For Thursday, January 10, please read Chapters 1, 2, and 3 of Freedman/Pisani/Purves and be prepared to answer the questions in the "Review Exercises" of these chapters.  These chapters should be easy to read on your own but I will answer any questions you may have as best I can. Despite their simple writing, they demonstrate beautifully some of the problems in creating statistical studies and drawing valid conclusions from statistics. (Freedman/ Pisani/ Purves is an extraordinary text. It does not cover all the statistical tools we will need -- some, together with illustrative cases will be drawn from Finkelstein/Levin -- but it is insightful with lucid expositions.) I will begin lecturing with material in Chapters 4 and 5 on the the normal curve.

Pay some particular attention to problem 11 on pp 54 - 55 of FPP. There is a very obvious answer given in the Instructor's Manual but it may not be the last word. Suppose  that you were defending the people charged with rigging the exam. Are there ways that you could counter (or at least attempt to counter) what seems like clear evidence of guilt? Is the evidence really that clear or is there more that we need to know? (What do grade distributions look like in some courses?  I haven't read  the reference given in the footnote (actually, endnote) and would be very interested if any of you have.  In particular, I don't know if there was a case that went to trial, and if there was, what evidence was presented. If any of you can find it you might want to share it with the class.)

When we come to counting problems I will briefly discuss a case from the years when New York State was running a "superfecta".  In this a person chose the first four horses in each of  four eight-horse harness races of a given day. If the horses he chose in the first race came in, then all his winnings were bet on the horses he chose in the second race, and so on through the four races. If he (or she, as it turned out) had the good fortune to pick correctly in all four races then his/her bet paid off very handsomely, something that the commissioners hoped would be a great attraction.  The superfecta was discontinued. Can you see any problem with it? (Should the commissioners have foreseen any problems?)

For Tuesday, January 15:
Please read through Chapter 6 of FPP, be prepared to answer the questions at the ends of the chapters and hand in written answers to questions 13, 14, and 15 on pages 107 and 108. (Note that question 15 asks you to "Comment briefly on the analysis." In all your answers, please use the least number of words that will make it unequivocally clear that you understand what you are talking about!

January 24, 2002:
For the next three metings we will be discussing the material in Chapter 4 of Finkelstein and Levin.  This may be difficult reading for non-mathematicians; some of it goes well beyond what a lawyer would need to know. Nevertheless, please look carefully at the cases cited and try to answer the questions in this chapter.  There is a discussion at the end of the book.  Please try to formulate your own responses before looking at what the authors say. (Do you agree?) We will later return to FPP, but some of the probability distributions with which you should become faimiliar are not covered there.