Objectives
		Our first objective is creating the five deprivation indices using the census data at various scales.  First of all, the indices are calculated at the Dissemination Areas (DA) level. Secondly, these entire areal data are aggregated into their respective aggregate based on the centroids proximity within the LHA administration units. Centroids of each DA are created in ArcGIS to stop the problem of joint ownership between DA’s and the LHA boundaries due to data inaccuracy and occurrences of multiple overlaying.  Thirdly, the areal data at DA level are also aggregated into CT level.				The second objective is to depict the correlation between the multiple indices for classifying socioeconomic deprivation with the study area at different scales.  It is also important to identify if multiple deprivation indices can equally identify inequality within the GVRD. As these indices were originally created separately, and as stand alone scores, it is interesting to study their framework under a comparison. Correlation tests were run at the DA, CT, and LHA level to test for similarities and aggregation as these tests are beneficial for determining the strength of the association within opposing datasets.
		The third objective is to identify if the areal data demonstrate a stronger spatial association at a larger scale or at a smaller scale. This analysis involves a test for spatial correlation. The method was chosen to observe if the study area exhibits a spatial pattern, such as clustering or random pattern.  As well, these patterns may also be dependent on scale. The test for spatial autocorrelation was a global measurement, using Moran’s I.				Our fourth objective is to address whether the multiple indices are adequate or too vague in their place-based representations of population aggregations and classifications of inequality within LHA units.  The standardized mortality ratios (SMR) are compared with each deprivation index to examine if there is a commonality between them at LHA level.  The Empirical Bayes estimation is used to smooth out the SMR.