How do we interpret the latest offering from Quinnipiac University showing Donald Trump and Hillary Clinton in a dead heat?
For those who do not know me or what I look like, allow me to offer you a brief description. I am about 6” tall, have killer abs, a full head of hair and possessed with disarmingly good looks.
Think George Clooney meets David Beckham.
Actually, that’s what I would LIKE to look like. But I don’t.
Think Michael Smerconish meets, well, someone who looks a lot like Michael Smerconish.
That’s the fundamental problem with the latest Quinnipiac poll. Respondents were not selected – or balanced – by what they ARE but by how they would like to be.
Q: “Are you a registered voter?”
A: “Uh, sure. Yeah, that’s the ticket.”
Q: “Generally speaking, do you consider yourself a Republican, a Democrat, an Independent or what?”
A: “Let’s see, today I feel like an Independent.”
Here’s the problem with that, particularly if you want to use this poll for predictive purposes. We know only about 75 percent of eligible adults even register to vote. So, in theory, 1 in 4 adults – those randomly dialed in this poll – aren’t able to vote.
But if they claim to be a voter, they are included. And that’s a real problem.
Very few respondents not registered to vote will admit to that fact. It’s called “socially desirable response patterning” – people will often answer the way they think they should answer. (It’s also why so many people watch Masterpiece Theater – or at least DID until Nielsen started tracking by interactive and electronic methods.)
Further, while there is value in knowing what people “consider” their party affiliation (and we test that question all the time in our polls), asking them to self-identify causes a host of other problems.
Let’s pick just one.
According to this poll, 37 percent of respondents belong to neither major party. THAT WILL NOT HAPPEN in the 2016 presidential election. “Independent” and “other” party voters will hover between 23 to 24 percent of the electorate. Based on this variable, one can easily see that the response set in this poll is highly skewed by as much as 40 percent.
So, how do you fix this?
Simple. You draw respondents from the active voter file and know what their registration is before asking. (it’s a little more complex than that, but you get the idea.) This way you can model your sample to look pretty close to what the electorate is most likely going to look like and you are not dependent on people telling you what they think you want to hear.
The folks at Quinnipiac did go through some pains to ensure bilingual respondents, chose a relatively robust cellphone sampling and did a great job calling and re-calling households … so kudos there.
However, based on the fact that respondents were asked to self-identify both the fact that they are registered to vote and then their party preference – and there is no indication that either of these vital factors was verified – we have to recommend taking this poll with a half-shaker of salt.
Key for the Salt Shaker test:
No salt needed: Solid pollster, solid methodology, and the sample appears to be nicely balanced.
A grain of salt: The poll has one or two non-critical problems and should be taken with a grain of salt.
A few grains: There are several concerns with how the poll was conducted, but not enough to throw it out entirely.
A half shaker: There are enough problems with the methodology to warrant serious concerns, and the poll should not be taken seriously.
A full shaker: The poll has so many problems it should not only be completely disregarded but pollsters receiving multiple “full shakers” will no longer have their polls covered by Florida Politics/SaintPetersBlog.
Steven J. Vancore is the president of Clearview Research. With a master’s degree in Marketing Communications from Florida State University, he has nearly 30 years’ experience conducting polls and focus groups throughout the state. He serves as an adjunct instructor in the Masters of Applied American Policy and Politics program at FSU. He can be reached at email@example.com.