Oof
A hell of a correction from the AP:
LONDON (AP) — In an Oct. 11 story about a study examining global abortion trends from 1995 to 2003, The Associated Press erroneously reported that nine out of 10 women will have an abortion before age 45. Researchers at the Guttmacher Institute in the United States and the World Health Organization calculated that an average woman would have 0.9 abortions in her reproductive lifetime (between the ages of 15 and 44), given currently prevailing rates. The figure was arrived at by combining higher abortion rates in some areas and lower abortion rates in others; some women have multiple abortions and others have none. The rate is an average and it does not mean that nine out of 10 women worldwide have abortions.
Think positive
Derek Lowe dives into a problem that is far too poorly understood by most of the public: the problem of false positives.
The news of a possible diagnostic test for Alzheimer’s disease is very interesting, although there’s always room to wonder about the utility of a diagnosis of a disease for which there is little effective therapy. The sample size for this study is smaller than I’d like to see, but the protein markers that they’re finding seem pretty plausible, and I’m sure that many of them will turn out to have some association with the disease.
But let’s run some numbers. The test was 91% accurate when run on stored blood samples of people who were later checked for development of Alzheimer’s, which compared to the existing techniques is pretty good. Is it good enough for a diagnostic test, though? We’ll concentrate on the younger elderly, who would be most in the market for this test.The NIH estimates that about 5% of people from 65 to 74 have AD. According to the Census Bureau (pdf), we had 17.3 million people between those ages in 2000, and that’s expected to grow to almost 38 million in 2030. Let’s call it 20 million as a nice round number.
What if all 20 million had been tested with this new method? We’ll break that down into the two groups – the 1 million who are really going to get the disease and the 19 million who aren’t. When that latter group gets their results back, 17,290,000 people are going to be told, correctly, that they don’t seem to be on track to get Alzheimer’s. Unfortunately, because of that 91% accuracy rate, 1,710,000 people are going to be told, incorrectly, that they are. You can guess what this will do for their peace of mind. Note, also, that almost twice as many people have just been wrongly told that they’re getting Alzheimer’s than the total number of people who really will.
People look at tests with small error rates--a false positive rate of, say, .5%, and conclude that if they test positive, that means it's overwhelmingly likely that they have the disease. But this is true only for conditions that are relatively frequent. Take a test for a disease that has a false positive rate of 5%, and a disease prevalence of 1 in 1000--lupus, say. If you test positive in a random assay, what are the odds that you actually have the disease?
Most people--even, apparently, a shocking number of doctors--would say that the odds are 95%. But this is all wrong. If you test 1,000 people for lupus, 1 of them will correctly test positive for lupus--and 50 of them will falsely test positive. The chances are only 1 in 51, less than 2%, that you actually have the disease.
These are in fact the actual numbers for anti-nuclear antibody tests and systemic lupus, at least as relayed to me by my immunologist after I got a borderline positive result on a screen. These suggest that no one should ever do a random ANA; the information it gives is garbage, particularly since they don't treat lupus until you manifest symptoms. Yet lots of doctors, including mine, do.
Department of awful statistics
I saw this on Feministing and thought: huh?
A new study by the Guttmacher Institute and the World Health Organization shows that abortion rates are similar in different countries whether the procedure is legal or not. Shocking, I know. Of course, what wasn't similar was the risk to women's health.
The study indicated that about 20 million abortions that would be considered unsafe are performed each year and that 67,000 women die as a result of complications from those abortions, most in countries where abortion is illegal.
Moral of the story? Safe, legal abortion is the best bet. Always.
This is a common meme among feminists; indeed, I myself was, in the long fled days of my youth, guilty of propagating it. I'm not quite sure why I thought that abortion was a magical exception to the rule that when you make something much harder and more costly to do, fewer people do it. In fact, the source for this oft-repeated claim turned out, when last I examined to to be fairly awful. This new claim comes with a better pedigree, but not, alas, noticeably better evidence. Here's what the New York Times has to say:
A comprehensive global study of abortion has concluded that abortion rates are similar in countries where it is legal and those where it is not, suggesting that outlawing the procedure does little to deter women seeking it.
Moreover, the researchers found that abortion was safe in countries where it was legal, but dangerous in countries where it was outlawed and performed clandestinely. Globally, abortion accounts for 13 percent of women’s deaths during pregnancy and childbirth, and there are 31 abortions for every 100 live births, the study said.
The results of the study, a collaboration between scientists from the World Health Organization in Geneva and the Guttmacher Institute in New York, a reproductive rights group, are being published Friday in the journal Lancet.
“We now have a global picture of induced abortion in the world, covering both countries where it is legal and countries where laws are very restrictive,” Dr. Paul Van Look, director of the W.H.O. Department of Reproductive Health and Research, said in a telephone interview. “What we see is that the law does not influence a woman’s decision to have an abortion. If there’s an unplanned pregnancy, it does not matter if the law is restrictive or liberal.”
I can't see the Lancet study, which is gated. But the summary does not back up this claim. The study says that abortions are generally high in the developing world, where it is usually illegal, and low in the developed world, where it is usually legal. It also tells you that abortion is relatively unsafe in the developing world.
But it seems mad to extrapolate this to a blanket statement such as "Law does not influence a woman's decision to have an abortion." For one thing, we know of cases where the law absolutely and indisputably did exert such an influence, such as Communist Romania, where abortion bans caused the birth rate to soar. For another, societies where abortion is illegal are probably different from societies where abortion is legal in other ways, such as attitudes towards birth control. Also, enforcement of laws varies even when the laws don't (abortion was technically illegal in Germany for most of the post-war period). And finally, since the variation is almost entirely among developed countries where access to birth control may be spotty for economic, political or social reasons, this would not necessarily tell us much about developed nations. As I understand it, most abortions in America are obtained by women who have had more than one abortion, which seems to indicate that for at least some segment of the population abortion is a substitute for birth control, rather than birth.
