Findings

Prevalence of witnessing violence in the home
Childhood aggression and witnessing violence in the home
Comparing aggressive behaviour among boys and girls, controlling for witnessing violence and other factors

Prevalence of witnessing violence in the home

In 1998/99, approximately 378,000 children between the ages of 6 and 11 in Canada had witnessed violence in the home at some point in their lives. This constitutes approximately 17% of the population in this age group. There was no statistical difference in the proportion of girls (17.7%) and boys (16.9%) who are reported to have witnessed violence in the home (Figure 1).

Childhood aggression and witnessing violence in the home

Table 1 shows the proportion of children in the sample who display a high level of aggressive behaviour for each of the risk factors including witnessing violence in the home before controlling for other factors in the child’s life.

Exposure to violence in the home has a strong association with aggressive behaviour among children. Approximately 32% of children who witnessed violence at home are reported to have high aggressive behaviour compared with 16% of other children in the sample. Overall, boys showed higher levels of aggressive behaviour than did girls, and this is true for those exposed to violence and those who were not (Figure 2).

According to previous research, aggressive behavioural problems are more common among younger children, and for most children these problems decline with age (Kerig, 2001). In this study, the average age of children who displayed aggressive behaviour (8.3) is slightly lower than for children who acted out with physical aggression less often (8.5).

Top of Page

Only one of the three community and parental social support factors were found to have an association with child aggression. Children whose families attend religious services infrequently had higher aggressive behaviour than those who attend regularly. Approximately 22% of children who rarely attend religious services displayed high aggressive behaviour compared with 17% with more frequent attendance. The neighbourhood cohesion and parental social support variables are both scales with a range of 0 representing low neighbourhood cohesion and a lack of social support to a high of 15 for a highly connected neighbourhood and 20 for very strong parental social support. The mean score on both of these scales were not significantly different for children who have an aggressive behaviour problem than those who do not.

Hostile parenting practices are related to aggressive child behaviour among children in the sample. The mean score on the ineffective and hostile parenting scale is 17.1 among children who often display aggressive behaviour compared with 14.9 for children who display aggressive conduct problems less often.

Children with high emotional anxiety were also more likely to display aggressive behaviour problems than other children. Approximately 38% with high emotional anxiety had an aggressive behaviour problem compared with 14% with lower emotional anxiety.

Top of Page

Comparing aggressive behaviour among boys and girls, controlling for witnessing violence and other factors

What is an odds ratio?

Table 1 shows that many factors may be associated with childhood aggression, but because of the interconnections among them, we must examine them together in a procedure that “controls” for the effects of all these factors simultaneously. Two variables that were included in the previous analysis – neighbourhood cohesion and parental social support – were not included in the regression model because neither was shown to significantly increase risk of aggressive conduct problems among children in the bivariate analysis. Table 2 presents the multiple logistic regression results, and shows which factors remain important correlates of child aggression after taking into account the impact of other factors.

What is an odds ratio?

An odds ratio is a statistic generated by a logistic regression and can be used to assess whether, other things being equal, children with specific characteristics are more or less likely to display aggressive behaviour than those in another group, referred to as the reference category. For example, consider the risk of aggression for children who witness violence at home compared to those who have never witnessed violence in the home (the reference category). An odds ratio near 1.0 implies there is no difference in aggression between the two groups; an odds ratio greater than 1.0 implies those in the group being considered (children who witness violence in the home) are more likely to be aggressive than those in the reference group (children who did not witness violence) and an odds ratio less than 1.0 implies those in the group being considered are less likely to be aggressive than those in the reference category.

When an explanatory variable is continuous (e.g. age measured in years), the odds ratio tells how many times the ratio P/(1-P) is greater or smaller for an individual who is one year older than another individual. For example, an odds ratio of 2.0 indicates that the odds of displaying high aggressive behaviour are twice as high for an 11 year old as they are for a 10 year old.

The odds of displaying aggressive behaviour continues to be higher for children who witness violence in the home, even after controlling for socio-demographic, social support, parenting and child emotional problems. Children who witness violence in the home have more than double the odds of acting out aggressively (2.2) than do children who never witness violence. The interaction between exposure to violence in the home and the sex and age of the child was also examined, but did not reveal a differential effect of witnessing violence on aggressive behaviour for boys over girls, or younger children over that of older children (results not presented in table form).

Among the socio-demographic variables tested in this model, three of the four were found to increase the odds of child aggressive conduct problems, being younger, male, and having a low income. The odds of displaying high aggression decrease by a factor of .93 for every year increase in the age of the child. Although this effect appears to be quite small, this is due in part to the small unit of measurement (1 year). If the effect of age were to be analysed over a three-year interval, the odds of aggressive conduct problems among children decrease by a factor of .80.

The odds of high aggression are 1.9 times higher for boys than girls in the sample, and 1.8 times higher for children in low income families. After adjusting for the effects of a number of other important factors in a child’s life, those living in single parent families were no more likely to display high aggressive behaviour than children living in two-parent families.

Similar to previous studies (Loeber and Farrington, 2000; Brannigan et al., 2001), hostile parenting was shown to increase the odds of aggressive behaviour for children. The odds of having an aggressive behaviour problem are considerably higher among children whose parents use hostile parenting techniques (such as getting angry or annoyed at the child, focusing on negative rather than positive child behaviours) as compared to those whose parents use these parenting styles less often. For every point increase on the hostile parenting scale, the odds of having high aggressive behaviour increased by a factor of 1.2.

When effects of other factors were controlled, the only measure of social support that was significant in bivariate analysis, religious attendance, ceased to be a significant risk factor for aggressive behaviour. However, child emotional problems continued to have a strong association with aggressive behaviour. The odds of having an aggressive conduct problem are more than two and a half times higher (2.6) for children who scored above the 80th percentile on the emotional anxiety scale.