Wouldja Like to Take a Surrrveeey?!

Do you like George Wendt? Five points to Gryffindor if you know where that's from.

I just received a call with an automated survey from the National Organization for Marriage. If you can't tell just by the name, this is a right-wing, Christian organization that of course believes gay marriage is evil and, somehow, a threat. I'm getting the call because Lincoln Chafee, Rhode Island's governor-elect, has made it clear that he supports gay marriage, so there's a lot of attention being paid by American religious groups to our tiny, little nugget of the union. It doesn't help that Rhode Island is so damned Catholic.

Regardless, there's a fantastic episode of Penn & Teller's Bullshit called Numbers, you can watch it here. It's a Hong Kong site called Megavideo, which is a little scummy, but it's at least safe. The important part comes in at about 12:20, if you want to fast forward.



The survey that I answered, and I'll try to be as accurate as I can, opened the survey, opened the survey, with the statement "This survey involves one of the most important subjects of our time: the institution of marriage." Nice primer, there, assholes. Look at what they said. This is IMPORTANT. We're talking about an INSTITUTION.

Then, the money question "Do you believe that marriage should be legal only between a man and a woman?" It's worded juuuust funny enough to possibly elicit a particular answer. Why not just ask the most straightforward question? "Do you believe that gays and lesbians should be allowed to marry?"

Because asking that question emphasizes that we are not allowing a group of people to do something. We have a rather strong freedom streak running through us, and pollsters want to avoid asking questions that collide with it. They phrase the question not in a way that makes it sound like we're preventing people from doing something, but that we are protecting something. That we are, in that protecting, preventing people from doing something is glossed over.

If you've watched the episode of P&T, you'll understand that this is not a merely semantic issue. It is a fundamental issue in the posing of the question. By phrasing it as though we are defending something and not as though we are oppressing a group, and priming people with the concept of "the institution of marriage" beforehand, the pollsters can easily shift the results of the survey in the direction that they want.

They then asked whether I was male or female, and whether I was over the age of fifty. This is likely important because men are shifting towards a pro-GLBT position faster than women, and there is a startling opinion shift in the mid-40's. People younger than 45 are strongly pro-gay, and people older than 45 are strongly anti-gay. Damned old people. Just die.

This also makes me question the timing of the call. I work out of my home, so I'm here, but calling on a weekday, at 2pm, when most people are at work; the pollsters are likely to get a disproportionate number of women and old people answering. Miiiiighty suspicious. It reminds us that the sample quality is important for any survey. Is it actually random? Or is there some other variable that can be affecting the qualities of the sample? And, even more perniciously, are the pollsters well aware of this and using it to their advantage? Judging from the wording of the actual survey, I'm quite sure that they know what they're doing in calling at 2pm.