Understanding the Early Years - Community Impacts on Child Development - August 1999

Previous

Contents

Next

2. Annotated Bibliography

2.1 Overview Based on the Literature

Though it is often assumed that there is a relationship between community characteristics and child development, research in the area is sparse. Of the past research that has been undertaken, most has been relatively small in scale, American-based and focused mainly on adolescents and older children. In addition, because few studies have looked beyond the impact of the socioeconomic climate within a community to the more social characteristics (e.g., community involvement, safety) that could potentially affect child development, the relevance of past studies to the examination of community impacts on preschool-age children is somewhat limited.

This work, rather than being an annotated bibliography in the traditional sense (i.e., in-depth review of all available literature), represents a more focused review in which articles were selected to guide the framework for research on community impacts. The focus is on measurement, and empirical research, to determine the most successful ways to measure the impacts of the community on child development for longitudinal surveys such as the NLSCY. Only the most recent studies examining the characteristics of communities, their residents, and their available facilities were examined.

The first section outlines the main theories, models and definitions used to conceptualize children's communities. Most often the first obstacle encountered when researching community impacts is due to the controversy over definitions of communities and neighbourhoods (two terms which are often used interchangeably). Although in the past communities have often been defined geographically (based on Census Enumeration Areas or zip codes) recent research has suggested that in order to more accurately represent a community as it is perceived by its residents, more sociological definitions (e.g., characterized according to individuals sense of shared space) should be employed.

Theories explaining the mechanisms by which communities affect child development are useful when framing data and research questions. There are several guiding theories linking community and child development, however theories of Social Contagion and Collective Socialization are particularly prominent. The Theory of Contagion hypothesizes the spread of deviant behaviour through imitation, modelling, and social learning from children's peers within the neighbourhood. At issue is whether such peer effects are equally strong for positive and negative outcomes. The Theory of Collective Socialization has a similar orientation except that it hypothesizes about the impact of adult role models and informally shared parenting functions by community residents. Both models therefore focus on the child's interactions with other members of the community and argue for a socially based definition of the neighbourhood.

Additional theories, which though less often employed are nonetheless important to consider when examining community influences, include theories of Relative Deprivation, Competition for Scare Resources, and Neighbourhood Resource Use. The theory of Neighbourhood Resource Use is particularly useful for work related to the NLSCY as it focuses on the manner in which the availability and accessibility of community resources can influence the use of programs and services by residents and, as a result, their child's healthy development. Each theory is described in greater detail in the body of the bibliography. The relative merits of each theory are not clear, since there are few empirical studies to test them against the reality of child outcomes.

The studies in Section 2 investigate community influences by examining socio-economic and family structure variables. Most often such investigations involve secondary analysis of existing data (quite often the census). Such data sources are often ill-suited to answer relevant questions because they do not provide sufficient detail on key variables hypothesized to affect child development, resulting in the use of proxy variables in the research. Variables that have demonstrated significance include for example, parent's labour force participation and marital status, the socio-economic characteristics of the family and larger community, and the concentration of people and the ethnic or cultural diversity in the area. Although these variables yield little specific information about how the community, through its resources and structure, can influence child outcomes, they do contribute important information about the influence of the child's wider socio-demographic environment.

The overall objective of this study is to determine the effects of community factors, over and above individual and family characteristics (including socio-economic ones) on child development. Census variables are clearly not comprehensive enough, as they do not provide sufficient data on relevant social characteristics of neighbourhoods (such as cohesion, safety, and resource use) that have been theorized to influence child behaviours. In Section 3, the results of empirical studies linking community factors, particularly social factors, to children's outcomes were examined. Factors that have been shown to increase positive outcomes include greater safety and cohesion, increased participation in community activities and higher levels of collective efficacy (social cohesion and a willingness to intervene for the common good) within communities. It became evident that the complex relationships between community variables and child outcomes, as suggested by the theories, had not often been empirically studied due to data deficiencies.

A variety of child outcomes have been examined in research studies investigating community impacts. Measures pertaining to young children include those that relate to cognitive and behavioural functioning, motor social development, and community participation, which has been used both as a dependent and an independent variable. Outcomes relating to adolescents and older children have been more numerous as the bulk of research has studied children in these age ranges, possibly because effects are stronger for older children as a result of increased interactions with their communities. Outcomes that have been studied include drug involvement, violent crime and other forms of delinquency, child maltreatment rates, sexual activity and education measures (including both levels of attainment and drop-out rates).

In Section 4, recent surveys with varying degrees of community content were examined. Some of the studies are in progress. Some, such as the Boston or Chicago instruments focused almost entirely on community relevant factors whereas others (Survey of Volunteering) contain only one or two questions of interest. A short description of categories and concepts measured on each survey was noted.

Based on the foregoing work, in chapter 4 of this paper a framework for research and data collection was developed. Since it was clear that the NLSCY alone would not be an appropriate tool for data collection, the data strategy for the pilot project in North York includes a Community Mapping Study. Decisions on data were made based on policy relevant research questions that were to be answered through the research. Definitions, variables to be measured, and the framework for analysis are presented.

In Chapter 5, the data strategy is laid out, showing how data collection would be spread out over the NLSCY and the Community Mapping Study. The final chapter provides an outline for the Community Mapping Study. The study is designed to provide results as stand alone research but also to be incorporated with the NLSCY.

The key subject areas for data collection for the NLSCY were community involvement, cohesion, safety/crime, resource use, and socio-economic characteristics. In Appendix A, the potential variables in the subject areas, the questions/instruments used to collect the data and their sources (previous use in a study) are listed. This inventory permitted comparisons, tests, and evaluations before decisions were made for the NLSCY.

Appendix B shows the instrument for the Community Program Survey. This instrument will provide information on community resources, whether they were used by the NLSCY sample or not. Comparisons of results among communities will show the areas of community investments that are most beneficial for families with children. Appendix C outlines the items used to assess the physical characteristics of the communities in the neighbourhood observation component.

2.2 Theories, Models, and Definitions

2.2.1 Definitions

Controversy over defining neighbourhoods for measurement

There is a great deal of controversy in the literature surrounding the best way to define and conceptualize neighbourhoods and communities for measurement. Most studies have employed a geographical definition, dividing communities according to municipal boundaries, or census tracts. Postal (zip) codes have also been employed because they provide a better indication of local characteristics than city, county, or provincial/state measures. However, concern has been raised that a neighbourhood as defined in this manner may still differ from the conceptualization of the neighbourhood that is held by the community's residents. The same difficulties arise when defining boundaries according to census tracts as they may be too large or even too small (depending on the area) to accurately reflect the environment with which the child interacts on a day to day basis (Brooks-Gunn, Duncan, Klebanov, & Sealand, 1993; Kohen, Hertzman, & Brooks-Gunn, 1998). Census tracts can represent upwards of 2000 people, which is likely much larger than a child's perception of his/her neighbourhood (Coulton, Korbin, Su, & Chow, 1995).

