Lies, Damn Lies, and Statistics July 10, 2006 Jill First, friend, the abortion rate is not the same as the unintended pregnancy rate. Second, you’re wrong about the abortion rate, too. But don’t let facts get in the way of your crusade against sex education.
First, Ma’m, I have no crusade against sex education or any other form of education. I will admit to becoming intemperate upon the subject of abortion but I agree that there’ll be no meeting of minds between us upon that. Looking at the situation in the UK (which is what my post was all about) the abortion rate has climbed as sex education has become more common in recent decades. As I commented at the second post you link to above: “OK, if you want to check statistics, try this. What was the abortion rate in, say, 1970? (that’s three years after legalisation in the UK). Do it per cpaita, per fertile woman, however you wish. Take today’s rate. In those 36 years has sex education increased or decreased? (That’s not a difficult one to answer actually). Fine, so what we would appear to be able to show here is a correlation between increased sex education and increased abortion rates. ( I’m convinced that that is indeed what the numbers will show but I’m not a statistician so I’ll leave that to you.) Now, as we all know, correlation is not the same as causation. But it does indeed seem a little odd to note that the correlation is for more of both and then insist that the causation is that one produces less of the other.” If you’d care to explain this to me please do. Type slowly and use short words for I am only a dumb Englishman. (And just to make it clear, no, I have nothing to do with any church or organised movement about sex education, abortion or anything else of a remotely similar nature.)
He would have been right if it was the US. The number of abortions has fallen quite a bit. The more interesting question here in the US would be whether or not the decrease in abortion was caused by a decrease in unintended pregnancies or in a decrease in the availability of abortion (or some combination of the two–which is what I think.) I know the number of teen pregnancies has also fallen (most of which I am assuming are unplanned/unintended). What I do not know about is whether the number of unintended pregnancies has gone down for older women.
One of the first rules in analyzing time-series data is that spurious calculations are the norm, not the exception. This has been illustrated many times by correlating the price or consumption of A with the price/consumption of B, where A and B are items selected at random. Somebody offering up such a naive analysis shouldn’t really be judged as a valid analyst.