!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
CHANCE News 2.07
(30 March to 12 April 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
===================================================
In the long run we are all dead.
John Maynard Keynes
===================================================
ARTICLES ABSTRACTED
>>>>>==========>> Poll delivers bad news to the media. Los Angles Times, 31 March 1993, A16 David Shaw Report of a Times poll on the publics attitude towards the news media. The results are compared with a similar poll carried out in 1985. While about 71% said the media was doing a "fairly good job" only 17% said they were doing a "very good" job down from 30% in the 1985 Times poll. Declines in favorable responses were seen throughout the poll, with the most favorable rating "very good" declining significantly. When asked why they had lost their confidence in the news media they most often cited sensationalism (28%), "selective reporting-- they don't tell the whole story" (24%) and bias (21%). The article has lots of analysis and graphics. The poll was an telephone interview with 1703 adults. <<<========<<
>>>>>==========>> Court hears case on science rules. The New York Times, 31 March 1993, A19 Linda Greenhouse A account of the beginnings of the Supreme Court consideration of the issue of the kind of scientific evidence should be permitted in the courts. The initial arguments were between those who support the Frye rule of 1923 that would allow scientific evidence only if it is generally accepted by the scientific community and the recent Ninth Circuit ruling that suggested the evidence should be published in a refereed journal, versus those who support the 1975 rule of Congress that provided a more liberal approach. that would allow "a witness qualified as an expert to testify in the form of an opinion or otherwise". The judges provided some amusing comments suggesting some concern with both positions. Harvard graduate Justice Blackmun complained "There are Harvard law professors on both sides of this case. I had hoped we could get together and lead us out of the wilderness" <<<========<<
>>>>>==========>> To yor health? The paradox of moderate drinking. Washington Post, 6 April 1993, Z10 Robin Herman The author reviews the numerous studies that have shown that moderate drinking of alcohol lowers a person's risk of coronary heart disease and discusses some of the reasons that public health officials do not recommend that those who do not drink start drinking. Special attention is given to studies that attempt to explain why red wine might especially helpful in preventing heart attacks. The writer discusses in some detail studies related to the so-called "France paradox" : despite a diet with a high amount of saturated fat French men die of heart attacks at half the rate of American men. <<<========<<
>>>>>==========>> Lefties don't die young after all. LA Times, 4 April 1993, A1 Thomas H. Maugh II A discussion of a UCLA study due out next month that deals with the controversy as to whether left-handers live less long than right handers. In the population as a whole about 9% of women and 13% of men are left-handed while at age 10, 15% of the population is left-handed, at age 20 it is 13%, at age 50 it is 5% and at age 80 it is less than 1%. Previous studies by Coren and Halpern claimed a significant less long lifetime for left-handers and suggested that this is because left-handers have more accidents than right-handers caused by tools, cars etc. designed for right-handers. Critics of these studies have claimed that the difference in older populations can be explained by the pressure for left- handers to switch to right-handers. Coren has said that there is no evidence for such a trend. The UCLA study asked 2,787 people if they had been switched. About half of whom were enrolled in the Multicenter AIDS Cohort Study at UCLA and the other half men and women from a retirement community. They found that the proportion of people whose handedness had been switched by parents and teachers increased with age, which they say largely but not entirely accounted for the apparent disappearance of left- handers from the elderly population. <<<========<<
>>>>>==========>> Ridding blood of lead may raise IQ. Newsday, 7 April 1993, City page 27 Gale Scott A report on two studies in the current Journal of American Medical Association designed to assess the value of removing lead from children's blood resulted in improved IQ scores. Researchers in one of the studies found that standardized scores of children aged 1 to 6 improved by one point for every 3-microgram decrease in their blood-lead level. The second study announced that after spending $9,600 per household to truck 15 tons of lead-contaminated soil out of the yards of 126 Boston house found only a modest decrease in the level of lead in children's blood and concluded that it was not worth the money. Other such studies are being conducted and will effect federal policy in combating this problem. <<<========<<
>>>>>==========>> New study questions use of AZT in early treatment of AIDS. The New York Times, 2 April 1993, A1 Lawrence K. Altman A report of a large European study aimed at evaluating the effect of treatment of early infection with HIV with AZT followed patients for an average of three years. Of those who were treated immediately after diagnosis as an HIV carrier 92 percent survived over the three year period, and 93 percent among those given placebo. Over the three-year period, 18 percent of both groups progressed from HIV to AIDS or death. Previous studies in the United States for a shorter length of time showed that those treated with AZT tended to have increased CD4-cell counts and this was the basis for recommending early AZT treatment. In this study there was also average increase of 30 in the CD4 count for those given AZT immediately but this rise did not seem to lead to any clinical benefit. Some of the earlier studies were stopped early for ethical reasons and the difference in findings from this study raise questions about the technique of stopping a study based on markers for a disease. The video series "Against all Odds" has in the last tape an excellent discussion of the first study that showed the effectiveness of AZT. <<<========<<