An alternative view is to define the neighbourhood from a sociological standpoint in which people's perspectives of their communities are used to define its boundaries (e.g., Boston Dorchester Cares Project - Neighbourhood Interview; Program on Human Development in Chicago Neighbourhoods - Community Survey Questionnaire, 1994). This would examine an individual's sense of shared space and institutions with which they interact. Although this method may more accurately conceptualize the neighbourhood as its residents view it, it remains a very difficult concept to operationalize. Geographically defined boundaries are much easier to measure and, therefore, more often employed. Furthermore, because until recently, few community-based studies had been undertaken, researchers wanting to examine neighbourhood effects had to rely on census data (geographic boundaries) as it was all that existed.

2.2.2 Theoretical perspectives

Though research on community influences is relatively new, a wide variety of theoretical models currently exist in the literature. This discussion will focus on five of the most prominent and relevant to our research: Social Contagion, Collective Socialization, Neighbourhood Resources, Competition, and Relative Deprivation. These models are outlined in the following two articles:

1. Furstenberg, Frank F. Jr., and Hughes, Mary Elizabeth (1995). "The influence of Neighbourhoods on Children's Development: A Theoretical Perspective and a Research Agenda," in Indicators of Children's Well-Being, Volume III. Cross-Cutting Issues: Population, Family, and Neighbourhood: Social Development and Problem Behaviours. Paper prepared for the Conference on Indicators of Children's Well-Being, Rockville, MD. 1995. Institute for Research on Poverty Special Report, No. 60c.

2. Jencks, Christopher and Susan E. Mayer (1990). "The Social Consequences of Growing Up in a Poor Neighbourhood," pp. 111-186, in L.E. Lynn, Jr. and G.H. McGeary (Eds.) Inner City Poverty in the United States. Washington D.C.: National Academy Press.

Social Contagion, also known as the epidemic model focuses on the role of imitation, modelling, and social learning from children's neighbourhood peers in shaping their behaviours. Generally speaking, it implies that "like begets like." That is, if good behaviour (e.g., graduating from university) is modelled it will increase the likelihood that other children within the neighbourhood will value an education and strive to advance academically. Conversely, negative behaviour (high crime rates, poor school attendance) should increase rates of deviant behaviours. The model does account for individual differences, which also frame behaviours, but states that even after accounting for personal characteristics, overall, the tendency will be for children within neighbourhoods to conform.

The theory of Collective Socialization looks at the impact of adult role models and informally shared parenting functions by residents in the community and in this sense they are somewhat similar the Contagion theory. The main difference between the two models is that in the former the child is influenced by and influences his/her peers, and in the later they are influenced by other adults within the community. According to this theory, adults can serve two main functions. First, they can act as positive role models, which help children to objectively assess the results of certain choices and behaviours (e.g., education creates opportunities), and at the same time exert social controls on the children by monitoring behaviours, being aware of and dealing with potential trouble.

The institutional model, better known as the Neighbourhood Resource Theory investigates the links between the quality and quantity of services available to residents such as police, parks, recreation, and health and the development of the children in the community. It implies that increased availability of services will lead to enhanced opportunities for development, enrichment of experience and reduced chance of problems. Resources within a community can also include social relationships developed among community members and therefore this concept known as social capital falls under the broader domain of resources. Three forms of social capital are particularly prominent in their influence on community members: shared norms, reciprocal obligations, and opportunities for sharing information, the presence of all of which can contribute to development (Coleman, 1988). Sampson (1992) argues that social capital is a key determinant of social organisation within a community and an important way to connect the child with his/her community. Social organisation among neighbours can in turn facilitate the generation of further social capital.

The final two theories that of Competition for Scarce Resources and Relative Deprivation are similar in that they focus on the negative impacts that certain neighbourhood structures (particularly being surrounded by affluent educated neighbours) can have on children. Theories of competition look at the effects of winning or losing in a competition for scarce resources within the community, while under the tenets of relative deprivation individuals appraise the impact of their situation relative to others in their communities, and subsequently adjust their behaviours accordingly. In these cases being surrounded by highly affluent neighbours would lead both children and adults to appraise their situations as worse than their neighbours, which may translate into lower achievement both academically and otherwise. For instance, Jencks and Mayer use the example of children who are not excelling academically because of a lack of effort. If these children are then moved from a lower to a higher socio-economic environment, their effort will only further decrease relative to their peers, if they appraise themselves as being much worse off. In terms of Competition theory more problems are likely to arise when neighbours are more heterogeneous (i.e., the gap between the rich and poor is wider). There appears to be no consensus on the relative merit of these theories and there are few empirical studies to test them against the reality of child outcomes.

Additional discussions of theoretical positions can be found in:

1. Cook, T.D., Furstenberg, F.F. Jr., Kim, J.R., Teitler, J.O., Geitz, L.M., Eccles, J., Elder, G.H. Jr., and Sameroff, A. (1994). Neighborhood differences in resources for promoting the positive development of adolescents: The roles of financial, human, social, cultural and psychological capital. Manuscript in preparation.

2. Coleman, James S. (1988). "Social Capital in the Creation of Human Capital," American Journal of Sociology, 9: S95-S120.

3. Sampson, Robert J. (1992). "Family Management and Child Development: Insights from Social Disorganization Theory," in J. McCord (Ed.), Advances in Criminological Theory (Volume III). New Brunswick, NJ: Transaction Books.

4. Wilson, W. J. (1987). The Truly Disadvantaged: The innercity, the underclass, and public policy. Chicago: University of Chicago Press.

5. Wilson, W. J. (1991). "Public Policy Research and the Truly Disadvantaged," pp. 460-481, in C. Jencks and P.E. Peterson (Eds.) The Urban Underclass. Washington, D.C.: The Brookings Institution.

6. Garner, C.L. and Raudenbush, S.W. (1991). "Neighbourhood Effects on Educational Attainment: A Multilevel Analysis," Sociology of Education, 64: 251-262.

2.2.3 Additional models

1. Kupersmidt, J.B., Griesler, P.C., DeRosier, M.E., Patterson, C.J., and Davis, P.W. (1995). "Childhood Aggression and Peer Relations in the Context of Family and Neighbourhood Factors," Child Development, 66: 360-375.

These authors address three additional models stressing the type of dynamic relationship between the child and his/her environment as an important determinant for outcomes:

• The Protective Model - The protective model, as its title applies, examines ways that children living in risky environments may be protected from developing problems. A healthy neighbourhood can play a key role in this interaction. Children in high-risk families for instance, only stand to benefit by living in low risk opportunity and resource filled environments as the neighbourhood can work to buffer the family-related risk factors. The model predicts no effect on children who are not at-risk.
• The Potentiator Model - This model focuses on the potential impacts of healthy neighbourhoods on the development of low-risk children. In this case, the only children affected by living in a low risk neighbourhood would be low-risk children whose development could be enhanced by this positive experience, no effects would be seen for other children.
• The Person-Environment Fit Model - This model looks at the relationship between the characteristics of a neighbourhood and the traits of an individual who lives within that neighbourhood. The more similarities between the two, the lower the likelihood of problems.

2. Theory of concentrated poverty and social isolation (Wilson 1987, 1991) as discussed in Furstenburg and Hughes.

This theory, linking the context of child rearing to child development, asserts that child development is dependent on the socio-economic conditions of the environment in which the child is raised. Children raised in impoverished environments may be isolated from social networks and resources that foster healthy development. Persistent poverty in neighbourhoods, in turn is created from combinations of a variety of economic social and cultural factors. Downward trends in the economy, re-structuring of the labour force, increased competition for jobs requiring higher levels of education, and changes in family structure have all contributed to concentrated poverty in neighbourhoods and therefore to the increased risk of developmental problems.

