> > > > Cheese is a cheeseburger topping, because I am dense. -41 (-0.3%)
> > > > Cheese is not a cheeseburger topping, although it may be considered a hamburger topping, because I have a clue. 10646 (100.4%)
> > >
> > > The fact that the total is 100.1% is simply a rounding error due to the fact that they're only given to one decimal place.
> >
> > The -.3% should be -.4%
> > Rounding error happens quite rarely. Usually, it's the statistic that's messed up.
>
> Okay, I'm not penetrating this. If that's NOT a rounding error, in what way is the statistic messed up?

I wasn't talking about the rounding error. I was talking about how it's technically impossible for more than 100% of those polled to answer the question. In other words, the total number of people who answered "Cheese is NOT a cheeseburger topping, it is a hamburger topping" is greater than the number of people who answered the question in the first place. It's also quite impossible for fewer than zero people to give a certain answer ("Cheese is a cheeseburger topping, because I am dense"). Sam, what gives?