>>>>>==========>> Homosexuals and the 10% fallacy. Wall Street Journal, March 31 1993, A14 J. Gordon Muir The author cites numerous polls in the United States and other countries that suggest that the 10% figure routinely cited for the percentage of the population that is homosexual is significantly wrong. This 10% figure is attributed to the Kinsey surveys in the 40's and the writer discusses some of the flaws in these surveys. The recent surveys vary quite a bit depending on how the question is asked but typically end up with percentages more like 1% to 2% for exclusively homosexual and 3% to 4% having had some homosexual experience. with percentages slightly lower for women than men. A recent study that gives higher percentages is the widely publicized "Janus Report" that estimated 9% men and 5% women. The author feels that his survey also has methodological flaws. <<<========<<
>>>>>==========>> Should the fight against AIDS be targeted? YES: Not everyone is at equal risk by John Gagnon NO: Denial of its spread is greatest danger by June Osbern. USA TODAY, 7 April 1993, 13A A recent report of the National Research Council has suggested to some that efforts to combat AIDS should be targeted in certain areas and within certain groups. Gagnon and Osbern debate the question "Should the fight against AIDS be targeted?" <<<========<<
>>>>>==========>> Doctors' pay underestimated but resented just the same. The New York Times, 31 March 1993, A21 Philip J. Hilts A survey asked people to estimate the income of doctors and executives in hospitals, drug companies and insurance companies and to say what they thought was a fair income. In most cases they felt that a fair income was less than their estimate of the incomes. On the other hand, their estimate of the incomes were significantly lower than the actual incomes. For example they were asked how much they believe doctors make in the higher-page specialty groups, radiology and anesthesiology. They estimated $100,000 after expenses (which they also thought was fair) and the average income in this two groups was about $225,000. Doctors in private practice average about $190,000 after expenses. They estimated drug company executives make about $600,000 and thought $150,000 to be fair. The top executive salaries in the six leading companies ranged from $2,000,000 to $12,000,000 in 1991. <<<========<<
>>>>>==========>> EMF and Cancer. Science 2 April 1993, Letters ORAU panel, A. Ahlbom and M. Feychting A panel established by the Oak Ridge Associated Universities at the request of the government reviewed the literature on EMF and cancer and concluded that epidemiological findings about EMF and cancer are inconclusive, inconsistent and there is no present plausible mechanism concluding that further research in this area should not receive high priority. Since the report was written, two Swedish studies appeared suggesting that there is a causal association. The first letter from the ORAU panel critiques the Swedish studies and explains why the panel does not think that these studies are sufficiently compelling to alter the conclusions of their report. The second letter is from authors of one of the studies giving their critique of the ORAU critique. A more complete discussion of the controversy over EMF and cancer can be found in Richard Stone's article "Polarized debate EMFs and cancer", Science, 11 Dec.1993, News and Comment, p. 1724) <<<========<<
>>>>>==========>> Poll Faulting. Chance Magazine, Winter 1993 Stephen Ansolabehere and Thomas R. Belin The authors use the disparity in the polls in estimating Clinton's lead on October 20 to raise questions about the variability of the polls and the pollster's methods of reporting their results. They point out that the margin of the error for the amount in the lead is larger than for a single candidate's percentage. They work out how much larger and suggest the rule of thumb: margin of error of lead = sqrt(3) x reported margin of error. They also discuss the disparity in the polls resulting from who the polls target -- all eligible voters, registered voters, or likely voters. The conclude with some recommendations to improve the reporting of poll results. <<<========<<
>>>>>==========>> Graphs in the presidential campaign: why weren't they used by more than one candidate? Chance Magazine, Winter 1993, Visual revelations Howard Wainer A discussion of the effective use of graphs by Perot in the presidential campaign with some suggestions of how he could have improved an already fine performance. The author suggests that they were not used by the other candidates because they tell the story too clearly. <<<========<<
>>>>>==========>> Jack or Jill. Lancet 20 March 1993 Editorial This editorial reviews where we are in sex selection. Data has been published for the success of Gametric Ltd (mostly in the USA). 72% of 1034 couples wanting a boy were successful and 69% of the 193 couples who wanted a girl were successful. A survey in 1970 and 1975 showed that for first pregnancies 45% preferred a boy and 34% had no preference and for second pregnancies, of those with a girl, 68% preferred a boy and 18% had no preference and of those with a boy 72% preferred a girl and 9% had no preference. The editorial suggests that this is a rare opportunity to plan in advance how a potentially valuable technique could be monitored and made available to the public. <<<========<<
>>>>>==========>> Enhancing statistical literacy: enriching our society. Journal of the American Statistical Association Katherine K. Wallman A presidential address at the 1992 annual meeting of the ASA calling for strengthening statistics and statistical thinking in all sectors of the population and suggesting some ways to do this. (As a small aside I was puzzled by the author's complaint of "The potential sampling error for smaller subgroups is larger" in the explanation of New York Times polls. Perhaps the complaint is that this is jargon but if you have a better explanation I would appreciate hearing it. <<<========<<
>>>>>==========>> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! CHANCE News 2.07 (30 March to 12 April 1993) !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Please send suggestions to: jlsnell@dartmouth.edu >>>==========>>|<<==========<<<
>>>==========>>|<<==========<<<