An interesting news story I heard on the radio the other day underscored something that I’ve long known since taking my PoliSci 300 class in Michigan with Professor Greg Markus. How you write a survey question can greatly influence the answers you get (and what your polls show as a result).

In addition to the normal examples that they gave, they gave an example of how not to write a survey question that I hadn’t heard before. Survey question: “Have you stopped beating your wife?” If you answer “Yes”, the assumption is you were beating your wife previously. If you answer “No”, then the assumption is that you are still beating your wife. The smart ones are the ones who I guess have no opinions on that survey. It seems like that would be the only “correct” answer....

So when listening to polls on *anything* for the upcoming election, find out how the question was asked. Or, if you’re like me, ignore the polls and vote the way you would anyways! It’s *your* vote. You don’t get paid for voting with the winning side.

And as a side note, not that I think anyone would ever *find* this entry looking for recommendations on PS300 at the University of Michigan, but that class was one of the best classes I took while being an undergrad at Michigan. Not only was it interesting and insightful, but the fact that grading was on a 100 point straight scale (meaning, 90+ was an A, 80+ was a B, etc.) and that you only had a chance to get 100 points in a 15 week semester really made the phrase “Making sure you dot your i’s and cross your t’s” really relevant.