2.3 Neighbourhood Socio-Economic Indicators (Census and Administrative Information)

1. Brewster, K.L., Billy, J.O.G., and Grady, W.R. (1993). "Social Context and Adolescent Behaviour: The Impact of Community on the Transition to Sexual Activity," Social Forces, 71: 713-740.

Keywords: sexual behaviour, adolescents, social disintegration, socio-economic status

Background: This article examined the role of community characteristics in influencing the sexual behaviours of adolescents. Sampled were 734 women (under 20) who took part in cycle 3 of the National Survey of Family Growth (NSFG-III) in the United States.

Dependent Variables:

• Contraceptive use
• Age at first non-marital intercourse

Independent Variables:

• Community variables: Socio-economic status (SES), female labour force participation; religiosity (proportion of religious adherents and religious conservatives); social disintegration (mobility, proportion unemployed, separated or divorced); racial and ethnic composition; service availability (family planning clinics, abortion providers); and the proportion of non-marital teen births.
• Family and individual predictors (e.g., education, living arrangements, religious affiliation).

Results: Characteristics of an adolescent's community can play an important role in determining his/her sexual behaviours. After accounting for individual factors, several of the community characteristics were predictive of increased risk for intercourse. A community's social disintegration was particularly important (with greater mobility in occupied housing units and the proportion of divorced or separated females increasing risk for sexual activity). The level of educational attainment in the immediate community also exercised an effect, the higher the education - the lower the risk). Higher socio-economic status of the neighbourhood, and higher proportions of foreign-born or Black residents served to decrease risk. A lower marital dissolution rate and a more active female labour force increased the probability that contraceptives will be employed.

2. Brooks-Gunn, J., Duncan, G.L., Klebanov, P.K., and Sealand, N. (1993). "Do Neighbourhoods Influence Child and Adolescent Development?" American Journal of Sociology, 99: 353-395.

Keywords: theories (Social Contagion, Collective Socialization), childhood, adolescence

Background: The association between neighbourhood socio-economic characteristics and developmental outcomes of children was expected to differ for two age points (early childhood and adolescence). In early childhood, the home environment was predicted to have the biggest effect, while for adolescents, neighbourhood factors (e.g., schools, peer groups, economic opportunities) were postulated to have more of an influence.

Methodology: This study looked at the association between selected neighbourhood characteristics (census data) and cognitive and behavioural functioning at 36 months of age. Data came from the Infant Health and Development Program (IHDP), which sampled 895 premature, low birth weight infants from among 8 medical sites in the United States. Additionally, 2200 women from the Panel Study of Income Dynamics (PSID) were used to investigate the two adolescent developmental outcomes.

Dependent Variables:

• Cognitive functioning (Stanford-Binet IQ at 36 months)
• Behavioural functioning (Child Behaviour Checklist for ages 2-3)
• High school drop-out rate, and teenage out-of-wedlock births

Independent Variables:

• Social isolation - Proxy variables were used to measure social isolation. Neighbourhoods where at least 40% of the people who were not elderly were poor and no more than 10% of families had incomes above $30,000.
• Neighbourhood characteristics - Percentage of 1) employed males who are in professional or management occupations; 2) lone female-headed families; 3) families receiving public assistance; 4) males unemployed during the past year; and percentage who are Black in the neighbourhood.
• Family variables - Family structure, economic resources (total income, mother's education, female-headed household, and mother's race).

Results: Overall, results indicated that after controlling for family resources, the neighbourhood factors most likely to affect child and adolescent healthy development were the presence of two-parent families and affluent neighbours of higher occupational prestige. Results appeared to be most consistent with the theory of Collective Socialization, which stressed the importance of the resources, role models, and informal monitoring provided by affluent neighbours. Some evidence was found for the Contagion theory.

3. Chase-Lansdale, P.L. and Gordon, R.A. (1996). "Economic Hardship and the Development of Five- and Six-Year Olds: Neighbourhood and Regional Perspectives," Child Development, 67: 3338-3367.

Keywords: neighbourhood resources, competition for scarce resources, problem behaviours, cognitive and reading abilities, childhood

Background: Using data from the U.S. National Longitudinal Survey of Youth (NLSY) this study examined the impact of neighbourhood characteristics on the incidence of problem behaviours, and the cognitive and reading abilities of 5 and 6 year old children. The authors investigated neighbourhood or community influences in terms of economic and social resources, hypothesizing that living among higher SES families will increase children's cognitive functioning because of widened opportunities within their neighbourhoods for education and development. Furthermore, the degree of crowding within a community may accentuate competition for scarce resources (e.g., when there is a limited number of kindergarten slots).

Dependent Variables:

• I.Q. (Peabody Picture Vocabulary Test), reading (Peabody Individual Achievement Test of Reading Recognition) and problem behaviour (Child Behaviour Checklist)

Independent Variables:

• Neighbourhood variables (SES, male joblessness, concentration of people, racial similarity and adult presence - for monitoring and supervision, defined as the ratio of adults 25-64 to children 0-17).
• Family variables (e.g., income, number of adults and children in household, mother's age at first birth).

Results: In certain regions of the United States (particularly those that have experienced declining economic growth - Northeast and Midwest) children in neighbourhoods with characteristics such as high SES and racial similarity showed higher levels of cognitive functioning. In areas where the presence of adults was higher, increased behavioural competence was also observed. This indicates that favourable neighbourhood characteristics can have a protective effect for families living in high-risk regions of the country and provides support for the Neighbourhood Resource Theory.

4. Chase-Lansdale, P.L., Gordon, R.A., Brooks-Gunn, J., and Klebanov, P.K. (1997). "Neighbourhood and family influences on the intellectual and behavioural competence of preschool and early school-age children," in J. Brooks-Gunn, G.J.Duncan, and J.L. Aber (Eds.)Neighbourhood Poverty: Context and Consequences for Children (Volume 1). NY: Russell Sage.

Keywords: theories (Collective Socialization, Neighbourhood Resource), affluence, cognitive and behavioural functioning

Background: This study examined the neighbourhood and family effects on the functioning of preschool (3-4 years) and early school (5-6 years) aged children in the United States. Cross-sectional data (i.e., sample of children who were 3 or 5 years in 1986) from the National Longitudinal Survey of Youth (NLSY), which began in 1986 and will follow 7000 children from infancy to late adolescence were used. A sample of children from the IHDP (an eight site study of an early educational intervention for premature and low-birth-weight children and their parents) was also employed to get a longitudinal look (same children sampled at age 3 and 5) at the development of these children.

