!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
CHANCE News 3.01
(1 January to 21 January 1994)
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Prepared by J. Laurie Snell as part of the
CHANCE Course Project supported by the
National Science Foundation and the
New England Consortium for Undergraduate
Science Education.
Please send suggestions to:
dart.chance@dartmouth.edu
Current and previous issues of CHANCE
News can be found on our chance gopher.
Just point your gopher to:
chance.dartmouth.edu
====================================
"Facts speak louder than statistics"
Mr. Justice Streatfield (1950)
====================================
IN THIS NEWS LETTER
My remark in the last CHANCE News that we might hear from the Skeptical Inquirer about the Dossey book on prayer showed that I have not been a faithful reader of the Inquirer. Jeff Witmer informed me that he and Michael Zimmerman had already written on this subject for the Inquirer. Here are some remarks on their article. <<<========<<
>>>>>==========>> Intercessory Prayer as Medical Treatment? An inquiry Skeptical Inquirer, 15 (2) 177-180 1991 Michael Zimmerman and Jeff Witmer
This article discusses studies that have attempted to show that prayer can have a beneficial effect on medical outcomes. One of the first such studies was carried out by Sir Francis Galton (1883), who hypothesized that monarchs were more likely to be the objects of prayers than commoners and should therefore live longer, all other things being equal. Of course, all other things are not equal, but no trend was found anyway. The authors comment on the double-blind study of Randolph Byrd published in the Southern Medical Journal. It was that study that inspired Dr. Larry Dossey to write the book mentioned in the last CHANCE News. Zimmerman and Witmer point out that Byrd compared the two groups according to complications rather than patients. The complications were not independent, and so a single patient was apt to have a number of these complications. They mention other problems with the study which make the results less convincing. The authors made a serious effort to find the complete literature on controlled studies on the effect of intercessory prayer and found essentially no papers. Concerned about publication bias, they wrote to editors to see if they had received such papers and the answer was again no. They express concern that so much of the information on a topic like this is anecdotal and so little is based on serious studies. <<<========<<
>>>>>==========>> James Hilton suggested the following interesting series of articles in the LA Times on genetic roots of violence. Fear Clouds Search for Genetic Roots of Violence. Part 1: LA Times, 30 Dec 1993, A1 Part 2: LA Times, 31 Dec 1993, A1 Sheryl Stolberg
Part 1 of this two part series discusses ongoing research aimed at determining if there is a genetic component to violent behavior. The article discusses the current research of Dr. Markku Linnoila at the National Institutes of Health. Linnoila states that his 13 years of research on the brain chemical serotonin has proved that people with low levels of this neurotransmitter are prone to impulsive, violent acts. This was established by examining the spinal fluid and blood of more than 1,000 Finnish prisoners including 300 violent offenders. Linnoila estimates that a serotonin deficit may be present in as many as one out of every 20 men. He is currently working on identifying the genes that control the manufacture of this brain chemical with the hope that this information will lead to the development of programs or drugs to help prevent violent behavior. The article discusses the questions this kind of research has raised such as concerns about its racist applications. Because of such concerns, researchers are wary about allowing their research to be reported in the news. Part 2 of the series discusses ongoing experiments to see if it is possible to predict and prevent violent behavior. In 1960, Leonard Eron, a psychologist at the University of Michigan, attempted to measure aggression among 875 third-graders in a semi-rural region on the Hudson River. He interviewed teachers and students, asking simple questions such as: who starts a fight over nothing? who pushes and shoves? and determined how aggressive these 8-year-olds were. Eron followed the children until they turned 30. He found that those who were highly aggressive at age 8 were about 3 times as likely as their peers to commit crimes by age 30; more often than not, their crimes were violent. Eron's method of picking out children at risk is now being used in a large violence prevention project in Chicago involving more than 8,600 children in 16 Chicago area schools. The study is paid for by the NIH and run by University of Illinois Psychologists. Researchers take children as young as 8 and attempt to identify those who are apt to become violent later in life. Those identified are "treated" by a variety of levels of counseling aimed at changing their behavior They will be followed for at least 5 years and hopefully until they are 22. The article discusses all the obvious problems: privacy questions, lack of truly informed consent, concern that similar attempts in the past have resulted in the subjects doing worse instead of better after the intervention. It is really weird to be reading in the LA Times that the experimenters do not feel that they can tell the parents the real purpose of the study. These two articles are remarkably complete discussions of all the problems involved in attempts to study causes of violent behavior. <<<========<<
