Multi-Level Effects on Behaviour Outcomes in Canadian Children - May 2001

1.4 Discussion

Previous

Contents

Next

1.4.1 Individual and Family Influences

Our findings suggested that children's behaviour outcomes were influenced the most by their age and sex and by characteristics within their families. Mothers reported greater internalizing (i.e., anxiety-emotional) problems for older children and greater externalizing (i.e., physically aggressive, hyperactive-inattentive) problems for boys than girls. As well, children from low SES, single-parent families whose mothers were younger, had greater depressed mood, and used more hostile and punitive parenting practices were reported to exhibit more problematic behaviour. In contrast, mothers reported less hyperactivity-inattention and more prosocial behaviour for older children as well as more prosocial behaviour for girls than boys. Children from families in which there was less dysfunction, greater social support, greater maternal depressed mood, positive mother-child interactions, and non-punitive consistent parenting were reported to be more prosocial. Positive mother-child interactions and consistent parenting were also related to fewer anxiety-emotional and hyperactive-inattentive problems.

While most of the findings were consistent with existing knowledge, one surprising finding was that maternal depressed mood was associated with more prosocial behaviour. It may be that these children learn prosocial behaviours, such as helpfulness and cooperation, to help mothers whose depressed mood may limit their everyday functioning. It may also be that these children learn to behave prosocially in order to develop relationships with other individuals who can meet some of their dependency needs, which mothers may not be able to successfully accomplish as a result of their depressed mood. A third possibility is that the association between parenting behaviours (i.e., positive interaction, consistent behaviour, less punitive parenting) and greater prosocial behaviour mediated some of the risks posed by having a mother with depressed mood. Non-substantive reasons may also account for the positive association between maternal depressed mood and prosocial behaviour. For example, there is the possibility that depressed parents may report more behaviour problems in their children. Serbin et al. (1998) advised caution when interpreting the meaning of parent-reported childhood problems under such circumstances, noting that findings may reflect a combination of the child's actual behaviour problems as well the parent's distress level. Other studies (Sawyer, Streiner, & Baghurst, 1998) have found the effect of parental distress on their reports of children's behaviour problems to be minimal and clinically insignificant.

Another surprising finding was the overall lack of significant association between family SES and childhood behaviours. It may be that the strong and consistent effects of maternal depressed mood and parenting behaviours attenuated the SES effects. This hypothesis is based on past findings showing that the relationship between family SES and youth behaviour outcomes is mediated by parental psychological well-being and parenting behaviours (Dodge et al., 1994; McLeod & Shanahan, 1993; Simons, Johnson, Beaman, Conger, & Whitbeck, 1996). As well, the hierarchical three-level models used in the study made it possible for the individual, family, and neighbourhood variables to influence not only the childhood outcome measures but also one another.

1.4.2 Neighbourhood Influences

The behavioural outcomes of children in our study were only minimally influenced by neighbourhood variables. This finding may be partly due to the fact that the children were young (2-11 years) and had limited contact with their environment, outside of the family. It may be that neighbourhood effects become more noticeable with age, as has been found in some studies of neighbourhood influences among adolescents (Aneshensel & Sucoff, 1996a; Seidman et al., 1998; Simons et al., 1996; Stiffman, Hadley-Ives, Elze, Johnson, & Dore, 1999). However, since behaviour problems have been found to decrease with age (Broidy et al., 1999; Nagin & Tremblay, 1999), neighbourhood effects on problem behaviours are likely to be complex. Our study did find that living in a rural area was associated with fewer anxiety-emotional problems and that neighbourhood poverty and problems were associated with physical aggression. Surprisingly, greater neighbourhood poverty was associated with less childhood physical aggression after having controlled for other neighbourhood and family variables. It may be that in poor neighbourhoods, parents' focus on economic concerns detracts from their ability to closely supervise and observe their children. As such, parental reports of childhood behaviour problems may be prone to greater inaccuracies as the level of neighbourhood poverty increases. Another possibility is that physical aggression may not be perceived as such in poor neighbourhoods but rather may be viewed as an adaptive response to adverse neighbourhood conditions (Kupersmidt, Griesler, DeRosier, Patterson, & Davis, 1995). Similarly, parents living in poor neighbourhoods may have higher thresholds for identifying physical aggression problems in their children. Finally, Foster, Hagan, Boulerice, and Tremblay (2001) used the same NLSCY data set but a clustering variable approach to show that very high SES neighbourhoods (i.e., low poverty) were associated with less childhood physical aggression compared with middle-class neighbourhoods. This finding, along with that of the present study, suggests that the effects of neighbourhood poverty on childhood aggression may be curvilinear.

1.4.3 Policy and Research Considerations

The scope of investigation into variables that influence children's behaviours has recently broadened from the traditional focus on a child's family to characteristics within the child's neighbourhood. Findings from our study indicated that neighbourhoods had an impact on child outcomes in that approximately 4-9% of the variation in problem and prosocial behaviours was associated with neighbourhood variables. However, neighbourhood effects were minimal relative to those associated with a child's family composition (e.g., single-parent household), family functioning (e.g., parenting practices), and parental psychological health (e.g., maternal depressed mood). These results are similar to studies that have compared school effects to individual and family effects (e.g., Willms, 1999).

