skip to primary navigation skip to content

Department of Geography


Identifying communities with census data for social science applications.

Research Team: Jane Law with CRISP


This research aimed to develop tools for identifying communities with census data in GIS. The communities identified are useful for social science applications (e.g., planning of health and educational resources). The methods being studied were cluster analysis with or without contiguity constraint, and a new approach that used a new spatial community statistic developed. Communities are interpreted as areas within which characteristics, such as income, are similar. Neighbours of a community have different characteristics (e.g., different income level). Cluster analysis methods were useful in classifying areas, but have their limitations in identifying areas with similar characteristics that were different from their neighbours. Using the spatial community statistics developed that incorporate the meaning of communities explicitly in the mathematical model, univariate communities were identified. Univariate communities were then overlayed using geographical information systems to identify areas that belong to univariate or multivariate communities.


Image as described adjacent

Image as described adjacent

Figure 1: Identifying communities by cluster analysis

Image as described adjacent

Figure 2: Identifying communities with similar education: analysing the similarities within areas compared to their neighbouring characteristics

Further readings

  • Law, J., and D. Willms (2002). “Provincial maps depicting neighbourhood types” in: D. Willms (ed.) Vulnerable Children, The University of Alberta Press, p. 389-406. (Winner, Canadian Policy Research Award).
  • Law, J. (2000). Spatial Analysis of Multivariate Demographic Data for Identifying Communities in GIS. PhD. thesis, University of New Brunswick, Canada.
  • Law, J., D. Coleman, J. McLaughlin, and J.D. Willms (1999). “Exploring communities using multivariate demographic Data.” Paper presented at Geo99: Mapping the Future: Tools, Techniques, Technologies, Fredericton, New Brunswick, Canada.
  • Law, J. (1998). “Spatial analysis of multivariate demographic data for identification of communities in GIS.” Paper presented at the Atlantic Institute Student Research Conference, Fredericton, New Brunswick.
  • Law, J., D. Coleman, J. McLaughlin, and J.D. Willms (1998). “Identifying communities through spatial analysis of multivariate demographic data.” Paper presented at Spatial Data Infrastructure’98: The 10th International Geomatics Conference, Ottawa, Canada.