even steven – I really would recommend that you take a basic stats course – this is an elementary issue that every social scientist gets drummed into them. If you would like to see a parallel demonstration (e.g. a discussion of why the fact that rich states tend to vote Democratic doesn’t mean that rich people vote Democratic, click through the link in the post and read the discussion of Gelman et al. Or even better, buy their book.

http://www.tlavideo.com/support/supportOption.cfm?v=3&sn=1&supID=34

You say: (Hint: think about what this ‘individualized data’ can or cannot tell us)

What the individualized data cannot tell us, and I believe this is Dendroica’s rather obvious point, is that Al Gore ran for president in 2004. If it does, in fact, tell us that then what you have there is corrupted data. The fact that that particular mistake is so prominent makes me wonder about the author’s attention to detail in other realms.

Online porn is just a few mouse clicks away from anyone with a broadband connection and can be downloaded in the privacy of one’s own home without the neighbors’ knowledge.

should read

Yes, it is POSSIBLE that there are a subgroup of non religious people in those states who are buying the porn, but you haven’t offered any evidence for that, so until you do, the study still stands to support the idea that religious people consume more porn.

even with the mistake, you still have to refute my argument.

Really. All we have is the data that indicates the higher porn rate in the set of religious people as a whole, which, all other things being equal, indicates there is a higher likelihood that each individual consumes more porn. Yes, it is POSSIBLE that there are a subgroup of religious people that there are non religious people in those states who are buying the porn, but you haven’t offered any evidence for that, so until you do, the study still stands to support the idea that religious people consume more porn.

Then you CHERRY-PICK an example where your ‘ecological argument’ is so compelling, wow—that’s impressive. You don’t think there’s a thousand examples where the aggregate data is higher because the individuals are higher. Oh my god, I know some tall woman, so it must be that the fact that men tend to be much taller than women must be FALSE.

Then you say

“Similarly, even if people in more religious or Republican states were more inclined to purchase porn online, this doesn’t imply that religious people or individual Republicans were more inclined to purchase porn online, and I can think of at least two or three plausible alternative causal mechanisms that would explain the observed correlation.”

Let’s hear ‘em. In the absence of those arguments. you don’t have a leg to stand on.

(Hint: think about what this ‘individualized data’ can or cannot tell us)

Nice error, you’ve got there!

Harvard’s Benjamin Edelman latest to be suckered by AVN’s $12B/year Figure
http://www.comstockfilms.com/blog/tony/2009/02/25/harvards-benjamin-edelman-latest-to-be-suckered-by-avns-12byear-figure/

When I read the paper and looked at the top 10/bottom 10 states for each measure (subscriptions per capita, per internet connection, per broadband connection, etc.) the thing I was struck by was that in several of the lists, staunch Republican states seemed concentrated at the top and bottom. I think Utah was top on every list, and the bottom in various lists was West Virginia (which jumped to the top in some lists) or Idaho. I seem to recall that California and New York were nowhere in the top or bottom 10 on any list, though Massachusetts was low in the top 10 on one of them.

And then of course, the other obvious explanation for this trend is that non-church-goers are simply more likely to do less anonymous things, like go to a strip club or buy a magazine at a gas station.

But my main reaction to this has been “so what”? I personally know people who go to church every sunday and don’t think there’s anything wrong with porn. Just because a significant group of church-goers consume porn, doesn’t mean another significant group of church-goers abjure it and criticize others for their support of it. So the temptation to conclude hypocrisy comes up against the fact that groups are made of individuals who make their own choices. Until we see that individuals who decry the evils of porn on religious grounds are also very likely to consume it, you’ve got nothing.