Similarly, saying that "making abortion illegal doesn't reduce its incidence, but only makes it more dangerous" is nonsense on stilts when the comparison is largely between developed countries with legal abortion, and developing countries with illegal abortion. Having an abortion in Burundi would be more dangerous than having one in America even if their government legalized the procedure, made it free, and awarded a medal and a complimentary fruit basket to every woman who had one. I am pretty sure that abortion, like almost every other activity, gets more dangerous when it is legally prohibited. But from what I can make out, this study doesn't do a good job of demonstrating that truism.
Cross-country comparisons--what statisticians call latitudinal studies--are fraught with difficulty because of all the differences in law, enforcement, data collection, social norms, political culture, health care systems, and so forth. That's why it's important to also look at longitudinal studies--studies that examine the same place over time. And all the reputable studies I'm aware of, which to be sure are not an exhaustive list, show pretty much the expected result: if you legalize abortion, you get more of it.
In America, especially, the evidence that legalizing abortion resulted in more abortions seems pretty rock solid. Steve Levitt wrote about this in Freakonomics, and I blogged an excerpt some time ago:
In the first year after Roe v. Wade, some 750,000 women had abortions in the United States (representing one abortion for every four live births). By 1980, the number of abortions had reached 1.6 million (one for every 2.25 live births), where it levelled off. . .
To be sure, the legalization of abortion in America had myriad consequences. Infanticide fell dramatically. So did shotgun marriages, as well as the number of babies put up for adoption (which has led to the boom in adoptions of foreign babies). Conceptions rose by nearly 30 percent, but births actually fell by 6 percent, indicating that many women were using abortion as a method of birth control, a crude and drastic sort of insurance policy.
I'm still in favor of legalizing it, of course, for moral and practical reasons that I've gone into elsewhere. But the case for legal abortion stands on its own. It doesn't need nonsense statistics to back it up.
Known knowns
I rarely pay attention to claims that "casualties in Iraq are down this month" because anyone who's followed the place even cursorily knows that they're remarkably cyclical, peaking each year right around Ramadan. (And don't get me started on the people who wrote that the peak in October 2006 was some sort of attempt to influence the American election.) But I didn't know why:
August 31, Juan Cole was shrill about administration and media statements that GI deaths in Iraq were down for the year. His point was that temperatures get to 120 degrees, so summer means low combat. He produced a table contrasting 2006 with 2007 by month: deaths are higher in 2007 than 2006. He asked for a visual display and many responded. I'm posting mine here.
But this is just part of a pattern of communicating, where true facts are presented within a limited context, so that the resulting asymmetry of information creates a false impression. Most of us don't know that Iraq temperatures reach 120 degrees in July so that physical activity like combat is down. On the other hand, if an official announced that lowland US snow fall in July was the lowest it's been all year -- duh. The difference is that here in the US we all know about July snow fall, so there's no asymmetry.
But as the author points out, many of the administration and military folks making those sorts of claims do know this (or at least, I certainly hope they do; if they don't, we're in worse trouble than I thought). They nonetheless produce meaningless claims about how casualties have fallen.
Department of awful statistics
It seems pretty clear to me that the drinking age is unconstitutional and immoral. If you can vote and get shanghaied to foreign wars, you are old enough to have a beer.
It also seems pretty clear to me that the drinking age will not be changed, because the main constituency against it is the small segment of the population currently between the ages of 18 and 21. And most of them are too busy building beer funnels to get a really solid political movement going. Also, it's tough to get momentum when your leadership abruptly stops caring every three years.
However, every so often I catch sight of something that brings home how silly the whole thing is. Such as this, from Radley Balko:
Research published in the Journal of Adolescent Health in 2004 found that adolescents whose parents permitted them to attend unchaperoned parties where drinking occurred had twice the average binge-drinking rate. But the study also had another, more arresting conclusion: Children whose parents introduced drinking to the children at home were one-third as likely to binge.And you could make a pretty good argument that drinking in the woods and getting bombed at unattended parties are the product of the minimum drinking age.
Of course, when the anti-alcohol activists cite the "earlier the age one starts drinking, the greater the chance of addiction" figure, they lump it all in together, which paints an incomplete picture, and makes for bad policy.
It's even worse than that; those figures usually leave in members of strict religious groups, a large segment of whom will never take more than a few drinks, if that. Obviously, it's hard to develop a drinking problem if you never taste the stuff, so those people drag down the averages. But they don't tell you anything about how the age at which one starts drinking affects your later alchohol consumption, except for the trivial observation that if you join a religious group that forbids drinking, you will probably not develop a drinking problem.
I'm a genetic determinist on these things; early drinking outside the home is most likely a sign that you're the kind of kid who has little parental supervision and a penchant for getting into trouble. Parents who don't supervise their children have probably bequeathed a substantial genetic legacy of irresponsible behavior to their children. And troubled, irresponsible people are more likely to develop drinking problems.
I'd like to see a study that compared upper-middle class kids from the suburbs to those in New York City, where, anecdotally, drinking seems to start earlier because the kids don't need to be driven everywhere. Are kids from New York City's private schools more likely to be alcoholics later in life than, say, kids from Englewood? I can't say I noticed any statistically significant differences in college.
| 
