Concepts in Public Opinion Research

Statistical Weighting

Weighting factors are used in sampling to make samples match the population. For example, let’s say we did a survey and 41% or respondents were female and 59% were male. We know from census data that females should make up 51% of the population and males 49%.  In order to model a more representative sample, we might add a little more “weight” to data from females.

Rather than dispose of additional surveys we can include the views of more respondents while balancing the overall responses to be more representative of the Canadian population at large.

When reputable pollsters apply weights to data they always disclose this. In Canada it is common industry practice to be transparent about methodology and to delineate between weighted and unweighted results.

An example of how a gender weight is calculated

To calculate how much weight you might need for a gender weight one would divide the known population percentage by the percent in the sample.

  • Known population females (51) / Sample Females (41) = 51/41 = 1.24.
  • Known population males (49) / Sample males (59) = 49/59 = .83.

Leading Questions and bad answer choices

It can be unsettling thinking that your opinions will disappear into a statistical model of the greater population, but it's especially frustrating to feel pigeon-holed into predetermined answers. It is the nature of our particular profession that we report on thoughts and opinions in aggregate. While it's often both useful and necessary, to present to you a finite array of options, it's never our goal to dismiss or misrepresent your opinion. If you encounter a question with answer variables you think are incomplete or egregiously misrepresent your thoughts then let us know. We're quite happy to get feedback on our questionnaire design.

Why do you need personal information?!

Unless you join our Pollara panel, we don't need personally identifiable information like an exact address or email address. But it is often very important to determine how your approximate location, age, gender and income levels affect your view. For instance, without your full postal code, it's not possible to know what electoral district you are in. Without knowing your income, it would be impossible to determine if billionaires and working-class people, say, had the same perception of service at their local hospital. We do not, however, present data to clients or the public with your personally identifiable characteristics visible and we don't have you in a database somewhere - and we are not building a profile about YOU. Each surveys data is stored separately and worked on in aggregate.

About Us

WHO ARE WE

Craig Worden

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Vestibulum ultricies sollicitudin ante nec mattis. Mauris eget ornare nibh. Nam venenatis porta diam scelerisque suscipit. Donec in vehicula ligula. Etiam efficitur quam ex, at consectetur magna porttitor at. 

Lesli

Mauris eget ornare nibh. Nam venenatis porta diam scelerisque suscipit. Donec in vehicula ligula. Etiam efficitur quam ex, at consectetur magna porttitor at. Nullam convallis, eros a vulputate tincidunt, mauris lorem tincidunt sem, sed cursus neque sem a est. Nam malesuada ipsum orci, et congue ligula fringilla