Yesterday, The Big E went over the results of the State Fair poll on the Anti-Family Amendment as it pertains to the task ahead.Today, I will use the same poll to point out the uncertain nature of the outcome of that task and why organizing will be the key to winning.
First, though, the results of the poll, conducted by the Minnesota House of Representatives, one more time.
"Should the state constitution be amended to define marriage as "only a union of one man and one woman?" This question will be on the November 2012 ballot."
Yes 29.8%
No 66.5%
Undecided 3.7%
What you see is that unlike with normal polls, this poll has no margin of error (but it has insignificant digits!). That's because it's actually not a poll at all, it's a straw poll.
There are two key differentiating factors between the State Fair poll and a poll from, say, Gallup; the nature of the sample and weighting.
At the State Fair anybody can just walk up and take the poll. In a traditional survey every effort is made to get as random a sample as possible. And by random I mean everyone who makes up the sample frame should theoretically have as equal a chance of being a respondent as anyone else.
Additionally, even though you go to every effort to get as random a sample as possible, the final result is often a little different from the actual demographics of the sample frame. To alleviate this, pollsters will weight their samples based on immutable characteristics like age, gender and race. The State Fair poll doesn't do any of these things.
So, the State Fair (straw) poll violates two of the most fundamental aspects of traditional public opinion polling, but just because it's a completely unscientific survey doesn't mean we can't do a little bit of analysis right?
