Massaging the Data

I feel it is my job, as a scientist and a sane human being, to point out everything that is wrong with Maggie Gallagher. She has an article in the NRO this week calling for more money to help "protect" marriage. Pretty much it's the same old crap. But I'd like to point out two small issues I have. One is with statistics. She says:

The most striking (and underreported) results are those of the 2004

UCLA freshman poll released earlier this year, which surveys 290,000

college freshman. Between 2003 and 2004 the proportion of college

freshman who support gay marriage dropped almost three percentage

points, from 59.4 percent to 56.7 percent. This is the first recorded

drop in support for same-sex marriage among college freshman since the

question was first asked in 1997.

Well, aside from the fact that I am completely unable to find this poll on the internet, we should look carefully at those numbers. I'm no mathematician, but even with 290,000 people surveyed, I know of no drop of 3% that can be considered statistically significant; the margin of error doesn't improve that dramatically with a greater sample size. So maybe the reason it's underreported is that it isn't terribly striking. Even all of the other polls she cites as showing that Americans are increasingly opposed to gay marriage aren't that dramatic.

The Pew poll, which asks “Do you strongly favor, favor, oppose, or

strongly oppose allowing gays and lesbians to marry legally?,” showed

Americans’ opposition to SSM climbing from 53 percent v. 38 percent in

July 2003, to 60 percent v. 29 percent in the latest August of 2004

survey.

Notice how the question specifically asks for one of four choices and yet she lumps them into two. It be interesting and important to know how committed the citizenry is to their opposition. These are minor points, but it's the subversive ways that people use numbers and facts to support their positions or make small changes seem significant, that are subtly manipulative.

And then there's an interesting omission at the end of the article:

Two years from now, one-third of the country is likely to be living

with gay marriage. Pending court decisions in California, Washington

state, New Jersey (along with Massachusetts) are likely to produce a

fragmented marriage system despite overwhelming public opposition. And

other states, like New York, are taking a different route: forbidding

the performance of gay marriages in-state, but recognizing gay

marriages performed in nearby Massachusetts or Canada.

Where is Connecticut, which just approved civil unions that are significantly close to gay marriage, and did so legislatively? I'm hard-pressed to believe that Maggie Gallagher missed that little development. No, the reason that Connecticut isn't mentioned is because it doesn't fit into her nice little model of judical tyranny and forced acceptance. It doesn't matter that some of the country might actually want to be living with gay marriage. Nor does it matter that public opposition to gay marriage in Massachusetts is waning and that the citizens there are not punishing gay marriage-backing legislators but rather rewarding them. Because for people like Maggie Gallagher, the voice of the people is sacred and absolute, but only as long as the people are agreeing with your position.