Complicated Polling Questions: Handle With Care!

May 10th, 2013 at 10:15 pm

Just a brief note on this polling result which struck me as potentially indicative of why you shouldn’t trust polling results on complex issues, in this case, how people feel about whether their state should accept the ACA’s Medicaid expansion.

See slide #11, “Few Minds Changed by Argument on Medicaid Expansion.”  The poll asks a pretty weedy question about whether your state should expand Medicaid to low-income, uninsured adults, given the Supreme Court decision that states are not required to take up the offer.  It finds that 41% of respondents say “no,” leave Medicaid as it is.

Then, if I’m reading this right, it offers those negative respondents a bunch of information about how the cost of the expansion will be picked up by the Feds and how without the expansion, a bunch of people won’t get coverage, and re-asks the question.

And according to that result, half of the people who responded “no” in the first place meant to say “yes” in the first place.

It’s just one poll, one example, and one particularly gnarly sequence of questions.  But it reminds one to take such results with a large grain of salt.

Print Friendly, PDF & Email

4 comments in reply to "Complicated Polling Questions: Handle With Care!"

  1. Nick Batzdorf says:

    And the multiple-choice questions in these polls invariably have problems, whether it’s being overly simplistic, leading the answer, missing the most salient points, simply forcing a black/white answer to a grey issue, and so on. In this case I’d also worry about code language in the question (avoidable or not) and also the way it’s framed – “low income,” “states not required”…

    The worst example I’ve heard was on a radio program a few years ago when the guest from Pew was talking about a poll they’d taken. It asked people whether they had a favorable or unfavorable impression of various races/religions (successively). Well, the correct answer is ” Are you freaking serious?! What a #$)#$!!!!! question!” But I don’t think that was one of the multiple-choice answers; instead people took the poll as permission to be racist.

  2. Arnold Ruff says:

    Jared, I have opened this site to get the word out how far our educational system has failed us. We have a “One-Sided Educational System. Check out; Most of our problems 90% stem from lack of education and mostly “One-Sided Education”. Let’s give our children a choice to life.


  3. Jill SH says:

    Wasn’t there a point a while back when only 40% or so were in favor of Obamacare , while 60% were against it? But what was really true was that 30% wanted single payer, not O-care; the other 30% wanted no reform at all.

  4. PeonInChief says:

    One of the stories that came out of the decision by some Republican governors to reject Medicaid expansion was that counties in those states wanted permission to sign up whether or not the state did. That’s because the burden of medically-indigent adults falls on county hospitals, while Medicaid co-payments fall to the state government. Not expanding Medicaid meant that the state expenditures wouldn’t grow, dumping the costs on the counties. (This was particularly evident in Texas, where some counties asked if they could sign up independent of the state government decision.)