RDC Proposal - Peter Smith, McMaster University
1. Project title
Sense of belonging to local community: Health, Socio-economic status and geography
Requesting access to the McMaster University RDC.
2. Rationale and objectives of the study:
Sense of community belonging is a concept related to levels of social attachment among individuals and is indicative of social engagement and participation within communities. Research has shown that there is an association between social networks and health outcomes. Social isolation can adversely affect health while social engagement and attachment can lead to positive health outcomes (Statistics Canada 2005; Choenarom et al 2005).
Recent work by Statistics Canada and others has demonstrated the link between sense of community belonging and certain health and socio-demographic outcomes. Using 2000/2001 CCHS data, Ross (2002) found that just over half (56%) of Canadians reported a strong or somewhat strong sense of community belonging and that this sentiment was associated with self-perceived health. In a follow-up study employing 2005 CCHS data, Shields (2008) examined sense of community belonging at the provincial/territorial and health region levels. The author found that nearly two-thirds of Canadians (64%) reported a strong sense of community belonging, highest in Atlantic Canada and the territories, and that there was a strong association between sense of belonging and mental health.
There is a need to explore more fully the regional geographic dynamics of sense of belonging and health in Canada, particularly across the urban to rural continuum and among settlements of different sizes (in terms of population). This will provide a more nuanced understanding of the association between belonging and physical and mental health issues according to certain social and environmental characteristics.
For example, in a study of San Francisco, Pendola and Chen (2005) found that sense of community was strongest among residents of neighbourhoods exhibiting characteristics of a ‘main street’ town. In a study of South Australia, Ziersch et al (2009) found that higher levels of social capital were significantly associated with better mental health for both urban and rural residents.
There is also a need to examine intra-urban variations in sense of belonging as we can assume that belonging will vary considerably among residents of different communities within a city or region.
The purpose of this project is to examine sense of belonging to local community in Canada. The study will address two principal research questions:
- What are the major health, social and geographic determinants of sense of belonging to local community and how does sense of belonging vary across the urban to rural continuum and among settlements of different size?
- Within the Hamilton CMA, how does sense of belonging vary among different communities according to their geographic location?
3. Proposed data analysis and software requirements:
The data analysis will be carried out using Stata 13. I am interested in examining the total population aged 18 and over in the provinces and territories and by urban-rural status and according to Metropolitan Influence Zone (MIZ). Only weighted results will be requested for disclosure from the RDC. To account for the sample design of the survey, bootstrap procedures will be employed to calculate confidence intervals and coefficients of variation, and to test the statistical significance of difference. A significance level of p<0.05 will be applied in all cases.
The data analysis will involve two steps. The first is descriptive statistics by way of bar charts and contingency tables to measure the relationship between sense of belonging and a number of independent variables (see table below). The second step in the data analysis is ordered logistic regression modeling. Compared to ordinary least squares (OLS), this technique is more appropriate given the categorical nature (4 point Likert scale) of the dependent variable: sense of belonging to local community. The ordered logistic model more appropriately accounts for nuanced differences across the categorical scale variable and controls for the constraints of the data; neither logistic or OLS address these issues sufficiently. Further, the ordered logistic model allows for a more parsimonious presentation of output given the proportional odds assumption (e.g. parallel regressions), compared to more generalized models.
4. Data Requirements:
I am requesting access to the confidential Master Data File for the à/08 Canadian Community Health Survey (CCHS). The CCHS contains a question on sense of belonging to local community as well as representative data on various aspects of health-care services, socio-demographic characteristics and health status. As a result, the CCHS is an appropriate data source for the research. The principal objective of the study is to examine regional variations in sense of belonging in Canada as well as its health and socio-economic determinants. The CCHS PUMF provides two levels of geography – provincial and health region. I intend to conduct a cross-sectional analysis for additional levels of geography. Specifically, I would like to identify rural residency in Canada based on distance from a larger urban area according to metropolitan influence zone (MIZ). Also, if the sample size permits (after consultation with the RDC staff at McMaster in compliance with data disclosure rules) I would like to examine sense of belonging at the sub-CMA level in Hamilton according to Forward Sortation Areas (FSA). Finally, I am interested in exploring sense of belonging among Canada’s Aboriginal population. The Aboriginal analysis will be conducted only at the national and provincial level and will not be part of the Hamilton case study.