>>>>>==========>> Darts Trounce the Pros by 42% to 2.2% in Latest Six- Month Stock-Picking Duel Wall Street Journal, 11 Jan 1993, A1 Georgette Jasen
The darts have won a smashing victory for the darts in the latest 6-month series in the ongoing battle between the darts and the investment experts. In this contest Wall Street Staffers throw darts at the stock tables for the darts choices and the experts use their wisdom to choose their stocks. The darts averaged a 42% gain in the period July 7 to Dec. 31 compared to 8% for the Dow Jones and 2.2% for the experts. The two most successful experts participate in the next six-month period, but in this case, they might better invite a couple of the darts. The Pros remain in the lead in a series of overlapping six-month bouts by a margin of 25 to 18, although they are only in the lead 22 to 21 over the Dow Jones. In terms of average gains, in 43 contests over the past four years, the investment professionals have had an average gain of 7.8%, compared with 4.4% for the industrial average and 3.5% for the darts. In a previous CHANCE News (2.18) we reported on an article discussing Burton Malkiel's attempt to show that the experts are not doing sufficiently better than the darts to be inconsistent with his random walk hypothesis. He will surely be pleased by the results of this six-months period. <<<========<<
>>>>>==========>> Psychology and Survival Lancet, 15 Jan 1993, 174-176 Letters to the editor
In a November 6 issue of Lancet (see Chance news 2.13) Phillips and colleagues concluded that Chinese- Americans born in an "ill-fated" year associated with a particular disease are more likely to die early from that condition than Chinese-Americans born in other years or in non-Chinese matched controls. This series of three letters points out why they are not convinced by this study. The first letter mentions a number of obvious biases and alternative interpretations. For example, a difference in age of death could indicate a difference in onset of the disease rather than in survival time. In the second letter the writer states that his experience in China leads him to disagree with the assumption that most Chinese people have an understanding of the five-phase theory. The writer of the third letter observes that, as mentioned in the paper, numerous studies have shown that group therapy increases the lifetime of patients with breast cancer and takes the authors to task for their remark, in their conclusion, that their "findings apply not just to breast cancer ..." pointing out "In fact, their findings do not seem to apply to breast cancer at all. (The difference between the average age at death from breast cancer of Chinese-American women born in the ill-fated year and those born in other years was .64 years which did not differ significantly from the difference (.58 years) for the matched non- Chinese controls.) This writer also points out that, while the authors found a significant association between birth-year and average age of death for acute myocardial infarction among Chinese-Americans, the difference was not significant for chronic heart disease and, for the group of all other heart diseases, those born in the ill-fated year actually lived longer than those born in other years. Unfortunately, we do not have a reply from the authors of the study, but checking these letters against the original article would be an interesting class exercise. <<<========<<
>>>>>==========>> Mammograms -- Your Breasts, Your Choice Wall Street Journal, 28 Dec 1993, Editorial page Bernadine Healy
The Mammography Controversy Wall Street Journal, 19 Jan 1993 Letters to the Editor
Berandine Healy was the director of the National Institutes of Health in the Bush administration. Here she argues that the recent decision of the National Cancer Institute to not recommend mammography for women under 50 is completely wrong. In fact, she argues that for this group there should be more effort to develop better mammography and better treatment. Healy observes that, if you look at the eight studies that included women in the age group 40 to 50 and then choose the five that followed women who developed breast cancer in their 40s for more than 10 years, you find that four of the five studies showed a decreased mortality for those who had cancers detected through mammography. The aggregate decrease is 17%. She remarks that, while not significant, this is a positive result. Healy claims that Clinton is using the change in position by the NCI to limit the support for mammography in his health plan. The first letter points out that Healy failed to mention that "over 10 years' time a women will have three-out-of-every 10 mammograms and examinations incorrectly interpreted as a possible cancer and that nearly every women who gets this report must undergo surgery to eliminate the diagnosis of cancer." This writer agrees with Healy that definitive tests are necessary and wonders why they were not started during Healy's term in office. The second letter is from a Professor of Orthopedics who writes: "if I lose my wife to the ravages of breast cancer, she is not a statistic, she was my wife and the mother of my children". He says that, as a family decision, his wife has had a mammogram every year from the age of 40. <<<========<<
