Instructor: Philip B. Stark. Office Hours
Course Description
Statistics 151 is an advanced undergraduate course in statistical models, emphasizing linear models.
The text is Statistical Models: Theory and Practice by David A. Freedman, Cambridge University Press (2005).
The course makes heavy use of linear algebra. You might want to supplement your reading with Matrix Computations by G.H. Golub and C.F. Van Loan, Johns Hopkins University Press (1983) or another book on linear algebra. See p.40 of Statistical Models: Theory and Practice for more recommendations.
The course also expects you to be or become reasonably fluent in R, a mathematical/statistical programming language and environment. There will be homework assignments roughly every week. There are nine lab assignments during the term. Course grades will be based on homework, lab assignments, a group term project, and a final exam. I have not yet decided what the relative weights will be. Grades will not be "curved." I will not offer an alternative time for the final. If you cannot attend the final, take a different course.
Show your work on all assignments. Homework must be legible, preferably typewritten. If the reader cannot read your homework easily, it will be returned to you ungraded. Labs must be typed, preferably typeset using TeX or LaTeX (this is a good opportunity to learn how to write a scientific paper). Homework and labs will not be accepted after the due date.
I encourage students to work together on homework, labs, and--of course--group projects. However, you must write up your own work. On the other hand, final exams are strictly individual work, and collaboration and other forms of cheating will not be tolerated.
Topics
Experiments and observational studies, regression, residuals versus error terms, matrix algebra, standard errors, generalized least squares, normal theory of regression, the F-test, path models, inferring causation from regression, response schedules, types of veriables, maximum likelihood, probit and logit models, latent variables, the bootstrap for estimating bias and variance.
Labs are from the textbook by Freedman. Data for the labs is online here.
Rough Course Schedule
The schedule for the term project is in a separate document.
The following is rough and subject to change. Be sure to read ahead in the book! You will get much more from the lectures if you have at least skimmed the material first. Homework problems are labeled <chapter>.<exercise set>.<problem number>. Homework is due in section the week after it is assigned.