Electronic Thesis and Dissertation Repository


Master of Science




Dr. Jason Gilliland

2nd Supervisor

Dr. Godwin Arku

Joint Supervisor


Geographical Information Systems (GIS) has emerged as a primary tool to study the built environment (BE) and its correlates. Accurate results employing GIS-based methods, however, depend on high-quality data sets and agreement on the appropriate nature and scale of areal units. Exploratory groundtruthing exercises illuminated possible issues of data accuracy. GIS-derived walkability measures for intersection density and land use mix were calculated for 2011 using Census of Canada data to determine whether commercially- or municipally-sourced urban data sets were better suited for BE studies. Road network data were ‘cleaned’ to remove unwalkable sections of road and to repair instances of intersection overrepresentation. Results suggested that municipal land use data and commercial road network data were ideal and both sources of data have benefits and limitations that should be discussed prior to any large research analysis. The second major analysis in this thesis utilized DMTI road and land use data and journey-to-work transportation data for three Canadian census years – 2001, 2006, 2011 – for thirty-three Canadian Census Metropolitan Areas (CMAs) to determine if a Canada-wide association between walkability and active and public transportation (AT/PT) exists. Using a Spearman’s Rank correlation matrix, walkability was positively associated with AT, PT, and AT/PT (r = 0.324 – 0.459; p < 0.01) and negatively associated with median household income and age of development (r = 0.124 – 0.222; p < 0.01) for all three census years. Study findings can be used to inform urban transportation and public planning policies dependent on multi-modal transit networks.