|
Abstract
|
VioLit summary:
OBJECTIVE:
The aim of this study by Salzinger et al. was to examine the family risk factors associated with child abuse. The effects of family violence on children's behavioral problems were investigated. The authors also examined whether different types of child behavioral disturbance developed as a function of different forms of family violence.
METHODOLOGY:
The authors conducted a primary analysis of quasi- experimental, cross-sectional data obtained from 106 physically abused children aged between 8 and 12 years old. Study participants were confirmed abuse cases registered in the New York State Child Abuse Register for New York City, from 1986 to 1989. A comparison group of 85 non-abused school children were matched with the abused children based on same gender, age within one year, and similar race and ethnicity. The authors claimed that the total sample comprised two, almost demographically identical, groups. Children's mean ages were 10.2 for the abused group and 10.3 for the comparison group. Of the abused group, 5% were White, 56% Black, 38% Hispanic and 1% Asian. The comparison group comprised 9% White, 42% Black, and 49% Hispanic. Children's mother's mean ages were 35.5 and 36.5 respectively. Maternal education mean was 11 years for both groups. Mean number of adults living in the children's homes was 1.8 for both groups, and mean number of children was 2.9 and 2.8 respectively. The authors reported that both groups comprised mostly low socioeconomic status families; 49% of the abused group and 42% of the comparison group were on welfare. 71% of the total sample was male. The authors stated that they found no significant interactions between gender and the association between abuse and major outcome factors.
The authors reported that family violence was determined based on information from an audiotaped, semistructured home interview of the child's caretaker, and for the abused group, a written account (taken from the Child Abuse Register) of the circumstances leading to the family being indicated for physical abuse. Selected risk factors for abuse included: (1) drug and alcohol abuse in the mother's family of origin; (2) severe beating and/or sexual abuse of the mother in her family of origin; (3) substance abuse in the child's household; (4) severe discord among the adults in the child's household; and (5) physical violence between the adults in the child's household, between the adults and the target child, between the adults and other children in the household, and between the children in the household. The authors reported that interrater reliability averaged 88% for the coding of information from narrative descriptions (i.e. agency accounts plus the Family Interaction component of the interview and information obtained from No.5 above). The Child Behavior Checklist (Achenbach & Edelbrock, 1983) was completed by the targeted child's caretaker as a measure of disturbed behavior. The Teacher Report Form (Achenbach & Edelbrock, 1986) was completed by the child's homeroom teacher. The authors stated that the instruments generated 4 parent rating measures--Total Behavior Problems, Internalizing Behavior Problems, Externalizing Behavior Problems, and Social Competence, and; 4 teacher rating measures--Total Behavior Problems, Internalizing Behavior Problems, Externalizing Behavior Problems, and Adaptive Functioning. The authors reported that because almost every maltreated child in the study attended different schools, children's behavior data were collected from teachers attending 100 schools in 23 of New York City's 32 school districts across the 4 urban boroughs of New York City. Path and multiple regression analyses were conducted to examine the data.
FINDINGS/DISCUSSION:
In order to determine the path by which earlier family violence makes its effects felt later as abuse directed towards children, the authors chose a set of suggested child abuse risk variables and entered them into a Path model. The following set of possible risk factors were listed as they would occur chronologically: excessive use of alcohol and/or drugs by the child's mother's parents; severe beating and/or sexual abuse of the mother in her family of origin; excessive alcohol and/or drug use and severe discord between adults in the child's household; and partner abuse in the child's household. The path analysis revealed that two indirect paths would most likely lead to child abuse. It was found that for both paths child abuse was most likely to take place in homes in which there was partner abuse which tends to happen in homes characterized by severe couple conflict and substance abuse. The authors reported that the paths then diverged but that both paths were almost equally as likely to occur. The less likely path revealed a direct association between the child belonging to a household in which there was conflict and substance abuse, and earlier substance abuse in the mother's family of origin. The other path revealed that the effects of substance abuse in the mother's family of origin on conflict and substance abuse in her family of procreation was mediated by whether she had been abused by her own parents.
