjelesno nasilnicko ponasanje na međurazrednoj razini nema. Socio-ecological models, in addition to individual factors, emphasize the importance of contextual factors for the occurrence of peer violence, notably family and school. The aim of the study is to verify the predictiveness of some structure and process family variables and variables from the mesosystems of families and schools in order to explain physical and verbal peer violence. The study included 880 students from fifth to eighth grades, their parents and their homeroom teachers. Data on peer violence was collected by using self-ratings, self-nominations and peer nominations from students’ classes. Information about parenting practices was collected from the children, while the parents provided data on the socioeconomic status of their family, knowledge of the child’s peers and involvement in the child’s education. Homeroom teachers reported about the frequency of contact between parents and the school. Data were analyzed using multivariate multilevel modeling. Demographic variables that explain both types of peer violence were the child’s older age and, for physical violence, the male gender. At the individual level, structural family variables which have an effect on the differences in violent behavior refer to the incompleteness of the family and, on the interclass level, the differences in family income of students. Concerning the process family variables on the individual level, peer violence can be connected to more frequent parental punishment and lenient parental supervision. Physical peer violence at the individual level is also connected to greater parental permissiveness and greater parental involvement in the child’s education. Verbal peer violence at the interclass level can be linked to family income inequality within the class, while physical peer violence at the interclass level did not have significant predictors.