Dependent Variables:

• Verbal abilities (PPVT-R) both age groups
• Behavioural Functioning - Child Behaviour Checklist (2-3 years), Revised Child Behaviour Profile (4-5 years)
• Cognitive Functioning - Stanford-Binet Intelligence Scale (age 3), Wechsler Preschool and Primary Scale of Intelligence (age 5)

Independent Variables:

• IHDP family variables - child's gender, race, mother's years of schooling, female-headed family, maternal age at birth, maternal employment (unemployed, part-or full-time, and family income to needs ratio).
• NLSY individual and family measures - same as IHDP measures, with addition of whether or not the child was enrolled in regular school during the survey week as a control variable.
• Neighbourhood factors - SES (low, high) male joblessness, family concentration, and ethnic diversity.

Results for Preschoolers: Preschool children showed few direct neighbourhood effects, with the only significant predictor of higher I.Q. scores being the presence of affluent neighbours (IHDP), supporting the authors' hypothesis that affluence would be the factor most likely to impact development in younger children. As such, it lends support to the Neighbourhood Resource theory, which postulates children from affluent neighbourhoods would fare better because of increased opportunities for enrichment and development.

Male joblessness had a paradoxical relationship, being associated with an increase in children's internalizing problems in the NLSY sample and a decrease in internalizing problems in the IHDP sample. Despite these neighbourhood effects, family factors explained most of the variance, leading the authors to conclude that "the family is the primary socializing unit for preschool children and that direct neighbourhood influences on such young children, as measured by the five factors, are small or filtered by family experience."

Results for Early School Age Children: Most of the effects on school age children were also accounted for by family variables. The neighbourhood variable affluence was strongly related to PPVT and reading recognition scores (NLSY) and verbal IQ (IHDP). Ethnic diversity had conflicting effects, and male joblessness was related to increased internalizing and externalizing problems in children. The effects of male joblessness could be explained by the theory of Collective Socialization, as it may result in lower levels of positive behavioural modelling and supervision. This study also discussed the problem of selection bias and the resulting difficulty of separating the effects of family from neighbourhood characteristics as an important limitation to research studying community impacts on development.

5. Crane, Jonathan (1991). "The epidemic theory of ghettos and neighbourhood effects on dropping out and teenage childbearing," American Journal of Sociology, 96(5): 1226-1259.

Keywords: Epidemic theory (Social Contagion Theory), neighbourhood quality, adolescent behaviours

Background: The relationship between neighbourhood effects and high school drop-out and teenage childbearing rates were analyzed to determine whether there was a sharp increase in these social problems in the worst neighbourhoods of large cities in the United States. Data were drawn from the Neighborhood Characteristic File of the PUMS (Public Use Microdata Samples) samples from the 1970 Census. It was only in 1970 that the Census Bureau defined a neighbourhood and this was the first time that neighbourhood data were available. Neighbourhoods were defined geographically and are about the same size as census tracts (averaging 4-5 thousand people). Only teenagers living with their parents were included in the analyses.

Dependent Variables:

• School drop-out and teen childbearing rates

Independent Variables:

• Neighbourhood Quality (percentage in the neighbourhood who hold professional or managerial jobs- % high status)
• Control variables such as SES, family structure, mobility, ethnicity

Results: Results lend support to the Epidemic (Social Contagion) theory, which predicted that a child's tendency for deviance will increase if he/she associates with deviant peers. Overall, findings indicated that incidence of both dropping out and teenage child bearing was increased for children raised in neighbourhoods of lowest quality (having the lowest proportion of neighbours in high status occupations). Effects were particularly strong for those in the lowest end of the Neighbourhood Quality range.

6. Garner, C.L. and Raudenbush, S.W. (1991). "Neighbourhood Effects on Educational Attainment: A Multilevel Analysis," Sociology of Education, 64: 251-262.

Keywords: school achievement, neighbourhood deprivation

Background: This study examined neighbourhood effects on the educational attainment of a group of 2,500 adolescents in Scotland. Each of the respondents had finished their last compulsory year of school between the years of 1984 and 1986. Data from the survey were linked to information collected in the 1981 Census and hierarchical linear regression was employed to determine the influence of individual ability, family background, schooling, and neighbourhood characteristics on the end-of-school attainment of these young people.

Dependent Variables:

• General attainment score upon secondary school completion (scale consisting of 14 items)accounting for both educational attainment and years of schooling.

Independent Variables:

• Individual characteristics: Two measures of prior attainment (verbal reasoning and reading ability obtained from tests conducted in all schools when children were aged 11 and 12).
• Family measures: Father's occupation (social class) and employment status, length of parental schooling, family size, single-parent family status.
• Schooling measures: School membership (to measure variation among schools).
• Neighbourhood measures: Enumeration district of neighbourhood (from Census), and a composite measure (12 variables) of deprivation (e.g., proportions of unemployed, youth unemployed, single-parent families, low-earning socioeconomic groups, overcrowding, and the percentage of permanently sick individuals.)

Results: Several characteristics including having an unemployed or lower occupational status father, being a member of a large, single-parent family, or having parents with lower levels of education had negative impacts on educational attainment. Prior school attainment had the largest impact on future educational achievements. However, after controlling for pupil ability, family background, and schooling, the authors found a significant association between neighbourhood deprivation and lower educational attainments. The full analytical model explained the majority of the variance, and the remaining unexplained variance was insignificant indicating that there were few additional neighbourhood effects that were not accounted for. Findings suggested that policies to alleviate educational disadvantage cannot be focused solely on schooling, but must form part of a broader initiative to tackle social deprivation in the larger society.

7. Kupersmidt, J.B., Griesler, P.C., DeRosier, M.E., Patterson, C.J., and Davis, P.W. (1995). "Childhood Aggression and Peer Relations in the Context of Family and Neighbourhood Factors," Child Development, 66: 360-375.

Keywords: community resources, peer rejection, neighbourhood companions, aggression

Background: This study examined the relationship of neighbourhood and family factors on childhood aggression and peer relations for 1271 elementary school children (all children in grades 2 to 5) in a southern city in the United States. Children were assigned to 1 of 8 family types based on income, ethnicity, and household composition. Census data were used to identify the 29 neighbourhoods (high or low SES) studied. Data on the students were obtained from school archives and teacher reports.

Dependent Variables:

• Aggressive behaviour: children were asked to nominate 3 peers who fight a lot.
• Peer Rejection: children were asked to nominate the three peers they liked the most and the least.
• Home and Neighbourhood Play Companions: children were asked to indicate the children they played with (from two rosters) in their neighbourhood, and in their homes or peers' homes. The number of reciprocated nominations received by each child was counted to provide an estimate of the number of companions (in the same grade) that each child had.

Independent Variables:

• Family and neighbourhood measures of ethnicity, family income (poverty), and household composition factors were used. Gender and developmental differences were also factors.

Results: Results indicated that neighbourhood context was an important indicator of childhood aggression and peer relations even after controlling for familial factors. Living in a low-income home increased children's risk of aggression, poor peer relations, and having fewer play companions. In general, results followed the same pattern for Black children and those of single parent families. Lower SES neighbourhoods were also associated with poorer outcomes, perhaps because children from lower SES environments had fewer positive role models. Because these children also had fewer playmates it emphasizes the important role of social class in facilitating peer relationships. High levels of stress associated with poverty can also be linked to antisocial behaviours, or it may be that aggression is a learned response from living in an unsafe environment.

