364 Statistics for Lawyers and Policy Analysts Spring 2000

Dick Ippolito

Office: 319

Phone: (703) 993-8243

Email rippolit@gmu.edu

Office hours: Wednesday: 9AM-NOON; 1-2PM and 4-6 PM

Other days: by appointment

Required text: Moore and McCabe, Introduction to the Practice of Statistics3

^{rd}Edition, WH Freeman, 1999. A fancy statistics calculator is not needed in the course, but I assume that everyone has a calculator of some kind. The most useful function is y^{x}.

Grades:If there are not too many students then grades will be determined by class performance and the final exam. I will assign problems from the text every week, and distribute one or two practice take-home exams during the semester with answer sheets.Overview:

The Moore and McCabe text assumes no background in statistics, and is very readable. I expect each student to carefully study each assigned chapter before class and to reread each chapter after the material is completed. It will be important to take careful class notes, to rewrite notes after each class, and to make sure that notes jive with the text material. Assignments will be made after each chapter for students to work through (answers are in the back of the book).

This is a hands-on course. Statistical calculations will be required throughout the semester. At the end of the course, you should have a reasonably good handle on the basics of statistical inference, a reasonable command of the vocabulary used in statistics, and some sense about the kinds of legal and policy issues that turn on statistical inferences. The text is laden with good examples, and the course will be intensive in applications.

We will cover as much of the text as feasible over the semester. I will cover Part I quite quickly. Part II ‘Probability and Inference’ is the core of statistics as you will use it, and I will spend as much time on this material as is required for you to assimilate it. I will pick and choose among topics from Part III, but will definitely cover the least squares chapters.

ScheduleThe following is a listing of lecture numbers and subjects. The reading assignments refer to Moore and McCabe. Please note: the schedule is tentative. I will either slow or advance the pace depending on the rate at which the students assimilate the material.

1. Statistical distributions; how to summarize them and interpret what they are ‘saying;’ how to read a normal distribution probability table. Chapter 1

2 Statistical relationships – the basics (scatter plots, correlation, least-squares regression, causation issues, and so on.) Chapter 2.

3. Thinking about how data are produced. The concept of a random sample. Chapter 3.

4&5. Probability – the basics: the study of randomness, probability models, random variables, the rules for means and variances, general probability rules. Chapter 4.

6&7. The basics of inference. The binomial distribution. What is it? When to use it? How to read the tables. The sampling distribution of a sample mean, the central limit theorem; how a binomial becomes like a normal distribution when the number of trials becomes large. Chapter 5.

8&9. Inference continued: we will become familiar with the jargon of probability and inference: confidence intervals, tests of significance, the power of a test, the null hypothesis, one sided tests, two sided tests, p-values, type I and type II errors. Chapter 6.

10. Comparing two distributions and deciding if they are different. Chapter 7.

11. Comparing proportions to decide if they are statistically different from each other. Chapter 8.

12. Looking at relationships in a table and deciding if there is a statistically significant relationship between two variables using the chi-square statistic. Chapter 9.

13. A simple regression. How does it work? how are the results interpreted? How is the ‘goodness of fit’ evaluated? An introduction to multiple regressions. Chapters 10 and 11.

14. Open. I will use this lecture to catch up if we fall behind. Otherwise, I will look at some additional applications.

15. EXAM May 13, 1PM open book and open notes.