Our findings would suggest that policy efforts aimed at decreasing problematic and promoting prosocial behaviours among 2-11 year old children should place greater emphasis on families, in particular parental difficulties related to financial stress and psychological well-being and to ways of disciplining and interacting with their children. While our study found a number of associations between family variables and children's behaviours, it seems important to remember that these variables are most likely linked with one another, with other family variables, and with child variables (e.g., age, sex) in a number of significant and sometimes complex ways. As well, Boyle and Lipman (1998) noted that some family characteristics (e.g., SES) are difficult to separate from those at the neighbourhood level (e.g., poverty, unemployment).

While our study found a limited neighbourhood impact on children's behaviours, research is still in the early stages and as such, there are several future research considerations. We defined neighbourhoods using data at the census tract level. Compared with previous research that examined smaller catchment areas (enumeration areas; Boyle & Lipman, 1998), our purpose in using a larger more heterogeneous neighbourhood definition was to reduce measurement error so that we could more clearly differentiate family and neighbourhood effects. However, this adjustment still resulted in only a few significant neighbourhood effects. Brooks-Gunn, Duncan, Klebanov, and Sealand (1993), who also defined neighbourhoods using the census tract, wondered whether this spatial unit was too large, too small, or too incongruent with the relevant neighbourhood environment of their sample. They noted that a number of neighbourhood size variables may also influence behaviour, such as the density of people residing in an area and the availability of public transportation.

There clearly are conceptual challenges associated with examining neighbourhood effects. First, there is no single operational definition of neighbourhood, with definitions typically being guided by the particular issue under investigation (Gephart, 1997; Jarrett, 1997). Second, there are various ways in which neighbourhood effects may be revealed. Boyle and Willms (1999) noted that neighbourhood differences may be the result of real contextual effects or artificial compositional effects, namely differences in inhabitant characteristics. As well, there may be interactions between individuals and their neighbourhoods so that each is influenced by the other. Our study, like many previous ones, used administrative boundaries to spatially define neighbourhoods. It has been suggested (Boyle & Lipman, 1998; Boyle & Willms, 1999) that this approach results in a great deal of heterogeneity which, in turn, makes it difficult to detect neighbourhood effects. Boyle and Lipman (1998) stated that the "next generation of studies into neighbourhood effects would be well served by linking census data to spatial boundaries that minimize within place heterogeneity of the hypothesized risk factors for study, such as neighbourhood disadvantage" (p. 31).

Policy makers and service providers who would want to use results from this study to decide on behaviour problem prevention policies and services should consider the fact that individual and family characteristics are most strongly related to children's behaviour problems. Children most at risk of behaviour problems are young males living in a dysfunctional family with young depressed mothers who do not live with the father. Thus, to substantially reduce the level of behaviour problems in our society, one is tempted to target these characteristics. However, correlates are not causes. We need to understand the mechanisms that are involved. Spending our resources on correlates (which are not causal factors) would be a waste of money. In most cases, the causes of behaviour problems are complex and appear to build up over long periods of time. Longitudinal and experimental data are needed to understand these mechanisms. With time, the NLSCY data collection cycles will allow us to more rigorously test mechanisms. Multiple data points are particularly important for modelling more complex relationships such as mediated and interactive effects. Longitudinal data will also further enable us to more properly conceptualize childhood behaviour adjustment as a developmental process that changes over time.

Because there is good evidence of intergenerational transmission of behaviour problems from the present study, and from many other longitudinal studies, the best advice to policy makers and service providers for the prevention of behaviour problems is for them to take a long-term perspective. To prevent behaviour problems, one probably needs to make long-term investments in early child development through support to adolescents and young adults who are and will be the next generation of parents of young children. From this perspective, although males are those with the highest levels of problems, females with problems (apparently less serious) could be a better investment in the long run, since they are the ones most involved in providing the early environment (pre- and post-natal) which appears to be of crucial importance for the development of a brain which will be in control of an individual's behaviour.

1.4.4 Limitations

Our study's use of a cross-sectional design did not allow for causal conclusions to be reached. One of the goals of the NLSCY, from which our data were derived, is to collect information on children every two years until adulthood. As such, we will be able to address important issues related to predictors of children's behaviour outcomes in future reports. Another limitation in the present study was the sole use of maternal reports, which are subject to biases, to assess children's problem and prosocial behaviours. The NLSCY also intends to collect information from children's teachers so we will be able to examine the issue of informant differences and bias in future reports. Third, our study used static conceptualizations of such variables as family composition and poverty. Past research (McLeod & Shanahan, 1993) has demonstrated that variable histories, such as poverty duration, may have important links with children's mental health functioning. Finally, we did not examine possible interactions associations between and among individual, family, and neighbourhood variables that may impact on childhood behaviour outcomes.

Previous

Contents

Next