Neighbourhood context, particularly neighbourhood income level often mediated the family effects. For instance, Black children from single parent, or low-income families who lived in middle SES environments appeared to be protected from aggression, and had more neighbourhood play mates. This indicated that communities with more resources may be able to better provide for children and buffer the risk family-related risk factors. These same children, however, were at increased risk of being rejected by their peers, providing support for the idea that people like to associate with others whom they see as similar to themselves.

2.4 Community Indicators

1. Boyle, Michael H. and Ellen L. Lipman (1998). Do Places Matter? A Multilevel Analysis of Geographic Variations in Child Behaviour in Canada. Applied Research Branch, Strategic Policy, Human Resources Development Canada. Working paper W-98-16E.

Background: This paper examined the influence of geographical location on emotional and behavioural problems of Canadian children aged 4-11 years. It assessed the impact of disadvantage of the family as compared to the neighbourhood on child problem behaviours.

Methodology: Data for this study came from 7,799 families (11,516 children) who participated in the first wave of the Canadian National Longitudinal Survey of Children and Youth (NLSCY) collected in 1994-1995. Multilevel modelling was used to estimate variations in child problem behaviour associated with geographic area and determine the explanatory power of socioeconomic disadvantage. The effects of geography were evaluated for three areas: the provinces of Canada; Census Metropolitan Areas (CMAs) and Census Enumeration Areas (EAs).

Dependent Variables:

• Scale score measures of hyperactivity, conduct and emotional problems (originally developed for use in the Ontario Child Health Study and Follow-up).
• Neighbourhood Disadvantage constructed from Census data on a) percentage of total neighbourhood income coming from government transfer payments; b) mean household income in 1000's of dollars; percentage of neighbourhood population aged 15 years and over c) without a secondary school certificate; d) with a university degree or certificate; and d) unemployed.

Independent Variables:

• Sex and age of child; sex, age, and birthplace (within or outside of Canada); number of siblings in the family, single or two-parent family, family income (above or below the poverty line) and SES of the Person Most Knowledgeable (PMK) about the child.

Results: Significant place-to-place variation existed in child problem behaviour, and the amount of variation depended on the size of the geographical area - about 6% of the variation was associated with between neighbourhood differences; about 2% between CMAs; and less than 1% between the provinces.

Neighbourhood disadvantage was also associated with child behaviour problems. Children were more likely to have conduct problems, hyperactivity or emotional problems if they came from a neighbourhood with a high percentage of single-parent families. Among the indicators of socioeconomic disadvantage, the strongest predictors of child problem behaviour were single-parent family status, and family SES. The independent variables in this study accounted for only 2.6 % (emotional problems) to 3.7% (hyperactivity) of the variance, after removing child sex and age from the model. Most of this variance was due to parent/family variables.

2. Brook, Judith S., Nomura, Carolyn, and Cohen, Patricia (1989). "A Network of Influences on Adolescent Drug Involvement: Neighbourhood, School, Peer, and Family," Genetic, Social, and General Psychology Monographs, 115(1): 125-145.

Keywords: drug use, cohesion, school, peer, and family influences

Background: This was a two-year study, set in upstate New York that examined the interrelationship of neighbourhood, school, peer, and family factors, and adolescent drug involvement. A total of 518 adolescents were seen twice throughout the study, while they were between the ages of 9-18 and 11-20. The sample was considered broadly representative of American families with respect to SES and family structure.

Dependent Variables:

• Drug involvement over time (frequency, duration, and severity of use)

Independent Variables:

• Neighbourhood factors: Cohesion (The people in this neighbourhood often share things to help each other out.); fear (People should not walk alone in this neighbourhood.); good/bad neighbourhood (Is this a good neighbourhood to grow up in?); satisfaction (I would move out of this neighbourhood if I could.)
• School factors: Autonomy (Students here choose a lot of their own academic program.); conflict (There is a lot of fighting between students in or around school.); personal ties (Good students here help out students who aren't doing well.); positive learning environment (Students and teachers are proud of the school.)
• Peer factors: Aggression against peers (How often does the child pick on his/her friends?); close friends (Do you have one or more friends who would turn to you for advice or help?); number of achieving friends (How many of your friends get all A and B grades?); general sociability (Do you enjoy yourself when you are with people your own age?); peer alcohol use, cigarette, and illicit drug use
• Family factors: Maternal and paternal affection and conflict, family SES

Results: Physical (bad neighbourhood) and social aspects (less cohesion, and overall dissatisfaction with the neighbourhood) were related to greater drug involvement over time. Greater school conflict, a negative learning environment and poorer peer and family relationships were also associated with increased drug involvement.

Each of the domains (neighbourhood, school, peer, and family) was significantly related to drug involvement over time. However, only the peer and family factors remained significant after controlling for demographic variables. Family and peer domains had direct effects on drug involvement while the neighbourhood and school domains had an indirect influence. Neighbourhoods with good living conditions that were socially supportive and perceived as satisfactory were linked to schools with little conflict and an emphasis on student independence in the context of a positive learning environment; good relations with achieving and non-drug-using friends; and a nonconflictual and affectionate parent-adolescent relationship.

3. Coulton, C., Korbin, J., Su, M., and Chow, J. (1995). "Community-level factors and child maltreatment rates." Child Development: 66: 1262-1276.

Keywords: community social organisation, child maltreatment

Background: This was a cross-sectional examination of the rates of child maltreatment in urban neighbourhoods in the state of Ohio, USA. It answered three research questions: 1) Are maltreatment rates a function of structural conditions associated with the level of community social organisation? 2) Are maltreatment rates and other behavioural outcomes (e.g., violent crime, drug use) interrelated? 3) Do the determinants of each differ?

Methodology: Using census and administrative agency data for 177 urban census tracts, the authors employed factor analysis to measure levels of social organisation within the area. The term community social organisation "refers to patterns and functions of formal and informal networks, institutions and organisations in a given area. Community social organisation is strong to the degree that these local structures are able to accomplish the goals of residents and exert social control from within the community."

Dependent Variables:

• Rates of child maltreatment, violent crime, drug trafficking, juvenile delinquency, teen childbearing, and low birth weight

Independent Variables:

• Community social organisation: measured by examining a variety of characteristics of a community and its residents. Factor analysis revealed that indicators of many of these concepts were highly inter-correlated and therefore three indicators of community structure were employed - impoverishment (poverty, employment, vacant housing, population loss, female-headed households) child-care burden (ratios of kids to adults, females to males, % of population that is elderly) and instability (proportion of residents that moved in last five years, those with household tenure less than 1 year and over 10 years). Factor scores were calculated for each census tract for each of the dimensions of community structure. Additionally, a geographic location variable was calculated for each tract because of the recognition that the resources in an area can be affected by resources available in surrounding areas.

Results: Impoverishment was highly correlated with maltreatment rates. Areas with the highest incidence of maltreatment among children were those that also experienced poverty, unemployment, female-headed households, racial segregation, abandoned housing, and population loss. The child-care burden factor had a significant but somewhat weaker effect (maltreatment rates were higher in areas with many kids, few elderly, and low proportions of adult males). Areas with higher mobility (i.e., instability) also had higher maltreatment rates. There was an interaction between instability and impoverishment, with the effects of instability on maltreatment being less pronounced in areas that are most impoverished. Risk of maltreatment was also higher for neighbourhoods in close proximity to high poverty areas. Finally, child maltreatment rates were indeed correlated with other types of deviant behaviours.

