PSYTE™ HD Financial geodemographic data is part of a suite of segmentation products belonging to the PSYTE™ HD family. Based on the PSYTE™ HD Canada geodemographic clustering system, PSYTE™ HD Financial geodemographic data leverages a substantial amount of geodemographic research that went into the development of atoms for PSYTE™ HD Canada geodemographic data in 2014. These atoms, numbering approximately 200, are the principal neighbourhood types in Canada and represent consistent geodemographic communities.
PSYTE™ HD Financial geodemographic data was built using several input databases. Primary clustering of the 2014 version of PSYTE™ HD Canada was built on Census 2011 data. Clusters were validated in 2018 using Census 2016 data. Other inputs, including the Canada Wealth database, were also used to determine PSYTE™ HD Financial geodemographic clusters.
The starting point for PSYTE™ HD Financial geodemographic data uses over 20 variables derived from numerous census variables related to financial and other characteristics of households, along with data from Canada Wealth, produced by Precisely. Due to the highly collinear nature of these variables (from a variance standpoint), the data was transformed into a reduced number of principal component factors, where each factor significantly represented a unique financial characteristics or financial behaviour of households. These atom-level factors represented the starting point of the atom-to-cluster development process.
Using a proprietary version of Ward's hierarchical clustering, atoms were agglomerated together until 15 financial clusters emerged. These clusters effectively characterize neighbourhood types according to their primary or dominant financial characteristics and behaviours.
For the 2018 update, PSTYE™ HD Financial geodemographic clusters were validated as statistically relevant clusters using updated Census 2016, Canada Wealth, and other socioeconomic datasets. Approximately 13 percent of Dissemination Areas were assigned to other clusters based on changes in population and household characteristics. In other words, some DAs experienced sufficient change in their socioeconomic characteristics to warrant reassignment to different PSYTE™ HD Financial geodemographic clusters. Nevertheless, all clusters were re-validated with 2016 census and wealth data so that they maintained their cluster names and descriptions. All descriptions were refreshed with updated data at the time of development.