Appendix D Sample distribution for the models

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Two regression models were developed to explore the question, "Who is likely to spend 30% or more of income on shelter costs?"

The goal of the first model was to compare the characteristics of those who ever (at least one year) spend 30% or more of household income on shelter to those who never do. The sample was divided into the ever group (with STIR at or above the 30% benchmark for at least one year between 2002 and 2004) and the never group (with STIR never at or above the 30% benchmark during all three years).

In the second model, the goal was to focus on the population which persistently spends 30% or more of household income on shelter, in this case for the full 3-year study period. The sample was divided into the group persistently spending 30% or more of their income for shelter and the never + occasionally group (with STIR below the 30% benchmark for at least one year during the three-year period).

Model 1 regresses the ever indicator and Model 2 regresses the persistently indicator against the socio-demographic and geographic characteristics of Canadians and the households in which they live. See tables 20 and 21 for more information about how the sample has been divided into the ever and persistently groups.

Note that the models exclude those aged less than 16 because certain questions such as Aboriginal status, immigration status and education level are not asked of children. Note also that income was not included in the models because it is part of the calculation of the characteristic of interest – the STIR.