Findings suggest that a community's level of social organisation, functioning, networks, and resources were important influences on their rates of maltreatment and other incidence of deviant behaviours. The seemingly paradoxical relationship between instability and impoverishment needs to be further examined. However, one explanation may be that because a lot of mobility in impoverished areas is over short distances (i.e., within blocks of the original location) perhaps social networks are not being disrupted and therefore problems are minimized.

4. Dewitt, D.D., Offord, D.R., and Braun, K. (1998). The Relationship Between Geographic Relocation and Child Problem Behaviour,Applied Research Branch, Strategic Policy, Human Resources, Development Canada. Working paper W-98-17Es.

Keywords: moving, problem behaviour (school failure, substance use), social control, attachment to social institutions

Background: The goal of this study was to determine the extent of the relationship between a family's geographical relocation and subsequent problem behaviours in their children. The research was guided by Social Control theory, which states that problem behaviour results from children's attachments to institutions and people (e.g., schools, family) being broken. The role of social control as a mediator of the relationship between relocation and problem behaviours was examined. Data from cycle one of the Canadian NLSCY were employed, and children aged 0-11 were studied.

Dependent Variables:

• Problem behaviour: Physical aggression-conduct problems, indirect aggressive and antisocial behaviours, property offenses, school failure, lifetime tobacco and alcohol use.

Independent Variables:

• Geographic relocation - total moves and recency of last move.
• Measures of social control (mediators): Family-related (harmony, parental monitoring, parenting), school-related (negative school attitudes, number of school changes, low achievement), and community-related (participation in sports and other organised activities outside of school).

Results: Moving house was a common experience for children in Canada, while 25% had never moved, 32% had moved three or more times. The number of residential moves a child had experienced was an important indicator of problems, with those who had moved three or more times being twice to three times more likely than non-movers to experience the variety of negative behavioural outcomes (with the exception of antisocial behaviours).

Some mediating effects were observed. Children were less likely to have problems if they had high parent/child attachment and family harmony and more positive attitudes toward school. Problems were more common in children experiencing inconsistent and punitive parenting practices. Increased vulnerability to the effects of moving was observed among children with lower levels of academic achievement and more negative attitudes toward school. Frequent involvement in out-of-school sports or other community activities (i.e., attending church or participating in clubs) also exerted a protective effect.

Overall, moving was not an inherently stressful process for many children. These results tend to support a commonly held view that moving contributes to aberrant child behaviour by intensifying problems (i.e., problem behaviour risk factors) that already exist within the family.

5. Kohen, Dafna E., Hertzman, Clyde, and Brooks-Gunn, Jeanne. Neighbourhood Influences on Children's School Readiness, Applied Research Branch, Strategic Policy, Human Resources Development Canada. Working paper W-98-15E.

Keywords: school readiness, safety, cohesion, toddlers' and preschoolers' competencies

Background: This study examined the effects of neighbourhood and family characteristics on children's school readiness. The sample included toddlers (aged 2-3) and preschoolers (aged 4-5) who participated in the first cycle of the Canadian NLSCY. For each age group and for each competence measure, the first analytical model estimated neighbourhood effects after controlling for province of residence. Subsequent models then examined the mediation role of family characteristics on these relationships.

Dependent Variables:

Measures of school readiness:

• Toddlers (aged 2-3) - Maternal reports of motor and social development and problem behaviour
• Preschoolers (aged 4-5) - Receptive verbal abilities (using the PPVT-R) and problembehaviour scores (maternal rating)

Independent Variables:

• Family variables - province, sex of the child, number of people in the household, PMK age at the birth of the child (teen or non-teen)
• Family socio-economic characteristics - household income, PMK's level of education and single-female head-ship
• Neighbourhood characteristics: Poverty (percentage of families in the neighbourhood with household incomes less than $20,000); affluence (percentage of families in the neighbourhood with household incomes greater than $50,000); neighbourhood family structure (percentage of single female headed families); neighbourhood unemployment (above the national average versus below the national average, 1994); neighbourhood safety -interviewer's ratings of the safety and general condition of the neighbourhood (e.g., volume of traffic; presence of garbage, litter, or broken glass; any people loitering, arguing, shouting or fighting; conditions of buildings on block); neighbourhood cohesion - a scale score of 5 items: (e.g., there are adults in the neighbourhood that children can look up to; people are willing to help their neighbours)

Results for Toddlers: Neighbourhood affluence was an important determinant of motor social competence while affluence and fewer single female-headed families were significantly associated with behavioural competence. However, family characteristics such as high levels of household incomes and maternal education mediate neighbourhood effects and were significant independent contributors.

Results for Preschoolers: Neighbourhood socio-economic characteristics such as affluence, neighbourhood safety and cohesion, poverty and single female head-ship had an effect (in the expected direction) on preschoolers' verbal ability scores. Family characteristics such as high levels of household income and maternal education were associated with higher verbal ability scores for children. These family characteristics mediated the effects of neighbourhood socioeconomic characteristics. The effects of neighbourhood female head-ship on children's verbal ability scores were mediated by neighbourhood safety.

Neighbourhood poverty, single female-headed families, and unemployment were associated with higher behavioural problem scores. Ratings of unsafe neighbourhoods were associated with behaviour problems but these effects were mediated by PMK ratings of neighbourhood cohesion. Family characteristics such as low levels of PMK education and single female-head-ship were associated with higher ratings of behaviour problems but neighbourhood effects persist over and above family effects.

The results of this study provided some evidence that neighbourhood variables appear to exert larger, more direct impacts as children get older. This can be confirmed by further examination of neighbourhood effects on school-aged children using additional outcome measures such as standardized tests and teacher ratings of competence, available for older children. One limitation of this study was the issue of selection bias (the process whereby families chose particular neighbourhoods in which to live), leading to difficulties in separating neighbourhood effects from those that result from the characteristics of the family that chooses to live in the neighbourhood. Few studies have controlled for this bias.

6. Offord, David R., Lipman, Ellen L., and Duku, Eric K. (1998). Sports, the Arts, and Community Programs: Rates and Correlates of Participation, Applied Research Branch, Strategic Policy, Human Resources Development Canada. Working paper W-98-18E.

Keywords: sports, arts, community participation, socio-demographic factors, problem behaviours

Background: This paper employed data from the Canadian NLSCY to assess children's rates of participation in sports, arts, and community programs according to selected socio-demographic variables, and whether such participation was associated with improved psychosocial adjustment in children.

Dependent Variables:

(Note: These variables were also used as independent variables in some of the analyses)

• Participation in any sports involving coaching or instruction outside of school hours; participation in unorganised sports or physical activities; taken lessons or instruction in music, dance, art or non-sport activities; and taken part in any clubs, groups or community programs with leadership (Scouts, Girl Guides, Brownies, Cubs, Church groups).

