Chance: A Course About Current Chance News

J. Laurie Snell

Dartmouth College

I was asked to talk about some of the successes and failures that we have had at Dartmouth College trying to make students in statistics more aware of the interdisciplinary nature of their subject. Naturally, I would rather speak about our successes than our failures.

We have had two years' experience with colleagues at other colleges developing and teaching a new introductory course called Chance. This project has been supported by the National Science Foundation's Undergraduate Development Program and the New England Consortium for Undergraduate Science Education.

The Chance course has been taught or is being taught at Dartmouth, Middlebury, Grinnell, Princeton, Spelman, and the University of California at San Diego. While it has not been taught at the level that we have been talking about in this symposium, it could be, and certainly some of the approaches we use would enrich more advanced undergraduate and graduate statistics courses.

Chance deals with current events that rely on statistical and probabilistic concepts. We got the idea for the Chance course from reading Chance magazine. This magazine was started by Springer-Verlag in 1988 and now is a joint publication of Springer and the American Statistical Association. One of its founders, Stephen Fienberg, is participating in this symposium.

In its brief existence, Chance has attracted some of the leading workers in probability and statistics to write articles that present important statistical problems in current affairs in the spirit of Scientific American. Topics that have been covered include:

  • Scoring streaks and records in sports;

  • Health risks of electric and magnetic fields;

  • The effectiveness of aspirin in preventing heart disease;

  • Statistics, expert witnesses, and the courts;

  • The undercount problem in the 1990 U.S. Census;

  • Extraterrestrial communication;

  • The use of DNA fingerprinting in the courts;

  • The story of Deming and quality control;



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 86
Modern Interdisciplinary University Statistics Education: Proceedings of a Symposium Chance: A Course About Current Chance News J. Laurie Snell Dartmouth College I was asked to talk about some of the successes and failures that we have had at Dartmouth College trying to make students in statistics more aware of the interdisciplinary nature of their subject. Naturally, I would rather speak about our successes than our failures. We have had two years' experience with colleagues at other colleges developing and teaching a new introductory course called Chance. This project has been supported by the National Science Foundation's Undergraduate Development Program and the New England Consortium for Undergraduate Science Education. The Chance course has been taught or is being taught at Dartmouth, Middlebury, Grinnell, Princeton, Spelman, and the University of California at San Diego. While it has not been taught at the level that we have been talking about in this symposium, it could be, and certainly some of the approaches we use would enrich more advanced undergraduate and graduate statistics courses. Chance deals with current events that rely on statistical and probabilistic concepts. We got the idea for the Chance course from reading Chance magazine. This magazine was started by Springer-Verlag in 1988 and now is a joint publication of Springer and the American Statistical Association. One of its founders, Stephen Fienberg, is participating in this symposium. In its brief existence, Chance has attracted some of the leading workers in probability and statistics to write articles that present important statistical problems in current affairs in the spirit of Scientific American. Topics that have been covered include: Scoring streaks and records in sports; Health risks of electric and magnetic fields; The effectiveness of aspirin in preventing heart disease; Statistics, expert witnesses, and the courts; The undercount problem in the 1990 U.S. Census; Extraterrestrial communication; The use of DNA fingerprinting in the courts; The story of Deming and quality control;

OCR for page 86
Modern Interdisciplinary University Statistics Education: Proceedings of a Symposium Randomized clinical trials in assessing risk; The role of statistics in the study of the AIDS epidemic; and Evaluating a needle exchange program to prevent the spread of AIDS. As you can see, Chance magazine is a wonderful resource for teaching statistics and should certainly be in every college and university library. Our original idea was to choose three or four of these topics and to try to understand the problems studied, the statistical or probabilistic issues involved, and the techniques that have been used to solve these problems. We soon found that many of these topics, as well as other equally interesting Chance course topics, were regularly reported in daily newspapers such as the New York Times, the Washington Post, and the Los Angeles Times and in weekly journals such as Science, Nature, and the New England Journal of Medicine. Thus we have used a blend of current news and articles like those in Chance magazine for our course. The Chance course is not meant to be a substitute for a basic statistics course or for an introductory probability course. Its purpose is just to make students better able to understand and make critical judgments about reports in the media that involve probability or statistical concepts. The first thing we discovered was that we ourselves needed some background information on topics as technical as DNA fingerprinting, and we needed some help finding current news articles. For this reason we established the CHANCE database on gopher on the Internet to make available to others our experiences in teaching current topics as well as to provide background material for these topics. We also started an electronic newsletter available by e-mail called Chance News. Chance News abstracts articles from the last couple of weeks' news that might be useful in teaching a course like Chance. Anyone interested in obtaining this Chance News can request it by electronic mail from dart.chance.edu. You can also obtain current and back issues of Chance News from our CHANCE gopher database. I think that the ability to share information on the Internet will greatly enhance the teaching of statistics in the future. For example, let me show how easy it is to get our Chance News from gopher. The procedure is a little different for each machine, but I will illustrate it in terms of a Unix machine. Assuming that gopher has been installed by your system manager, and this is the case at most institutions, you just type gopher chance.dartmouth.edu You will then see a list of items, one of which is CHANCE Database. You move the pointer to this and then you will see:   CHANCE Database —> 1. Search CHANCE database <?>   2. Welcome.   3. Chance_Course/