Specifically, I am requesting access to the 2007/08 CCHS Master File to access five variables to the data analysis that are not available in the public use files:
- Statistical Area Classification (GEODSAT) (CMA, CA/Strong/Moderate/Weak/No MIZ).
- Rural and Urban Grouped (GEOBDUR2)
- Postal code (GEODPC) **the first 3 digits (FSA) will be used and only for Hamilton if sample size permits according to Statistics Canada disclosure rules**.
- Population size group (GEODPSZ)
- Aboriginal identity (SDCDABT)
I will consult with the staff at the McMaster University RDC to ensure that acceptable sample sizes can be derived from each variable and that appropriate weighting and bootstrapping procedures are applied to the data.
Population of Interest
The population of interest in the study includes CCHS respondents aged 18 years and older in Canada (2007/08 n=120,838) (research question 1) and CCHS respondents aged 12 years and older in the Hamilton CMA (2007/08 n= 2,128) (research question 2).
A number of variables have been identified from the 2007/08 CCHS master file to meet the objectives of the study. The table below lists dependent and independent variables of interest. Several variables will be re-coded in STATA 11 to reduce the number of categories representing values.
|Sense of belonging (dependent)|
|GEN_10||Sense of belonging to local community|
|GEN_02B||Self-perceived mental health|
|GEN_07||Perceived life stress|
|GEN_08||Perceived work stress|
|PACDPAI||Leisure Physical Activity Index|
|GEO_PRV||Province of residence of respondent|
|GEODUR2||Urban and rural areas (2 levels)|
|GEODSAT||Statistical Area Classification – Metropolitan Influence Zones (MIZ)|
|GEODCMA6||2006 Census Metropolitan Area (CMA)|
|GEODPC||Postal code (data for first 3 digits will be analyzed and only if sufficient sample sizes can be obtained according to disclosure rules).|
|GEODPSZ||Population size group|
|INC_3||Total household income – best estimate|
|EDUDR04||Highest level of education - respondent, 4 levels|
|DHHDDWE||Type of dwelling|
|DHH_OWN||Dwelling – owned by a member of household|
|Source: Statistics Canada.|
5. Expected project start and end dates:
This project is expected to start September 1, 2010 and continue until August 31, 2012.
6. Expected products:
It is expected that this project will result in the following products:
- A book-chapter
- 2-3 journal articles
- Presentations at academic conferences
Choenarom, C. Williams, R.A. and Hagerty, B.M. (2005). The Role of Sense of Belonging and Social Support on Stress and Depression in Individuals with Depression. Archives of Psychiatric Nursing, Vol. 19, No. 1, pp 18–29.
Neal, S. and Walters, S. (2008). Rural Be/longing and Rural Social Organizations: Conviviality and Community-Making in the English Countryside. Sociology. Vol. 42, no. 2, pp.279-297.
Pendola, R. and Gen, S. Does “Main Street” Promote Sense of Community? A Comparison of San Francisco Neighborhoods.Environment and Behaviour. Vol. 40, No. 4., pp. 545-574.
Ross, N. (2002). Community belonging and health.Health Reports, Vol. 13, No. 3, March. Statistics Canada, Catalogue 82-003.
Shields, M. (2008). Community belonging and self-perceived health. Health Reports, Vol. 19, No. 2, June. Statistics Canada, Catalogue 82-003.
Statistics Canada, Health Statistics Division (2005).Community Belonging and Self-perceived Health: Early CCHS Findings (January to June 2005). Statistics Canada, Catalogue 82-621-XIE.
Ziersch, A.M., Baum, F., Darmawan, I, Kavanagh, A.M. and Bentley, R. Social capital and health in rural and urban communities in South Australia. Australian and New Zealand Journal of Public Health. Vol. 33, No. 1, pp 7-16.