Independent Variables:

• Sociodemographic and family variables - age groups (6-8, 9-11), gender, income, single-parent status, number of siblings, and family functioning (scale of six dimensions, including problem solving, communication, roles, affective responsiveness, affective involvement, and behaviour control).
• Community variables - availability of good parks and playgrounds, and extent to which area was seen as good place to bring up children (deemed a civic neighbourhood).
• Child characteristics - presence of one or more problems (emotional or behavioural disorder, repeated a grade, or impairment in social relationships). In one of the analyses, this variable was used as an outcome measure.

Results: Incidence of one or more problems decreased if children were younger, female, or participated in unsupervised sports or the arts. Low income, single-parent status, family dysfunction, and living in non-civic neighbourhoods were independently associated with an increased rate of problems.

Overall, participation rates were low in all types of programs, with many children reporting that they never participated in these activities. Furthermore, participation was lower for children from lower income groups, single parent families, and except for community programs, those with more siblings.

In terms of community variables, strong agreement or agreement by the PMK that there were good parks, playgrounds, and play spaces in the neighbourhood indicated increased rates of participation in each of the activities, (except community organisations). A non-civic neighbourhood was associated with higher rates of almost never participated in all activities (strongest relationship seen with sports).

7. Sampson R, Raudenbush S, and Earls F. (1997). "Neighbourhoods and Violent Crime: A Multilevel Study of Collective Efficacy." Science, vol. 277 (August 15): 918-924.

Keywords: collective efficacy, social cohesion, informal social control, violence

Background: This article analyzed data from a 1995 survey of 8,782 residents of 343 neighbourhoods in the community design component of the Project on Human Development in Chicago Neighbourhoods. It was hypothesized that collective efficacy, defined as social cohesion among neighbours combined with their willingness to intervene on behalf of the common good, was linked to reduced violence.

Dependent Variables:

• Incidence of violence in neighbourhood
• Personal victimization (e.g., mugging, fight, or sexual assault)
• Census variables to reflect neighbourhood differences in poverty, race and ethnicity, immigration, the labour market, age composition, family structure, home-ownership, and residential stability

Independent Variables:

• Collective efficacy constructed using the combination of two like-type scales - a) informal social control; and b) social cohesion and trust (See Appendix for complete details on variables.)

Results: Personal background variables - high SES, home-ownership, and age were positively associated with collective efficacy, whereas high mobility had a negative association. At the neighbourhood level, after controlling for these personal background variables, concentrated disadvantage and higher immigrant concentration decreased incidence of collective efficacy, whereas residential stability had the opposite effect.

Even after controlling for the relationship between neighbourhood social composition and violence, a strong association remained between higher levels of collective efficacy and lower incidence of violent acts. Increasing collective efficacy in a neighbourhood was also related to decreased personal victimization, even after taking into account the effects of neighbourhood characteristics such as concentrated disadvantage, residential stability and immigrant population. Lack of collective efficacy within a neighbourhood was also associated with increased homicides.

Three other scales derived from the community survey were also examined: neighbourhood services, friendship and kinship ties, and organisational participation. When these factors, along with neighbourhood characteristics (e.g., prior homicides, concentrated disadvantage, concentration of immigrants, and residential stability) were controlled, collective efficacy remained the largest predictor of violent crime. Results suggested that personal ties, organisational affiliation, and local services by themselves were not sufficient to reduce violence, which was more directly attributable to informal social control and cohesion among residents.

The authors noted that other dimensions of neighbourhood efficacy that may be potentially important should be explored (e.g., political ties) as well as additional influential factors that were linked to the wider political economy need to be examined.

8. Simcha-Fagan, Ora, and Schwartz, Joseph E. (1986). "Neighbourhood and delinquency: An assessment of contextual effects," Criminology, 24(4): 667-703.

Keywords: organisational participation, personal ties, criminal subculture, adolescents

Background: The authors differentiated between the effects of individual and family factors from community contextual factors on delinquency for a sample of adolescent males in twelve New York City neighbourhoods. The 12 neighbourhoods had relatively similar social-demographic profiles. Measures of neighbourhood characteristics were derived from census data and information reported in the Simcha-Fagan neighbourhood questionnaire.

• Neighbourhood characteristics: A pool of 90 items was factor analyzed to identify 3 separate neighbourhood dimensions: 1) Deviant-Criminal Subculture (e.g., low community attachment and network size and breadth, social disorder, conflict subculture and illegal economy, and neighbourhood anomie); 2) Informal Structure of Personal Ties/ Community Informal Structure (average residential stability, informal neighbouring, extent of local personal ties);3) Formal Institutional Structure / Community Organisational Participation (average parental education, and community organisation and involvement).
• Individual measures included school attachment-commitment, association with delinquent peers, self-reported delinquency, and official recorded delinquency.

Results: The community's level of organisational participation and presence of social disorder-criminal subculture were the two factors most strongly associated with adolescent delinquency. As expected, high community participation was linked to low delinquency, while higher levels of social disorder led to high delinquency. Low family income, delinquent peers, and older ages were related to high self or officially reported delinquency.

Indirect effects were also present. For instance, residential instability was associated with decreased community participation, which in turn led to lower school attachment and higher reported delinquency. Furthermore, low economic characteristics of communities were associated with the presence of disorder-criminal subculture, which in turn had a direct effect on delinquency.

9. Duncan, Greg J., & Stephen W. Raudenbusch (1998, #3). Neighborhoods and Adolescent Development: How Can We Determine the Links? Joint Center for Poverty Research, Northwestern University Institute for Policy Research, Evanston, IL. http://www.jcpr.org/neighbor_dev.html

Keywords: neighborhood context, adolescent development, behaviour, measurement, methodology

Background: This study aimed to identify promising survey methods that may be used to obtain an unbiased understanding of neighbourhood effects. It was argued that: 1) the use of small samples from a few select neighbourhoods diminishes the possibility of distinguishing among the various ways that context may influence youth; 2) studies that measure neighbourhood characteristics based on parent/youth self reports will likely produce spurious results, especially when youth outcomes are based on self report as well; 3) more reliable neighbourhood data may be drawn from independent samples of residents or by systematic social observation (SSO); 4) employing outcome correlations for youth living in close proximity to one another is an effective means of estimating the upper bound of neighbourhood effects; 5) quasi- and random-assignment experimental studies represent the most promising method to date for determining neighbourhood influences.

Methodological issues: The discussion of methodological issues was framed within a model where adolescent i's achievement or problem behaviour (y) is an additive function of i's family (FAM) and extra-familial context (CON):

y_i = A'FAM_i + B'CON_i + e_i

Simultaneity: First addressed was simultaneous causation, where contextual conditions may be caused by the behaviour of the adolescent just as contextual conditions influence behaviour. This was considered especially important in the extra-familial context, for example, where individuals self-select themselves into particular peer groups, hence determining the context within which they find themselves.

Omitted-context variables: When regression analysis was conducted with adolescent outcomes as the dependent variable and family and contextual characteristics as independent variables, estimates of respective influences may be biased should context variables be omitted in the equation. This type of error becomes particularly problematic when using administrative data approaches (i.e. census data), which may use single measures of neighbourhood characteristics.

Endogenous membership: Neighbourhood context has been identified as a function of individual constraints and decision-making, rather than an entirely random process. For example, the propensity of families to live in better or worse neighbourhoods depends in part on a combination of parental background, characteristics and choices. The complexity of this issue is problematic for measurement, hence making the direction of the relationship and biases difficult to identify.

