8 December 2015

Montana Poll’s authors need to release critical data

In the last paragraph of my report on the Montana Poll, I noted it’s a student project that does not have the resources available to commercial pollsters such as Gallup or Public Policy Polling. There, the error bars are wider.

After reading my post and the poll, a sharp-eyed reader observed:

The poll included both registered and unregistered voters. Most polls are of registered voters.

Women in this sample were far more conservative than most Montana polls I’ve seen.

I couldn’t find any demographic description of the sample. If you don’t know the representation of various demographic groups it’s impossible to determine sample bias.

Those are good points. We can infer the sample breakdown by political party from Questions 14 and 15, the Republican and Democratic Presidential primaries:

The neither Republican nor Democrat category probably includes Libertarians and other third parties, and self-identified independents. Only ten percent or so of voters are true independents, and many of these tend to be low information voters who do not always vote. Many self-identified independents are closet partisans whose true political affiliation can be smoked out through questions such as “Do you lean Republican or Democratic?”

I hope the authors of the Montana Poll will release the demographic data and the crosstabs so that readers can assess the reliability and accuracy of the survey.