>>>>>==========>> Heart Diseases Are Persisting in Study's Second Generation New York Times, 5 Jan 1994, C12 Jane E. Brody
Don't Make Wives a Cardiac Risk Factor New York Times, 21 Jan 1994, Letter to the Editor, Louise Cohen
The Framingham study began in 1949 and followed the living habits of 5,200 Framingham residents. For the past twenty years 5,100 children of the original participants have also been studied. These offspring, now middle-aged, are in many respects healthier than their parents were at a comparable age. Blood pressure and cholesterol levels are lower, and far fewer smoke. But the men are heavier and less active, and both men and women have much higher rates of diabetes than their parents had at a comparable age. There seems to be no explanation for this increase in rates for diabetes. Dr. William Castelli, director of the Framingham Heart Study, says that the health of the Framingham children mirrors the health trends in the country. He explains that falling cardiovascular death rates since the 1970's have sparked a false sense of optimism. The rate at which people suffer heart attacks and strokes has not fallen comparably to the death rates. Drugs and fancy operations and special treatments are keeping people alive who would earlier have died. Dr. William Kannel, who has been monitoring the Framingham offspring, remarked that "20% of the 50- year olds, 40% of the 60-year olds and 50% of the 70- year-olds in Framingham are taking drugs to lower their blood pressure. A study of the Framingham participants showed that men with highly educated wives who work outside the home faced a greater risk of a heart attack than did men with less educated wives. Dr. Kannel suggested that this reflected dining out and perhaps "a lack of nurturing," thus inviting the letter to the editor. The letter writer suggests that proper nurturing would be for these men to "cook dinner themselves, exercise or eat a heart-healthy meal at their favorite restaurant." <<<========<<
>>>>>==========>> Study Shows Wider Utility for Aspirin New York Times, 6 Jan 1994, A20 Lawrence K. Altman
A study headed by Dr. Richard Peto of Oxford shows that nearly everyone who is in the midst of a heart attack should take aspirin and those who have survived a heart attack or a stroke should be taking it regularly. The study is a meta-analysis of nearly all the studies undertaken of the use of aspirin and similar drugs for heart attacks, strokes and other circulatory disorders. This included some 300 controlled studies with 140,000 patients. Aspirin resulted in a 25% reduction in the risk of another heart attack for those who had suffered one. The researchers involved in this study are urging the F.D.A to change their current recommendations on the use of aspirin to reflect the results found in their study. It is observed that these recommendations will have to be assessed in relation to the well-known side effects of aspirin. The article remarks that "the jury is still out on daily aspirin use with healthy adults with no history of cardiovascular disease." <<<========<<
>>>>>==========>> Experimental Drug is Prize In a Highly Unusual Lottery. The New York Times, 7 Jan 1994, A1 Tamar Lewin
Berlex Laboratories devised a lottery, the first of its kind, to distribute a new drug, Betaseron, which is the first medicine thought to slow the course of multiple sclerosis. More than 300,000 people in the United States have multiple sclerosis. To be eligible for the lottery patients had to be certified by their doctors as having a form of the disease where symptoms come and go unpredictably, and must be in the early stages of the disease. This led to 67,000 people entering the lottery; a computer program chose a random order for those to receive the drug. Tests of the drug showed such promising results that approval was sped up, and the company could not produce the drug quickly enough to meet demand. Speeding up of approval also means that the drug's long term effects have not been properly tested, making it a kind of double lottery. <<<========<<
>>>>>==========>> "Snow Job on the Slopes" US News & World Report, 17 January 1994, pp 62-65 Richard J. Newman
