CHANCE News 1.11
              (27 Nov to 3 Dec 1992)

Prepared by J. Laurie Snell

Please send suggestions to dart.chance@dartmouth.edu 




>>>>>==========>> Scientists hike probability of a major quake. Lost Angeles Times, Dec 1, 1992 Part A page 1 A panel of California's leading earthquake scientists put out a report revising upward the probability that a quake of magnitude 7 or greater will hit Southern California. They report a 5% to 12% chance of the quake each year (A kind of confidence interval for probabilities.) They say that the chance of such an quake in the five year period starting September 1 may be as high as 47%. However, at another place in their report they say "the most likely case is that no large earthquake will occur in the next few years". An editorial in the next day's LA Times pokes a little fun at the report but also says that "we should not chase them into hiding with ridicule. These earthquake predictors are going through the painful process of finding their way in public and someday they may start hitting the mark with more regularity." <<<========<<

>>>>>==========>> Genes may play a role in divorce; Twins study reveals personality link. The Houston Chronicle, Sec A, page 2 A discussion of a report by two Univerisity of Minnesota Psychologists, Lykken and McGue, that appeared in the November issue of the Journal of Psychological Science. The study involved 722 pairs of identical twins and 794 pairs of same-sex fraternal twins aged 34 to 53. The authors report that if one twin in a pair had been divorced the probability that the second twin had been divorced was 45 percent for identical twins and 30 percent for fraternal twins. This difference suggests a genetic factor playing a role in personality or something that might relate to divorces. (Note that if the fraternal twins were assumed independent the probability for a twin having a divorce would not be 30 percent but rather higher (more like 46 percent). The article says that about 20 percent of the twins in study had been divorced so I guess that neither identical or fraternal twins act independently as regards divorce. <<<========<<

>>>>>==========>> The DNA Wars; touted as an infallible method to identify criminals, DNA matching has mired courts in vicious battle of expert witness. Los Angeles Times, 29 Nov, 1992, Magazine Section Edward Humes A long and well written article on the saga of DNA fingerprint in the courts including the hassle of expert witnesses face with the lawyers. <<<========<<

>>>>>==========>> Rape conviction overturned on DNA tests. New York Times, 2 Dec 1992 Sec B page 6 Jonathan Robinovitz Harry Kotler was freed after serving 11 years in prison for rape on the basis of DNA tests that were not available in 1982. This is one more in a growing number of cases that have been reconsidered and often overturned by showing the lack of a match in DNA fingerprinting. Unlike the situation matching DNA fingerprints the case where they do not match is generally considered to be completely convincing. <<<========<<

>>>>>==========>> U.S. companies push for perfection. USA Today, Dec 1, 1992 John Hillkirk A memory chip contains 16 million microscopic transistors and Kodak says a 35mm film negative is made up of an almost infinite number of photographic elements. Such examples have motivated companies to aim for a "six sigma" level of quality. Motorola is making pagers and cellular telephones that of 99.99966% defect free. <<<========<<

>>>>>==========>> Annals of radiation: The cancer at Slater School. New Yorker Magazine, Dec 7, 1992 Paul Brodeur In December 1990 teachers at the Slater Elementary School in Fresno, California became concerned about the number of their colleagues who had gotten cancer and who had worked on the side of the school building with high-voltage transmission lines a hundred feet from the building. This article describes the teachers attempts to determine the extent of the cancer among the teachers and the parents to determine if the transmission lines are a danger to the teachers and their students. The article is a typical infinitely long New Yorker article but in the process it reviews the studies that have been made to investigate the dangers of high-voltage lines and the political, legal, and other forces that a group faces when they have an environmental concern. There is a brief discussion of how the Poisson distribution is used to evaluate the probability of a cancer cluster but the most interesting part is the dynamics of working with the government agencies to investigate such a concern and the role of the press in reporting current investigations. It would be useful to have a more technical discussion of the statistical problems invovled in estimating probabilities of cancer clusters accuring by chance. <<<========<<

>>>>>==========>> AIDS and sexual behavior in France ACSF investigators and Sexual lifestyle and HIV risks. A.M. Johnson and others. Nature, 3 Dec 1992 These two articles report on two major surveys (sample size of order 20,000) one in France and one in England to determine the sexual behavior of these two populations to help in AIDS prevention programs and to provide data for models which attempt to predict the spread of AIDS. The French chose to do a telephone survey and the British to use face to face interviews. Both studies discuss the design of their survey and the various problems in doing surveys of this kind. An accompanying editorial comments on the importance of these studies and some of the uses to which they will be put. It also discusses possible biases which could explain inconsistencies such as men finding more heterosexual partners than there are women acknowledging partnership. (It reminds us that Margaret Thatcher refused to have government funds used for the British survey). !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! CHANCE News 1.11 (27 Nov to 3 Dec 1992) !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Please send suggestions to: jlsnell@dartmouth.edu >>>==========>>|<<==========<<<