The authors conducted multiple linear regression to test 4 types of family violence and their effects on children's adaptive functioning and generally disturbed behavior. Variables placed into the equation included: 1) presence of physical and substance abuse in the mother's family of origin, 2) presence of substance abuse and conflict among adults in the child's home, 3) the degree to which the child was severely physically victimized by adults, and 4) the degree to which abuse took place among any other house members. Parent and teacher Achenbach ratings of a child's total behavior problems were used as a measure of a child's general behavioral disturbance. Adaptive functioning and social competence were also rated by teachers and parents, respectively, using the Achenbach instrument. The authors found that a child's behavioral disturbance, as rated by the parents, was significantly related to all of the above four variables (R squared=.23, F(3,112)=8.58, p=.0001). The authors found that child abuse accounted for a part of the variance (sr squared=.12, F(1,111)= 15.19, p<.01); witnessing violence among others in the family accounted for some of the variance also (sr squared=.07, F(1,111)= 7.14, p<.01); after controlling for family stress and conflict in the child's household and mother's family of origin, there was a joint effect of being a victim and a witness of familial violence (sr squared=.17, F(1,111)=21.25, p<.01); and the combined effect of all types of family violence after controlling for child abuse victimization (sr squared=.12, F(1,111)=13.75, p<.01). The authors stated that according to teacher ratings children's disturbed behavior occurred as a function of all four variables (R squared=.09, F(3,98)=2.41, p=.05). Child abuse accounted for a part of the variance (sr squared=.05, F(1,97)=5.1, p<.05). There was a joint effect of being a witness and victim of familial violence (sr squared=.09, F(1,97)=9.57, p<.01). The authors contended that teachers tended to rate the children's behavior according to the immediate direct effects of violence victimization. Mothers were said to have taken into account the history and conditions of both their children's and their own abuse.
The authors reported that parent ratings of their child's social competence did not relate to any of the 4 family violence variables. The authors found an inverse relationship between children's teacher rated adaptive functioning and the 4 family violence variables (R squared=.09, F(3,96)=2.43, p=.05). By itself, child abuse accounted for 9% of the variance (sr squared=.09, F(1,95)=9.38, p<.01); there was a joint effect of child victimization and witnessing familial violence (sr squared=.08, F(1,95)=8.33, p<.01). There was a combined effect for all types of family violence when the effects of child victimization were controlled (sr squared=.08, F(1,95)=8.25, p<.01).
The authors tested the effects of family violence on internalizing and externalizing behavioral disturbance. The severity and pervasiveness of family violence were found to predict internalizing behavioral disturbance according to parents (R squared=.07, F(2,187)=6.77, p=.001), and externalizing behavioral problems as per parent ratings (R squared=.16, F(2,187)=17.94, p=.0000) and teachers ratings (R squared= .07, F(2,187)=7.47, p=.0008). The authors found that by controlling for pervasiveness, abuse severity predicted internalizing behavioral disturbance as rated by parents (R squared=.02, F(1,187)=5.04, p=.025) and externalizing behavioral disturbance according to both parents (R squared=.09, F(1,187)=20.56, p=.0000) and teachers (R squared=.06, F(1,187)=11.75, p=.0007). The authors argued that abuse severity and family violence pervasiveness were risk variables for both internalized and externalized behavioral disturbance in children. The authors reported that there was a stronger association between externalizing problem behavior and severity and pervasiveness of family violence.
The authors concluded that the most likely paths leading from family violence to child abuse were indirect, commencing with substance abuse and violence problems in the mother's family of origin, arbitrated by the levels of stress and conflict in the mothers current household, and resulting in mother to child victimization. Children's disturbance, according to parents and teachers was found to be an effect of violence in the child's immediate family. Finally, the authors stated that severity and extent of family violence were found to be associated with internalizing and externalizing behavior problems.
AUTHORS' RECOMMENDATIONS:
The authors suggested that researchers study the problem of child abuse using a longitudinal approach in order to track children's emerging problems. Conducting interviews with fathers and male partners was also suggested as a means for explaining the variance that previous researchers have not been able to account for.
(CSPV Abstract - Copyright © 1992-2007 by the Center for the Study and Prevention of Violence, Institute of Behavioral Science, Regents of the University of Colorado)
KW - Family Risk Factors
KW - Child Abuse Causes
KW - Child Abuse Risk Factors
KW - Child Problem Behavior
KW - Child Behavior
KW - Child Victim
KW - Victim Behavior
KW - Behavior Causes
KW - Behavior Risk Factors
KW - New York
KW - Child Abuse-Behavior Link
KW - Child Abuse-Problem Behavior Link
KW - Child Physical Abuse Victim
KW - Child Physical Abuse Effects
KW - Child Physical Abuse Causes
KW - Child Physical Abuse Risk Factors
KW - Spouse Abuse Effects
KW - Children of Battered Women
KW - Child Witness
KW - Witnessing Violence Effects
KW - Witnessing Spouse Abuse
KW - Adult Substance Use
KW - Adult Parent
KW - Adult Offender
KW - Adult Violence
KW - Parent Offender
KW - Parent Substance Use
KW - Offender Substance Use
KW - Substance Use Effects
KW - Alcohol Use Effects
KW - Drug Use Effects
KW - Domestic Violence Effects
KW - Domestic Violence Causes
KW - Domestic Violence Victim
KW - Domestic Violence Offender
KW - Child Abuse Victim
KW - Child Victim
KW - Child Abuse Offender
KW - Child Abuse Effects
KW - Child Abuse Causes
KW - Domestic Violence Risk Factors
KW - Psychological Victimization Effects
KW - Intergenerational Transmission of Violence
KW - Violence Against Women
KW - Partner Violence
Language: en |