This is the latest entry in the "how to lie with statistics" file, which students may find interesting because of the popular subject. The subtitle of the article reads: "Ski resorts market themselves with numbers that tend to be slippery. Here's what counts." Selected examples follow. Crested Butte Colorado advertises the highest average snowfall of any ski town in Colorado, which is true, but 11 other resorts claim more snow on the mountain, which after all is where one skis. Many resorts trumpet high-capacity or high speed lifts, in the hopes that people will automatically assume this means the shortest waits in line. Examples show that this is not the case. Lake Tahoe's Heavenly Valley boasts 4800 acres but includes wooded areas in this total. Thus Vail, which claims 4014 acres of cut runs and open bowls, actually has more skiable terrain. <<<========<<
>>>>>==========>> "Beyond the Voodoo Stick: If a picture is worth a thousand words, now many words is an information graphic worth?" The Boston Globe Magazine, 16 January 1994, pp. 10-18 Mark Feeney
The "voodoo stick" is what Ross Perot called his chart pointer during his campaign "infomercials." The author of this article comments on what he sees as the rise of information presentation in American culture. There are quotes from Edward Tufte ("The Visual Display of Quantitative Information"), whose favorite data graphic, Minard's statistical depiction of Napoleon's invasion of Russia is reproduced here. According to the article, the current era of data is 1978, when Time Magazine hired graphics director Nigel Holmes. His highly illustrated graphs are cited as examples of what Tufte unflatteringly calls "chartoons." The heavy use of graphics in USA Today is cited as an outgrowth of Holmes influence. The article has some interesting discussion of the goals of information graphics and some speculation on where current trends are leading. <<<========<<
>>>>>==========>> Note's: The 'Gambler's Fallacy' in Lottery Play" Management Science, Vol. 39 (no. 12), December, 1993, pp 1521-1525 C.T. Clotfelter and P.J. Cook
The authors present data from the Maryland State Lottery's daily numbers game. They find a consistent tendency for the amount of money bet on a particular (3 digit) number to drop immediately after that number is drawn as a winner, and then gradually to recover its former level over a period of several months. This behavior is consistent with the "gambler's fallacy," as described by Kahneman-Tversky and others. One benefit of the present analysis is that, unlike the majority of studies of the subject, it is based on real-world data as opposed to laboratory experiments. As a result, though, their data is open to another explanation: perhaps winners who regularly play a favorite number stop playing the lottery altogether after winning, since they have achieved their goal (e.g., money for a new car). While the present study does not allow one to distinguish directly between this explanation and the gambler's fallacy, the authors do cite other lottery studies which have shown that million-dollar winners, for example, tend to play more, on average, after winning. This is seen as an argument against the wealth effect as an explanation for the data observed here. <<<========<<
>>>>>==========>> Study: Vitamin E, Veggies Reduce Lung Cancer Risk Newsday, 8 January, 1993, Part II Pg 22 Ridgely Ochs
A study reported in the current issue of the Journal of the National Cancer Institute shows that a diet rich in raw fruit, vegetables, cheese and Vitamin E can reduce the risk of lung cancer for non-smokers. Researchers at Yale Medical School interviewed 413 men and women with lung cancer about their diets and compared them to the diets of 413 men and women who did not have the disease. They found that those who ate the most raw fruits and vegetables - 2 1/2 servings a day or more - had a 40% lower risk of lung cancer than those who ate them least often. This article says also: "those without lung cancer had taken doses of Vitamin E regularly, and those with the disease had not". I wonder if this is literally true. It is estimated that about 15% of the lung cancer deaths - about 22,350 deaths each year - cannot be attributed to active smoking. <<<========<<
>>>>>==========>> Statistics: A Cosmic One-in-20,000 Chance The Washington Post, 10 January, 1994, Section 1 Page A2 Reuter
An article in the January 6 issue of Nature reports that people run as great a risk of dying from an asteroid impact as they do from an airplane crash or a flood. The authors estimate this risk to be one in 20,000 in an average 65-year lifetime. Recall that we reported in the last CHANCE News that the Farmers Almanac estimated that the annual risk of being hit by an asteroid is 1 in 17 billion. The authors suggest that threats we do worry about but that have a smaller chance of happening include: certain carcinogens, poisoning by commercial foods and pills that have been deliberately tampered with, wild animals like grizzly bears, fireworks accidents, terrorist bombs and airline hijackings. They recommend that we should consider seriously having a sky survey to track comets and asteroid and using nuclear weapons to smash or divert them. (Today's New York Times claims that we have had such a secret survey.) <<<========<<
