Public polls projectthe business of reading our minds
The state of public polls in Canada tells us a lot about the state of public debate about issues in Canada. It also informs our understanding of the nature of the public opinion business. The on-going analysis of the publicly available polls in Canada is provided by Jenkins Research.
Public opinion polls, whether one likes them or not, are critical ingredients of public discourse.
Lessons from the Canadian Public Polls Project
Public opinion polls, whether one likes them or not, are critical ingredients of public discourse. There is no arguing that they provide our companies, not-for-profits, and democratic institutions a peek into the minds of Canadians. This in turn helps them affect or respond to social, economic and political events.
Polls are ubiquitous. The stats below are somewhat mind boggling. And, these are only the ones that are publicly released.
It turns out that the Public Polls Project provide a window on the topics that are dominating public discourse and where the public falls. It also shows some interesting facts about the methodologies that polling firms are using (and some of the implications of them)
# of Releases Per Month
Questions reported on
# of questions per release
How We Built the Database
In April 2016, I began tracking all of the public opinion polls that were being released in Canada. The result is the Public Polls Project, which is the only systematic review of public opinion in the public domain.
Multiple strategies are used to identify public opinion releases. These include Google searches, a review of press releases on the newswire and by going to the research firm websites.
Each time a survey is identified, the field dates and research firm are compared against existing surveys in the database for duplicates. In some cases, the same survey has more than one release. For example, an omnibus survey might have multiple clients and/ or topics. As such, we separate poll releases from unique surveys.
- Unique releases of survey results are identified not every mention of public opinion. In the search process, mentions of poll results that do not appear to be a unique release of a survey result are not included.
- Only surveys of Canadians are included. Surveys of special populations of the public, seniors, women, young people are included. B2B surveys are not included. Surveys of people who are on proprietary lists (e.g. work sites or travel sites) are also excluded.
- Non-survey methodologies are not included.
- Government of Canada surveys, which are archived on the internet, are not included unless they garner public attention.
Best efforts are made to identify releases in the correct month. Numbers are updated for previous months if new poll results are identified that correctly should be allocated to those months.
Each unique release is coded based on a number of variables. These include the topics (specifically and in general categories), the sponsors, the modes of data collection, the sample sizes and other survey characteristics.