!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
CHANCE News 2.05
(27 Feb to 14 March 1993)
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Prepared by Laurie Snell
Please send suggestions for articles or format to
dart.chance@dartmouth.edu
Back issues of this chance news and full text of the articles referred
to are provided on the Chance gopher. Just point your gopher to
chance.dartmouth.edu
Gopher is available to anyone with access to internet. Your
computer guru should be able to show you how to use it if you
have not used it.
===================================================
Science is built up with facts, as a house is with stones. But a
collection of facts is no more a science than a heap of stones is a
house.
Jules Henri Poincare
===================================================
ARTICLES ABSTRACTED
>>>>>==========>> Abe Ross writes: A recent program on CBC discussed DNA fingerprinting as a means of identifying murderers, rapists, etc. When asked how accurate the test was one of the researchers said that the probability was less than 1 in 10,000 of two people having the same print. She want on to say that once you get a rate like that it might as well be 1 in a million. (I assume she was referring to classic statistical p levels, e.g., p <.00001 and p <.0000001.) However, a defense lawyer then pointed out that close to 5 million people lived around the Toronto metropolitan area. If the odds of a mismatch were 1 in a million then only 5 other people were likely to have the same DNA print. But, if the odds were 1 in 10,000 as many as 500 people could have the same DNA print. Interesting perspective. <<<========<<
>>>>>==========>> Note: There is a recent Nova program on DNA typing. I have not seen it yet. It will be shown in our area Thursday night. However, the New York Times critic Walter Goodmen was not impressed and writes: "Whether the producers were driven to use re-enactment's because the process of DNA typing, mostly a comparison of dots on a computer printout, does not make for visual excitement, or whether they were attracted to the subject in the first place by its appeal to audiences who go for tales of rape cannot be decided even by DNA testing. Either way, "Nova" stands indicted on suspicion of the worst." <<<========<<
>>>>>==========>> Commenting on our report on the Personal Health column of Jane Brody The New York Times, 24 Feb, 1993, C12 Gerrold Grossman writes: Her example of a "300 percent increase", was an increase from 1% to 3%. Obviously, that is only a 200 percent increase (since a 300 percent increase would take it from 1% to 4%). Innumeracy is widespread, annoying, and harmful!" <<<========<<
>>>>>==========>> Tor Tosteson comments I saw an article yesterday in the Valley News (our local newspaper) about "401(k)" plans. A stockbroker was quoted as saying that the investment community was concerned that individuals now given a choice about retirement investments were choosing to avoid stocks more than the experts would recommend. I find it interesting that the brokers criticize the risk preferences of consumers. <<<========<<
>>>>>==========>> Design your own. Forbes, March 15, 1993 Kenneth L. Fisher Since pension funds and others now allow the employee to allocate the assets, brokerage firms sell "asset allocation" charging about 3% per year. As remarked above there is concern that a person's natural risk aversion will lead to their lacking enough money to retire. In this article you are told how to do it yourself. The author provides the percentage you should have in equities, bonds and cash as a function of age. This would be an interesting topic for a chance course if anyone knows what these percentages are really based on. <<<========<<
>>>>>==========>> Medicine as a stochastic art. Lancet 27, Feb 1993 Katerine Ierodiakonou, Jan Vandenbroucke The authors state that the ancient Greek philosophers called medicine a stochastic art. The Greeks observed that two patients might be given the same careful treatment but have different outcomes caused by chance and proposed that the aim of medicine should not be seen as the achievement of the desired end, but as doing everything possible to achieve this end. Greeks in the second century argued about where the stochastic nature came from. Alexander of Aphrodisias held that the problem was that medicine was not an exact science but others such as Galen believed it was caused by inherent variation in the patients. The authors tie up some modern concepts such as chaos theory with those of the Greeks but suggest that we have still not resolved the stochastic nature of medicine and we will continue to "face the tension between empiricism and scientific insight." <<<========<<
>>>>>==========>> Do-it-yourself health tests. LATimes, March 11, Business; D1 Michael Schrage A claim that a new generation of low-cost medical technology that allows testing in the home will drive up health care costs. The FDA is trying to figure out how reliable a test should be to be approved for home use. All the more reason to teach students about false positive etc. <<<========<<
>>>>>==========>> Confusion in the name of science. Toronto Star, March 13, 1993 Don Sellar The article provided the quote at the beginning of this chance news. It comments on the fact that headlines of stories about studies can often be misleading. For a example a Reuter wire service story in the Star with headline "Migraines stopped 'altogether' by arthritis drug, study finds" led to a lot of people rushing out to get the drug when in fact the study of 23 patients had reduced the frequency by 23 percent and their length by 41 percent. Similarly, the Star published a Reuter wire story under the headline: "Scientists identify cause of multiple sclerosis." on the same day the headline in the Globe and Mail about the Nature article was "Cause of MS not found, scientists say" <<<========<<
>>>>>==========>> Studies rebut controversial aids theory. LATimes, March 11, 1993 Sheryl Stolberg Peter Duesberg has argued that drug abuse and not the HIV virus causes AIDS. He often claimed that no study had been done to specifically compare incidence of AIDS among groups with the same drug-use patterns. The article reports on two studies which responded to the challenge and claim to refute Duesberg's theory. The first reported in the current issue of Nature. This study looked at drug use and AIDS among 1027 San Francisco men and found no correlation between the two. The second is reported in the current issue of Lancet. This study reports that among 715 homosexual men in Vancouver, those who tested positive for HIV and went on to develop AIDS had the same drug-use patterns as those who were HIV-positive but had not developed AIDS during a comparable time. Furthermore, only HIV-infected people developed AIDS. Duesberg does not accept the findings. <<<========<<
