Discussion

Discussion
Limitations and future research

Discussion

Not all children who witness violence in the home develop aggressive behaviour problems themselves. Although children exposed to violence in the home are more likely to act out aggressively than are other children, the majority (approximately 68%) of children who witnessed violence in the home do not act out with aggressive behaviour. There are, however, important factors that can increase or reduce the negative impact of this experience.

This study suggests that parenting practices can reduce or intensify conduct problems among children. Children are less likely to act out aggressively when their parents use more effective parenting techniques than those who rely on hostile techniques such as reacting to their child’s behaviour with anger, and using negative rather than positive reinforcement. As Lytton (1990) suggests, the relationship between child behavioural problems and parenting is likely a circular problem. Children themselves differentially trigger parenting styles as a result of their behaviour, which can in turn exacerbate the antisocial behaviour that it is intended to address.

Consistent with earlier research (see Sprott and Doob, 1998), this study suggests that children who are generally happy, with lower levels of depression and anxiety are less likely to have aggressive conduct problems than children with higher emotional anxiety. Gaining the ability to regulate one’s emotions and resolve conflicts without resorting to aggression is an important step in the child development process. Parents have a key role in socializing and supporting this emotional development, and this is particularly important for children dealing with the confusing and often traumatizing emotional demands of living in high conflict households (Committee on Integrating the Science of Early Childhood Development, 2000). The NLSCY shows that child depression and anxiety is highly correlated with parental depression, which makes addressing child conduct problems even more challenging, as they are “often deprived of the parent as a resource for managing these powerful emotions” (Committee on Integrating the Science of Early Childhood Development, 2000:108).

For this reason social support outside the family unit is particularly important as an alternative resource for children. This study, however, did not find a significant relationship between low parental and community social support and child aggression, nor did it find that social support mediated the relationship between exposure to violence and aggressive behaviour. The measures of social support used in this analysis largely focused on the support available to parents through religious attendance, community ties and the availability of friends and family to turn to for support. It may be, however, that the social resources directly available to children are more important than what they gain indirectly through their parents. As Hetherington (1992) suggests, children exposed to marital conflict are better able to cope if they have close peer relationships and positive and distracting activities outside the home that enable them to distance and disengage themselves from the conflict. Direct measures of peer support, academic and extracurricular engagement will be available in future cycles of the NLSCY as the children in this sample enter adolescence but are not available for these younger groups of children.

Consistent with previous research on gender differences in aggression, boys in the study had higher odds of engaging in aggressive behaviour than did girls. This observation is not surprising given that girls experience more pressure to refrain from aggressive behaviour than do boys and awareness of this disapproval starts at a young age (Crockenberg and Langrock, 2001). Given this strong disapproval of female aggression, girls learn to suppress their anger making them more susceptible to internalizing symptoms – such as indirect forms of aggression (see Dauvergne and Johnson, 2001).
Although boys in the study were more likely to display aggressive behaviour, this study did not find that exposure to violence in the home differentially impacted boys’ aggressive behaviour over that of girls showing some similarity in experience.

Child development research shows us that the ability to self-regulate, or manage stress through cognition rather than behaviour, improves with age as children expand their range of coping skills (Kerig, 2001). Among children in this study high aggressive behaviour declined with age. This is true of both children exposed to violence in the home and those who were not. Ideally one should explore the impact of first exposure to violence on children at different stages of their emotional and psychological development. Previous research has identified some unique symptoms associated with exposure of violence at different developmental stages, ranging from the neural development of infants to emotional and behavioural problems among school age children and adolescents (Perry, 1997; Jaffe et al., 1990; Sternberg et al., 1993). Unfortunately, the timing of first exposure to violence was not captured in the NLSCY, and parent respondents reported on the frequency of witnessing violence in general making it difficult to isolate these occurrences at any specific point in time.

Limitations and future research

As in any study, readers should be aware of possible limitations with the data source or measures used in the analysis. The NLSCY was designed to measure a variety of family, peer, school and community influences on children and it was not intended to directly estimate the number of children exposed to domestic violence. Parent respondents were asked to estimate “how often their children see adults or teenagers in the home physically fighting, hitting or otherwise trying to hurt others”. This is an indicator of exposure to violence that may involve persons other than parents, such as older siblings or other adults in the home.

Further, responses to this survey question were given by parent respondents, who may under-report the incidence or frequency that their children witnessed violence in the home. Previous studies of family violence have shown that parents often falsely report that their children were unaware of parental violence when the children report awareness of it (O’Brien et al., 1994; Jaffe et al., 1990). Lastly, the NLSCY does not directly measure child abuse and we do not know if the children who reportedly witnessed violence in the home were also targets of violence. Some studies have found that witnessing violence in the home and suffering abuse is doubly disadvantageous for children. Children who were both abused and witnessed parental violence were found to exhibit the most problem behaviours, followed by children who only witnessed the violent event(s) (McCloskey et al., 1995; Hughes et al., 1989).

This study also raises a number of important questions to be addressed in future research.
The current analysis relied solely on parent respondent assessments of their child’s use of physical aggression. For children in this age group it is also possible to explore child conduct disorders from the perspective of teacher respondents. Previous analyses have shown that children identified as being “highly” aggressive can differ considerably depending on who is being asked (parent, child or teacher), and we should not necessarily expect behavioural consistency across situations (Sprott and Doob, 1998). For example, children may not display conduct problems at school but act out at home (or vice versa). The developmental pathway towards serious offending usually starts with persistent disruptive behaviour in the home, which then spreads to delinquent acts at school and in the community (Loeber and Farrington, 2000). Further analysis examining the impact of exposure to violence at home on behaviour in other social contexts such as the school environment would prove beneficial. Future work will build on this analysis to examine the likelihood that children who witness violence in the home will develop more serious emotional problems, and will engage in delinquent activities as they enter the age of criminal responsibility.