The target population is comprised of all Canadian residents who were 16 years of age or older at the time of data collection, excluding long-term institutional residents, families of members of the armed forces on military bases, and individuals living on Indian Reserves.
Residents of sparsely populated regions were also excluded from the survey population for operational reasons. Even when combined with the exclusions listed above, this represented no more than 2 percent of the total population, well within the international 5 percent minimum under-coverage requirement. It is estimated that the coverage for the survey was 98.5 percent nationally, with provincial coverage ranging from 95 percent to nearly 100 percent. In the northern territories, reduced levels of coverage (70 to 90 percent) were obtained because only the communities covered in the national Labour Force Survey were included. Table C1 provides the estimated coverage rate by province and territory.
TABLE C1
| Estimated coverage rate (%) |
|
|---|---|
| Newfoundland and Labrador | 98.1 |
| Prince Edward Island | 99.7 |
| Nova Scotia | 99.3 |
| New Brunswick | 98.8 |
| Quebec | 98.9 |
| Ontario | 99.3 |
| Manitoba | 95.3 |
| Saskatchewan | 95.3 |
| Alberta | 98.2 |
| British Columbia | 97.1 |
| Yukon | 90.0 |
| Northwest Territories | 86.0 |
| Nunavut | 70.0 |
| Canada | 98.5 |
Canada was the only country that opted to include adults over the age of 65 in its target population; a liberty that was available as the sample design already satisfied the minimum suggested international sample size requirement for those aged 16 to 65 years.
The most recent Census of Population and Housing, with a reference date of May 15th 2001, was chosen as the frame for the survey. This already existing frame offered the ability to use reported household-level characteristics to identify dwellings with greater probability of containing an individual belonging to specific target subpopulations of interest. This auxiliary information greatly assisted the efficiency of the sample design. Specifically, the survey frame consisted of households enumerated by the Census long-form (20 percent) sample.