>>>>>==========>> Death American style. Don Kirk Newsday, March 9, 1993 It is claimed that the recent popularity in Japan for American fast food symbolized by the 958 Japanese McDonald's has led to a less healthy Japanese population. "Japanese, male and female, are taller and heavier than their parents and they are suffering from diseases associated with high-fat-eating people, notably Americans. The death from all cancers have gone up by at least one-third in just one generation. The rate of heart attacks and diabetes has also increased." Most Japanese epidemiologists believe that fatty foods are responsible for most of the increases. However there are other causes for some of the increases. For example, the increased incidence of breast cancer could be associated with the earlier menstruation and a later age for having children for the current generation. <<<========<<
>>>>>==========>> Where have all the lefties gone? Star Tribune, March 4, 1993 Lewis Cope About 13 percent of Americans are left handed, but only about 5 percent of those in their fifties are left handed. A recent study by Robert Glynn concluded that left handed people live as long as right-handers suggesting that the fact that their are fewer left handers among older people can be accounted for by pressure to become right handed. The author of earlier studies which claimed that left handers live less long, Stanley Coren, believes that left handers have a higher rate of deaths from accidents because tools, driving customs etc. favor right handers. The fact that there are a higher proportion of left handers among younger people and that younger people have a higher proportion of accidents has allowed both Coren and Glynn to claim that the other has not answered the question. Coren can claim that Glynn, by limiting himself to subjects over 65, has taken out the very group where the danger of being left-handed has an effect. Glynn can claim that Coren by not controlling for age will necessarily find a shorter lifetime for left handers just because there are more left handers among the younger people and younger people have a higher accident rate. <<<========<<
>>>>>==========>> Study links heart disease to margarine. The New York Times, March 5, 1993 Marian Burros A study in the March 5 issue of Lancet shows that women who eat four or more teaspoons of margarine a day have a 50 percent greater risk of developing heart disease than women who eat margarine only once a month. Women who eat larger amounts of other forms of solid and semisolid equivalent to six or more teaspoons of margarine a day increase their risk for heart disease by 70 percent. Such vegetable fats are found in cookies, cakes and fried fast foods. The findings came from the 1987 Harvard Nurse's Health Study which followed the dietary habits of 88,751 women from 34 to 59 years old for eight years. The article claims that the findings are likely to apply to men as well. Dr Walter Willett who conducted the study does not suggest returning to butter but rather that vegetable oils should be used in their natural state like olive oil, corn oil, or canola oil. Others claims that margarine that is low in trans-fatty acids is fine also. The study found that white bread was associated with higher risk of heart disease but red meat and butter were not. Willett commented that the claim that eating a lot of red meat is linked to colan cancer was supported by the Nurse's study and that the lack of association with butter may be because of the relatively small use of butter as compared to margarine these days. He also observed that the white bread association could be explained by other differences between people who eat white bread and those who eat other kinds thought to be better for you -- showing once more the difficulty of studies that are not controlled. <<<========<<
>>>>>==========>> New theory suggest cell regulator may hold the key to fighting AIDS. The New York Times, March 9, 1993 Gina Kolata Researchers Leonard Adleman, a mathematician at University of Southern California and Joseph Margolick, an immunologists at Johns Hopkins have suggested an explanation for the fact that the immune system does not produce more CD4 cells to compensate for those lost to the HIV virus. There aretheir two kinds of T cells, CD4 cells and CD8 cells which differ by just one molecule on the surface. CD4 cells can be attacked by the HIV virus but CD8 cells cannot. Normally, CD4 cells outnumber CD8 cells by 3 to 2. They suggest that the immune system does not distinguish between the two types and so when it sees the decrease in CD4 it adds more cells but again in the 3 to 2 ratio. This leads to a selective loss of CD4 cells. The analogy is suggested with a bag of quarters and dimes that has a small hole that allows dimes to drop out but not quarters. As dimes fall out dimes and quarters are put back in the original ratio. Eventually, the bag will have mostly quarters. If their theory is correct they suggest that they might be able to maintain the normal ratio of the two types of cells by killing off a certain proportion of the CD8 cells. In other words make a hole for the quarters also. <<<========<<
>>>>>==========>> The testing issue revisited. About Women on Campus, Winter 1992 A study by Howard Wainer and Linda Steinberg of ETS found that the average math SAT score for women is 35 points higher than for women on average even when the two groups are matched by college courses and grades. They suggest that an average of 52 points should be added to a women's SAT score if it is used in equations used for placement or admissions purposes for men. The report is Research Report No. 91-45 and is available from ETS in Princeton. It's title is "Sex Differences in Performance on the Mathematics Section of the Scholastic Aptitude Test: A Bidirectional Validity Study. <<<========<<
>>>>>==========>> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! CHANCE News 2.05 (27 Feb to 14 March 1993) !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Please send suggestions to: jlsnell@dartmouth.edu >>>==========>>|<<==========<<<
>>>==========>>|<<==========<<<