>>>>>==========>> Bad News May Not be as Bad as it Seems Chicago Tribune, 23 Jan 1994, Section Jobs, Page 1 Carol Kleiman
As reported in an earlier CHANCE News, the January unemployment rate to be released in February by the Labor Department will employ an improved survey technique that is expected to increase the unemployment rate by about .5% just by eliminating certain biases that were in the original survey. This article gives a few more details about the nature of the changes in the survey. For example, the questions will be more "sharply worded" and the interviewers will have laptop computers that will allow them to run checks as they go along and to make comparisons with answers given by the respondents the month before. Robert Reich takes credit for noticing the gender bias that was caused primarily by a question that led unemployed women to answer that they were currently doing housework resulting in their not being counted as unemployed. It will be interesting to see if Labor Department will feel that their .5% difference, based on their experimentation with the new survey is, in fact, correct when the January figures are announced. <<<========<<
>>>>>==========>> Needle Exchange Cuts Risks, Study Finds. Los Angeles Times, 12 Jan 1994, A3 Sheryl Stolberg
Two articles in the January 12 issue of the Journal of the American Medical Association show the effectiveness of needle exchange programs in preventing the spread of AIDS. One study was based on an underground needle exchange program in San Francisco. Researchers interviewed 5,644 intravenous drug users and found that, as participation in the needle exchange program increased, the percentage of addicts who shared syringes went down, from 66% before the exchange began four years ago to 35% now. The study also found that the average age of intravenous drug users has increased, while the number of first-time drug users has dropped dramatically. The authors state that this shows that needle exchange programs do not promote drug abuse as critics of these programs suggest. Doctors in Los Angeles have called for legalization of the small underground program that exists in their city. For the last two years, Governor Pete Wilson has vetoed legislation that would have made needle exchange programs legal. Needle exchange programs are legal in New York City. While these, and other similar studies, show that needle exchange programs cut down risky behavior of drug addicts they have not yet demonstrated that they have decreased the spread of AIDS. The researchers for the New York program feel that the exchange program has helped stabilize the spread of AIDS among drug users and expect to be able to show data in the next few years that it actually decreases the incidence of AIDS. <<<========<<
>>>>>==========>> AIDS and Population "Control" Scientific American, February 1994, Page 124 Essay by Gerard Piel
The author gives his ideas on population control. He remarks that during the three decades including World War I and II, these wars killed ten percent of all the people living then and did not make a big dent in the long run. Similarly, as bad as predictions are for the AIDS epidemic in Africa (estimated at 200 million people infected by 2010) he states that this will not be the solution to the population problem of Africa. Instead he argues that, just as the industrialized nations went through an explosive increase in population in the first phase of the industrial revolution and then moved to a phase of essentially zero growth, the same will be the case for the non- industrialized countries which he characterizes as being in the first phase of the industrial revolution. There are a lot of interesting statistics comparing birth rates, death rates etc. for the various countries. <<<========<<
>>>>>==========>> Understanding Probability and Statistics. A Book of Problems by Ruma Falk A K Peters, ISBN 1-56881-018-0, $39.95
This is a nice collection of problems that can be used to enrich an elementary probability or statistics course. Part 1 has problems to be worked out in the usual way and Part II consists of multiple choice problems. Answers are provided to all questions, and some of the more interesting problems have extensive discussions included with the answers. High school mathematics and elementary combinatorics suffice to work most of the problems. The author's expertise in psychology and special interest in the understanding of probability and statistical concepts shows up in many of the problems and their discussion. It is a pleasure to find such a nice integration of probability and statistics, making it possible for both probabilists and statisticians to take credit for developing such a fascinating subject. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! CHANCE News (1 January to 21 January 1994) !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Please send suggestions to: jlsnell@dartmouth.edu >>>==========>>|<<==========<<< >>>==========>>|<<==========<<<