Survey of Electors Following the April 3, 2017, By-elections in Calgary Heritage (Alberta), Calgary Midnapore (Alberta), Markham–Thornhill, (Ontario), Ottawa–Vanier (Ontario) and Saint-Laurent (Quebec)
- Phoenix Strategic Perspectives was commissioned by Elections Canada to conduct a survey to assist in evaluating the April 3, 2017, by-elections held in the federal districts of:
- Saint-Laurent (Quebec)
- Markham–Thornhill (Ontario)
- Ottawa–Vanier (Ontario)
- Calgary Midnapore (Alberta)
- Calgary Heritage (Alberta)
- This public opinion research was based on a 10-minute dual-frame (landline and cell phone) random-digit-dial telephone survey, conducted from April 7 to 24, 2017, with a total of 2,000 eligible electors (400 interviews per riding).
- The respondents were Canadian citizens at least 18 years of age on polling day who were residents of the electoral district (i.e. had an address of ordinary residence in the electoral district) from the first day of the revision period until election day. A profile of survey respondents can be found in Annex 1 of this report.
- Based on a sample of this size, the overall results can be considered accurate to within ±2.2%, 19 times out of 20. The results for each of the electoral districts can be considered accurate to within ±4.9%, 19 times out of 20.
- As is often the case with telephone surveys of the general public, younger electors (those under 35 years of age) were under- represented in the final survey sample and older electors (those 55 years of age and older) were over-represented. The survey weights corrected for this issue of disproportionate representation of age in the survey sample. This serves to reduce bias resulting from survey non-response should it be present, but it does not necessarily eliminate it. If electors under 35 who responded to the survey differ in their views on the issues addressed in the study from those under 35 who did not respond to the survey, the age bias (if there is one) might have been magnified. It is not known, however, whether there is any such difference.
For a more complete description of the methodology, please refer to the methodology note available under separate cover.