Sample frame

Representativeness

A representative sample is critical to quality survey research. Today, the level of Internet penetration in the developed world is very high. Approximately 28.5 million Canadians have online access either at home or at work, with most having access at both locations. Online access in Canada has spread to every significant demographic group, providing deep coverage of seniors, youth, and people with above and below average household incomes, visible minorities and many hard to reach rural populations. Telephone response rates (and the massive decline in land lines) have essentially disqualified that methodology for wide population sampling.

The Angus Reid Forum includes tens of thousands of engaged members — enough people in all major demographic groups to draw randomized samples that represent the population as a whole. Survey samples are randomly selected to carefully reflect the known universe on core demographic variables, socio-economic status and, as appropriate, specific attitudinal or other characteristics as well.

Angus Reid Forum has Canada covered

Arguably, geography is the most important factor of all – perhaps especially so in a vast country like Canada. Angus Reid Forum, Canada‘s best and most representative panel, includes engaged Canadians from coast to coast –- across all 10 provinces, and in every corner of every province. Holding to the adage that “birds of a feather flock together” in matters of the marketplace and the polity, regional coverage and representativeness have always been the foremost considerations for the construction and ongoing maintenance of the Angus Reid Forum.

Our approach uses Federal Electoral Districts (FED’s) as that underpinning building block. FEDs are a strong choice as a basic sampling point for a number of reasons, including:

· They are roughly equal in size (at 100,000 per);

· The (currently) 338 FEDs naturally cover the whole country; and

· Statistics Canada reports detailed population data at this level allowing for very focused panel and sample calibration (on core demographics and on the actual voting profile per FED if relevant).

Conducting Surveys

The first step in conducting a survey is to create a balanced sample matrix of the Canadian population – beginning with the regional-level foundation and including the other core characteristics. A randomized sample of Angus Reid Forum members are then selected to

match this matrix. The selected community members receive an e-mail invitation and are asked to complete the survey by clicking on a link to our secure online survey platform. In order to participate in the survey, these members must log in with their username and password. It is important to note that our panelists do not receive survey invitations too often or repeatedly on the same subject.

Careful monitoring during data collection ensures the targeted sample composition is achieved. In data preparation, statistical weighting may also be applied to ensure the most representative (and trackable) final survey sample as possible.

Respondents receive a small monetary incentive for completing each survey, as well as occasional prize draws. This encourages respondents to participate regardless of surveys’ subject matter. The Angus Reid Forum abides by strict confidentiality and privacy guidelines. The personal information of our community members is never divulged to anyone.