Page 278 - Special Topic Session (STS) - Volume 2

STS493 Stéphane D. et al.
The best example of an effective new communication strategy is the one
used for the 2016 Census of Population. Based on results from the previous
Census of Population, the population of Canada could be divided into five
different groups, each with its own unique communication strategy. The
groupings were based on the likelihood of a fast response to the census. One
group, the easiest to reach, got relatively light communication. People who are
more difficult to reach, on the other hand, received communications at various
stages of collection. This segmentation strategy is an important reason why the
2016 Census of Population was considered the best ever in Canada, with the
highest response rate on record, and with an impressive cost and quality
performance.

3. Phase 2: Experimenting with new data collection methods
As mentioned earlier, focusing only on optimizing Statistics Canada’s
current collection operations would be insufficient. New ways of collecting data
must be explored to reflect the new reality of a population less interested in
completing surveys and to take advantage of technologies now available that
could transform primary data collection operations. This section presents
Statistics Canada’s research focus areas, considering the anticipated
operational implementation feasibility.

Developing a crowdsourcing service
Crowdsourcing has been an early success in the introduction of new
primary data collection techniques. Crowdsourcing involves asking the
population to proactively provide information rather than wait to be contacted
when selected as a respondent. The risk of such an operation is obvious to the
statistician— crowdsourcing data quality is difficult to assess, with metrics on
quality near-impossible.
Nevertheless, Statistics Canada began to experiment with crowdsourcing in
areas deemed relatively low-risk. The technique was first used for a project to
improve the available information about dwellings in Canada. The “crowd” was
asked to provide GPS locations for a set number of dwellings posted on the
Statistics Canada website. That drew considerable interest from the population,
who provided the requested information faster than expected. Secondly,
Statistics Canada crowdsourced the price of cannabis through its StatsCannabis
web application in the months before the legalization of cannabis in Canada in
fall 2018, when cannabis consumption was still illegal (except for approved
medical use). This resulted in over 20,000 entries to the questionnaire, and
reasonable price estimates (i.e., within expectations, upon validation).
Subsequently, Statistics Canada implemented a crowdsourcing service
within its survey operations branch. The service is relatively simple—a short e-
questionnaire, accessible to all website visitors (i.e., there are no barriers such

267 | I S I W S C 2 0 1 9