OCR for page 86
Modern Interdisciplinary University Statistics Education: Proceedings of a Symposium 4. Chance_News/ 5. Chance_Profiles/ 6. Current_Chance_News. 7. Fair_Use/ 8. Other_Gopher_Servers/ 9. Texts_of_Articles/ 10. Tools/ If you want all the articles on a given topic such as SAT scores, you can use the search option. If you want the current Chance News, just move the pointer to Current_Chance_News and you can scan through the abstracts of recent articles. For our own class we browse through these abstracts to find one that might be interesting to discuss in class. For example, we might choose the recent New York Times article about the study showing that homosexuality might be partially a genetic trait. Here is the abstract of this article: Report suggests homosexuality is linked to genes. New York Times, 16 July, 1993, A, 1. Natalie Angier An article in the 16 July issue of Science reports on a study led by Dean Hamer that suggests that sexual preference has a genetic component. The study began by looking at the family histories of 114 homosexual men. The researchers discovered a higher than expected number of gay men among the men's maternal uncles and male cousins who were sons of their mother's sisters. This suggested a gene or genes that pass through the maternal line and therefore through the X chromosome. This led them to study the genetic material from 40 pairs of brothers who were gay. They found that in 33 of the pairs the brothers displayed the same cluster of five markers bunched into a small region on the X chromosome. Since there should be a fifty percent chance that two brothers share the same allele by descent, this finding is regarded to be significant. The authors are careful to say that this study will have to be replicated and, if correct, is surely only one part of the answer, and so on. A similar finding two years ago that suggested an anatomical difference between the brains of gay and straight men received a lot of publicity and led to heated debates, mixed emotions, and so forth. We would then provide the students with the full text of the article to read and ask the class to break up into groups and discuss for 20 minutes or so three or four questions concerning the study. For example, we might ask: How convinced are you by this study that there is a genetic component to homosexuality?

OCR for page 86
Modern Interdisciplinary University Statistics Education: Proceedings of a Symposium Does this study tell you anything about the proportion of the general homosexual population or the non-homosexual population that have these "gay genes"? What further studies will be necessary to clarify the extent of a genetic component of homosexuality? What legal and other implications would establishing a genetic connection for homosexuality have? When students discuss current issues in small groups they get totally immersed in the subject, and the most successful discussion questions develop a high level of noise and excitement. We are constantly surprised by the new and interesting ways that students look at a problem when they have not first been told how they should view it. It is an interesting challenge for the instructor to try to respond to the diverse ideas developed by the groups. If we wanted an article that would lead to a more extensive discussion, we might choose from the Chance News the article on the recent supreme court decision on determining criteria for admission of scientific evidence in the courts. We could discuss this in terms of the current controversy over the use of DNA fingerprinting. Here the students can see arguments about statistical conclusions in a life-and-death situation. By using the search capabilities of our database we could search for previous articles on DNA fingerprinting, and we would find reports on a number of legal cases in which the following statistical problem played a key role. A locus used in forensic DNA fingerprinting has a number of repeats of a particular sequence of DNA letters. For example, AGGAGGAGGAGG might be a particular allele for a locus having four repeats of AGG. For the loci used, there is a great deal of variation between individuals in the number of repeats. These numbers can vary from a small number to several thousand. Each individual has two such alleles, one obtained from the mother and one from the father. When there appears to be a match for these alleles between the DNA of the suspect and the DNA provided as evidence, how do you evaluate the chance of such a match for a randomly chosen person in the population? The Hardy-Weinberg law says that if our population has random mating and no immigration or migration, then the alleles in an individual can be regarded as the result of independently choosing two alleles according to the proportions of each allele in the whole population and we can use the infamous product law for independent events. The FBI and commercial companies that do this kind of analysis invoke Hardy-Weinberg applied to several loci to get remarkably small probabilities for a match. How can they justify this independence? Obviously, if they were looking at alleles that determine the color of your skin, they could not. But the loci used for DNA fingerprinting are taken from the "junk" part of the DNA that is not thought to serve any particular purpose. If it is just part of your junk DNA, why should Hardy-Weinberg not apply? After all, your parents hardly decided to marry based on the junk part of your DNA that has no known purpose. But even then, might different ethnic groups have different distributions for the alleles? What statistical tests can be used to justify the Hardy-Weinberg assumption? What about the measurement errors in establishing the number of repeats? These are the kinds of questions that are being raised in the courts in trials

OCR for page 86
Modern Interdisciplinary University Statistics Education: Proceedings of a Symposium using DNA fingerprinting. If you ask students to imagine that they are testifying in court, they will get a good chance to think about statistics in action. There are many other potential uses of the Internet for more advanced students of statistics. For example, students might be encouraged to participate in one of the statistics discussion groups. Previous discussions for these groups are archived on gopher, and so you can again search for all discussions on a given topic. For example, if you search on the archives of STAT-L or EDSTAT-L for messages on Simpson's paradox, you will find some interesting real-life examples described and referenced. You can also get data sets and information about statistical software from the STATLIB archives, and you can even read in those archives the current IMS Bulletin significantly before it appears by mail. All this and more can be found by pointing your gopher to the electronic mail address jse.stat.ncsu.edu, where you will also find the new electronic Journal of Statistics Education.   Journal of Statistics Education ---> 1. About the Journal of Statistics Education Information Service.   2. Methods of Access.   3. Index to the Information Service.   4. The Journal of Statistics Education/   5. EdStat-L Archives/   6. Software Tools and Information/   7. Other Discussion Groups/   8. Other Services (Census, Statlib, etc)/ In item 4 you will find the first issue of this new journal, and you can browse through it to find more information about learning in groups in an article by Joan Garfield, and an interview with Fred Mosteller explaining the evolution of his work up to the present time that finds him deeply involved in interdisciplinary work in the finest sense. Robin Lock's Datasets and Stories section of the journal will be a valuable resource for new ways to introduce interesting data into our courses. If you find something you like, just print it out or even mail it to a colleague or student. Previous speakers have explained the need to get students to think about the issues that cross disciplines. One place to start is to try to explain what you read in your daily newspaper. That has been our approach, and it has been a lot of fun.