SCI~199Y: ``Lies, damned lies, and statistics"

from the First Year Seminar Handbook


The title of the course is a quotation attributed to Disraeli, a British statesman of the nineteenth century. Does it still apply today? This course will consider how statistics and statistical thinking get used (and abused) in a variety of activities, including polling, public health, marketing, advertising, lotteries. Some questions that will be addressed are: Why do newspapers report a ``margin of error'' for poll results, and what does it mean? How can graphs and charts provide information (or misinformation)? What makes a good graph? How do new cancer drugs get tested, and why doesn't the same protocol work for AIDS? How do studies on mice get extrapolated to humans, and do the results make any sense? What is quality control, and why is it currently so fashionable in North American industry?

Examples of `current' interest in which statistical thinking is a relevant component


What's required in this course -- what you'll do


What's offered in this course -- what I'll do


How will the course be organized

We will consider between five and ten themes for discussion, background, and further investigation. These will be selected from current topics according to interests of the class participants, and may consider any of the topics already mentioned in this outline, and any of the following: statistics in sports (is there a hot streak in basketball, how are tennis players seeded, what would have been the outcome of the 94 World Series, is figure skating judging fair, ...); how to assess risks (should bicycle helmets be required, do power lines cause cancer, ...); probability in everyday life (the Monty Hall problem, picking winning lottery numbers,..); environment vs. heredity ( The Bell Curve, twin studies, genetics and crime); ... Please feel free to suggest other topics for consideration, and be warned that you will be asked about your preferences for choice of topics.

The first theme which will be treated in depth is the use of pictures (usually graphs) to display numerical information. Because numbers are often difficult to assimilate, especially if there are a lot of them, and because statistics is often considered to be very mysterious, many popular accounts in newspapers, magazines, and so on summarize information with some type of graphic. This can be done well or badly, as we will see, and there is a scholarly field of inquiry into how to do it well.


Sources of information

More specific reading lists will be provided for each theme.

Required for next week