Methodology

The Minnesota Community Survey is a probability-based online panel that is representative of the state

​

We recruited people to our panel by mail using an address-based sample (ABS) of households included in the U.S. Postal Service (USPS) computerized delivery sequence file (DSF). We then identified census block groups (CBGs) with large shares of high-priority populations and sent these areas more mailers than elsewhere. 

​

Our methods also reduce fraud. For example, since we send invites by mail, we can be reasonably certain that the person taking the survey lives at the address we sent the mail to.

​

Why we like ABS

​

We like address-based sampling because the address list maintained by USPS covers 90-98% of the U.S. population. This ensures nearly everyone in a state has a chance at being invited to join a Lumaris Community Survey.

​

It also allows us to oversample census blocks with high percentages of traditionally underrepresented groups, including Black, Hispanic and Asian adults. In addition, we sometimes use commercially modeled racial and ethnic variables that are attached to addresses in the DSF.

​

Why we like probability panels

​

Conducting surveys on a probability panel is usually more affordable than one-off probability surveys with cross-sectional designs (i.e. phone or mail surveys). At the same time, probability panels allow us to maintain high standards of research method quality.

​

A notable strength of probability panels like ours is accurate measurements of smaller populations, such as young people or Black, Hispanic and Asian Americans. Opt-in surveys (or non-probability surveys) are increasingly popular but often struggle to accurately measure smaller groups. In fact, opt-in surveys tend to have twice as much error as probability surveys and can produce misleading results on sensitive issues.

​

This is a high-level summary of our methodology. If you'd like more details, please contact us.