PODCAST:
June 22, 2026 ~ Chris Renwick and Lloyd Jackson spoke with Steve Mitchell, President and CEO of Mitchell Research and Communications. They discussed criticism of his poll, specifically about cross-tabs and open link polling.
MICHIGAN ~ A pair of new political polls released last week caused an uproar from Michigan campaigns, claiming the surveys should not be considered accurate.
The surveys in question were conducted by Mitchell Research & Communications and aimed to gauge public opinions on candidates in the GOP primary race for governor and the highly competitive Democratic primary for Michigan’s open Senate seat, according to an article from the Detroit News. The surveys were recorded through an online SurveyMonkey poll, sparking criticism that the link to the survey could be shared between supporters of certain campaigns, thus manipulating the results.
Steve Mitchell, president and CEO of Mitchell Research & Communications, responded to the criticism, claiming his polling group has protocols in place to prevent this kind of manipulation.
“We’ll send out 100,000 to 150,000 text messages, which is to about 1% of the whole state of Michigan,” Mitchell told WJR. “But we send them out in small batches. We’ll send out, say, 25,000. Those 25,000 (people) have link one, and link one will go directly to a survey. The identical survey as link two will go too, but it’s a completely different survey. In other words, the results in link one will be separate from the results in link two, link three, link four, and link five. So if we have four links that show that John Jones leads Mary Smith by 10%, and the fifth one shows that Mary Jones leads John Jones, we presume that somebody may have tampered with that fifth one, and we simply don’t use the data.“
