Lab 4: Data Uncertainty

What are the rules of Statistics Canada on data suppression in terms of Census Tract (CT) and Dissemination Area (DA)?

Generally, census areas with a smaller population size have data suppression due to confidentiality. Especially if there is only one postal code in an aggregated area. Further, the postal code will not be available to a commercial business (possibly asking for client’s information), only the household’s name is available if permitted by the household.

 

So what are the rules that Statistics Canada tries to follow in the delineation of CT boundaries?

Statistics Canada (2015) tries to follow a 2,500-8,000 person count per delineation of CT boundaries. This census changes every 5 years because of person-mobility; the general population of people living in that area may have moved away, reducing the person count in the area so boundaries have to be moved in a wider frame. Moreover, there might be an influx of new people in the boundary, resulting to a higher population, so the CT boundary needs to move into a smaller frame to make sure there are still 2,500-8,000 within a CT boundary.

 

The following questions are relating to the lab 4 'housing affordability' map uncertainties:

How many have a “0” recorded for shelter or Median value of dwelling?

  • 19 CTs are recorded at “0”

Why do they have values of 0?

  • They have values of 0 not necessarily because these boundaries does not earn money, but rather we don’t have the available information since these are Indian Reservations.

What percent of Metro Vancouver CTs do the ‘0’s represent?

  • 4% of CTs represent ‘0’

How many 2016 Das have 0s?

  • 127 DAs have 0

What percent of DAs does this represent?

  • This represents 3.68%

Why do these DAs have zeros for shelter cost?

  • These DAs have zeros for shelter cost because of the overwhelmingly amount of persons considered homeless and without a stable income who are housed under free housing. Moreover, according to Statistics Canada, “no characteristics or tabulated data are to be released if population of the area is less than 100”, so it’s not because people there are <100 people living in certain boundaries of DTES, but because there might be <100 homeowners in the DTES that there is not enough data to publicize, therefore considered an error.

Share this learning activity with others