Families as mediators and moderators: Neighbourhood characteristics, while affecting children directly, also shape family characteristics such as income, living conditions and parental mental health, which may in turn affect children. If this is indeed true, accounting for mere family differences may underestimate the effects of neighbourhood context, and the authors claim it is more useful to conceive of these relationships recursively, with families acting as mediators.

Variability in contextual characteristics: Diversity among neighbourhood contexts implies a need for widely dispersed samples across a variety of communities. Further, it is claimed that researchers must go beyond assessing "good" and "bad" neighbourhoods, recognizing all contexts that exist along the continuum and how competing theories of neighbourhood effects may be applied. In attempting to achieve a representative sample while remaining within budget constraints, many surveys use cluster samples. The authors cautioned against this practice due to decreased geographic variability, which may prevent a true modelling of neighbourhood context.

2.5 Measures / Surveys

See Appendix A for a full listing of community-relevant questions.

1. Program on Human Development in Chicago Neighbourhoods - Community Survey Questionnaire - 1994

This is an excellent source of questions that address a variety of community issues including: perception of crime; neighbourhood quality characteristics; neighbourhood cohesion (e.g., ties to others in the community, ability to rely or to be able to turn to neighbours when there are problems in the community, or when in need of assistance); and perception of one's own neighbourhood (in terms of the boundaries governed, name, major landmarks or stores, satisfaction with neighbourhood as a place to live, qualities of previous neighbourhood, community organisations and services). Interviewers also observed neighbourhoods and commented on visibility of garbage, lighting, people seen, feelings regarding own level of comfort/safety, and land use in neighbourhood).

2. Boston Dorchester Cares Project - Neighborhood Interview

This survey contains a variety of questions pertaining to the respondent's own spatial-geographic conception of their neighbourhood and general attitudes toward their neighbourhood as a place to live and bring up children. There are also questions relating to relationships with others in the neighbourhood such as familiarity with neighbours and social ties. One section addresses social problems in the community, both from the respondent's perspective and from the perspective of the interviewer. Other questions address participation in community efforts to address problems, attendance of religious, educational, social, political, or other groups, and their awareness of community programs, organisations, and people. Questions pertaining to the respondent's own personal health over the past few weeks (regarding anxiety, stress, other symptoms of poor health), parenting practices, and their relationship with a partner are also included.

3. Simcha-Fagan Questionnaire

The questions in this survey are very similar (some identical) to those employed in the Chicago Neighbourhood Survey. For example, this questionnaire has questions about the visibility of various social problems in the neighbourhood and general sentiments of people in the neighbourhood regarding being out alone at night. Respondents are asked to identify (by name, and on a map) their neighbourhood, and whether they believe the boundaries identified are shared by most people in the area. The survey also includes questions about the availability of special programs and services directed at children of all ages and about the existence of a neighbourhood newspaper, bulletin, or newsletter. Involvement in community organisations and associations is also probed, both within and outside of their own neighbourhoods. Respondents are also asked to compare their own neighbourhood to others in the city, in terms of safety (danger), and whether they believe their neighbourhood has changed in this regard in the past couple of years. The last section includes selected questions pertaining to social cohesion.

4. Calgary Youth Violence Survey

Part C of this survey contains a number of pertinent questions pertaining to self-identification of one's community, general attitudes toward the community, and what could be done to improve it. There are many questions that address participation in community activities, membership in clubs or groups, and social life (including type of activities engaged in, how often, and with whom). Another section deals with issues related to crime and criminal activity including: criminal involvement, related resources available within the school and frequency of use of resources, and contact with/general perception of the police. Respondents are also asked about their perception of youth crime and its frequency in their community as compared to other areas of the city, sources of anxiety/fears about their own community, and steps or practices taken to protect self (while at home, or in the community).

5. Quality of Life Survey - York University

Section A of this survey is focused primarily on the neighbourhood a respondent lives in, with a few questions pertaining to the broader community (city). There are a series of questions that address neighbourhood characteristics including the quality and number of resources (schools, parks and playgrounds), physical condition of streets, houses and buildings, and the physical environment (noise and air pollution). There are a few questions that address perceptions of others in the neighbourhood, relations and ties with neighbours, and sense of belonging (feeling a part of a particular neighbourhood - socially or culturally). Some questions also address satisfaction with the neighbourhood, duration of residence, and safety from crime. The questions pertaining to the broader community include those that ask about the availability of shopping, entertainment, cultural, recreational and sports facilities/ opportunities, as well as job opportunities and the general economic environment.

6. The Panel Study of Income Dynamics (PSID)

This is a longitudinal survey of a representative sample of U.S. men, women, and children, and their families. Data on employment, income, wealth, housing and food expenditures, transfer income, neighbourhood characteristics, and marital and fertility behaviour as well as intergenerational data and information on life events such as early childbearing, divorce, illness are included. A component on child development covering children from birth through age 12 is also available.

7. Survey of Volunteering

This survey includes questions regarding social activities and organisations to which individuals belong and the frequency of their involvement in these activities. The types of organisations listed include charitable, professional, political, cultural, educational, hobby or sports organisations, religious affiliation, neighbourhood, civic, community or school groups. Respondents are also asked about the type of involvement they have with any group (pay membership dues, attend meetings, participate in meetings, other) and the number of associations or organisations of which the person is a member or participant. There are also questions about voting in the last federal, provincial, municipal or local elections.

8. Australian Living Standards Survey

This survey defines neighbourhood as 'your street and the few streets around it', or ' the area close by'. Section 124 of the survey includes a series of questions (17) pertaining to neighbourhood quality such as the physical environment, access to public parks or playing fields with play equipment for young children, and the visibility of police services. Section 125 of the survey asks respondent ten questions that pertain to how well they know their neighbours (i.e., social cohesion), what they would most like to change or improve about their neighbourhood; and crime victimization. The last section of the survey has 5 general questions regarding the family's present economic circumstances.

9. Coulton, C.J. (1995). "Potential and Problems in Developing Indicators on Child Well-Being," in Indicators of Children's Well-Being, Volume III. Cross-Cutting Issues: Population, Family, and Neighbourhood: Social Development and Problem Behaviours. Paper prepared for the Conference on Indicators of Children's Well-Being, Rockville, MD. 1995. Institute for Research on Poverty Special Report, No. 60c.

Though this is a paper focused largely on a theoretical discussion of the outcome orientation vs. contextual orientation in measuring community variables, it does provide good sources of community indicators. Community outcome measures for children are available from the following sources: U.S. Census, municipal housing and police departments, county coroner, children's services, vital registry, juvenile court, and through boards of education. These outcome measures include those that pertain to social behaviours (teen childbearing, delinquency, and drug violation arrest rates), health and safety, cognitive development and achievement, and economic well-being. Community context indicators (sources such as - Census, public assistance counts, municipal housing, police, recreation and parks departments) include those pertaining to socio-economic composition, age and family structure, residential mobility, and environmental stress (vacant and boarded houses, housing code violations, and personal crime). Support for effective parenting (school locations, recreational opportunities, and community participation), and drug arrests are also examples of contextual indicators.

Previous

Contents

Next