CHANCE News 1.07
              (15 Oct to 22 Oct 1992)




>>>>>==========>> Mattel says it erred; teen talk Barbie turns silent on math. The New York Times 21 Oct 1992 Mattel Inc. has decided that the computer chip that randomly selects four phrases for each Barbie doll will will now pick from 269 selections, not 270 leaving out the phrase "math class is tough" in response to protests by the American Association of University Women and many others. <<<========<<

>>>>>==========>> Parkfield project attempts to predict earthquakes. CNN transcript #141 21 Oct 1992 Patrick Greenlaw An interview with Jim Davis, a geologist of the California Department of Conservation discussing the prediction of a 40 percent chance for an earthquake of magnitude six to occur near Parkfield, California within a specific 72 hour period. Davis explains that with no information if you pick a three-day interval at random, the probability of a 6-level earthquake is about three hundredths of one percent. Using information relating to a 4.7 earthquake that occurred in Parkfield makes this probability 1300 times the normal random value. The discussion in this transcript is rather garbled as relates to this calculation and it would be interesting to see the details. <<<========<<

>>>>>==========>> High court to rule on admissibility of scientific evidence in U.S. courts The New York Times, 14 Oct 1992 Linda Greenhouse "Opening the door to a far-reaching debate over the role of science in the courtroom, the Supreme Court agreed today to decide how widely accepted a scientific process or theory must be for it to be admissible as evidence in Federal court." The court agreed to consider a case where two Federal courts in California refused to allow a jury to hear evidence of studies relating to birth defects due to the drug Bendectin. Their ruling is expected to relate to the other important uses of scienfific evidence such as the use of DNA fingerprinting in the courts. Incidently, the drug Bendectin was used as an example in Peter W. Huber's article "Junk Science in the Courtroom" in the June 1992 issue of Scientific American. The current Scientific American (November) has letters critizing Huber's treatment of this example with a reply from Huber. <<<========<<

>>>>>==========>> Bayesian resolution of the exchange paradox American Statistician, Nov 1992 Ronal Christensen and Jessica Utts A timely discussion of Marylin vos Savants latest paradox discussion. <<<========<<

>>>>>==========>> Change in body weight and longevity JAMA 21 Oct 1992 Lee, Paffenbarger Results of a study on the effect of weight change on mortality using the results of the Harvard Alumni Health Study an ongoing study of Harvard Undergraduates who matriculated between 1916 and 1950. The authors found an increase in all-cause mortality among those who gained weight but, somewhat surprisingly, also for those who lost weight. The same result was obtained for heart disease but for cancer the relative mortality rate was essentially independent of weight change. <<<========<<

>>>>>==========>> A matter of opinion. Teaching statistics Autumm 1992 Gwen Royle A discussion of the outcomes of the polls in the 1992 British election. A daily account of the polls shows consistently the promise of a very close election with labor generally slightly favored. On the day before the election Labor had 42 percent and conservative 39 percent and then the very surprising final vote of 34 percent for labor and 42 percent for the conservatives. A good chance to ask your students -- could it happen here this year? !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! CHANCE News 1.07 (15 Oct to 22 Oct 1992) !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Please send suggestions to: jlsnell@dartmouth.edu >>>==========>>|<<==========<<<