I have a little longer than that.
They are drawn from an online panel of approximately 130,000 to 140,000 Canadians. These are people who, yes, self-select into the panel, but that doesn't necessarily mean that they will all be asked all the same questions or be invited to participate in every survey. There is an element of randomness to the way selection is done.
We are able to control to ensure that the sample is drawn from and balanced to census data. On income levels, on age, on gender, on region, on first language, on education levels, and all the main basically census demographics we are able to balance the sample to ensure that the people we're speaking to and those who are responding to us do represent and reflect the Canadian population as a whole. When we talk to people who are dealing with income levels lower than $50,000 or higher than $100,000, they are reflective of the actual totals within the population.