Correctional Service Canada
Symbol of the Government of Canada

Common menu bar links

The Impact of Experiencing and Witnessing Family Violence during Childhood: Child and Adult Behavioural Outcomes

Warning This Web page has been archived on the Web.

Summary and Conclusions

The purpose of this report was to provide an update on the cycle of violence research published in recent years. Research conducted with children indicates that children who are both witnesses and victims of abuse have more severe aggressive behavioural outcomes than do those who are solely victims, who in turn have more difficulties than those who have witnessed violence but have never been victimized. Investigations conducted with adolescents and adults in the general population suggest that both witnessing and experiencing family violence in childhood are associated with violence in the adult family. Research with offenders indicates that a very high proportion of this population have a history of witnessing violence or being directly victimized. File review studies reveal that an abuse history is in turn related to family-directed and stranger directed violence that comes to the attention of authorities.

It is interesting to note that despite Widom's (1989b) call for prospective designs in research addressing the "violence begets violence" hypothesis, the majority of the studies reviewed in this report are retrospective in nature. Indeed, many of the other methodological shortcomings pointed out by Widom can be observed in articles published in the five years since her report became available to the academic community at large11. Based on the few prospective studies that do exist, numerous reviewers maintain that the majority of abused children do not continue the cycle of violence as adults. Even in childhood, the proportion of children who exhibit clinical levels of emotional/behavioural problems as a result of living with family violence is lower than one might expect: about one-third of boys and one-fifth of girls are in the clinical range (Cooper, 1992). As discussed earlier, estimates of the proportion of abused children who then go on abuse their own offspring are around 30%. Witnessing violence between parents is thought to be only modestly related to marital aggression in the second generation; about 16-17% of witnesses report aggression in their own intimate relationships (Widom, 1989b) 12. Nevertheless, a number of the studies described in this report indicate that the proportion of variance in violent behaviour accounted for by abuse and/or witness status is significant, thus further research on protective factors and potentially helpful treatment interventions is required.

11Given the lag time between data collection and publication of results, it is probably the case that many studies were designed and completed in the years prior to Widom's landmark review yet published in the years that followed. Perhaps an increase in the number of studies that incorporate Widom's design recommendations will be seen in the near future.

12It is difficult to generate corresponding figures for the relationship between witnessing violence and later perpetration of child abuse and between experiencing abuse and later partner abuse since none of the prospective studies directly addressed these questions.

Although few studies specifically address the issue of intergenerational transmission of violence among offenders, the available research suggests that the overlap between childhood victimization and later violent behaviour in this population is actually quite substantial (Dutton & Hart, 1992a, 1992b; Robinson & Taylor, 1994). As explained earlier, numerous characteristics of incarcerated populations are mentioned in the literature discussing the attributes of batterers (Dutton & Hart, 1992b). For example, Hotaling and Sugarman (1986) discovered in their extensive review of the risk markers for husband to wife violence that witnessing violence during childhood or adolescence was a consistent risk marker for battering, as was alcohol use, low assertiveness, low income and educational level and partner- directed sexual aggression. A criminal arrest record was also (inconsistently) associated with marital aggression. This correspondence between criminal and family abuser profiles reinforces the importance of making treatment programs available to offenders. The Correctional Service of Canada has already taken steps to provide programming for offenders who have a history of wife assault and/or parenting problems. Given the findings of recent file review studies, it is important that the programs offered directly address the concept of the cycle of violence.

The next logical steps for future research on the cycle of violence include studies exploring the effect of amount of violence witnessed (Jaffe, Wolfe & Wilson, 1990; Hughes, 1988), more prospective studies following victimized children into adulthood (Widom, 1989b), and studies of the factors that may prevent individuals from repeating the abuse that they have experienced and/or witnessed (e.g. "protective factors;" Garmezy, 1981).