Applied and Mathematical Statistics

Instructors: David Feldman and Chris Petersen
Offices: Second floor, Turrets Annex and Second floor, Davis Center
Phone: x249 and x410
Mailing List: stats at h0rnacek dot coa dot edu
Office Hours: By appointment

Catalog Description

This is an intensive course that investigates the three main frameworks for modern statistical analysis: parametric analysis, Monte Carlo and related computational methods, and Bayesian statistics. Topics during the term will include research design, power analysis, meta-analysis, parametric statistics, randomization tests, probability models, likelihood, and maximum likelihood. Students will learn how to apply key tools from each analytic framework as well as gain a theoretical understanding of each tool and an overall sense of the mathematical structure of statistics. Students will also gain proficiency in R or a related statistics platform. Students will meet at least weekly with the instructors and will also convene regularly without the instructors to collaborate on problem sets and for further discussion. Evaluation will be based on regular problem sets and participation in seminar-style class meetings. Permission of instructors required. Calculus III and Linear Algebra are strongly recommended as pre- or co-requisites.

Intermediate/Advanced. *QR* No lab fee.

Goals
  1. Be able to apply a range of modern statistical techniques to problems in a variety of settings. This includes parametric and non-parametric analysis, Bayesian methods, and if time permits, computational techniques.
  2. Understand the mathematical structure of modern statistics.
  3. Gain experience using the statistical package R.

Course Details Our text for this class is Larry Wasserman, All of Statistics: A concise Course in Statistical Inference, Springer, 2004. We are uncertain of the exact pace of the class, but our goal is to cover chapters 1-14. To do so, we will move through the first five chapters at a brisk pace, but will then slow down.

This class will meet at least weekly. The students should meet as a group at least one or two additional times each week. There will be six or seven challenging problem sets. Evaluation will be based on problem sets and participation in seminar-style class meetings.