Emerson Statistics

Lemmon Creek Waterfall Camp - photo by Scott taken May, 2009

 Emerson Statistics Home  | UW Courses  | Online Courses  | Supplementary Materials  | RCTdesign.org


Scott S. Emerson, M.D., Ph.D.
Professor Emeritus of Biostatistics
Department of Biostatistics
University of Washington

 

Emerson Statistics Home

Courses
Teaching Statistics
RCTdesign / S+SeqTrialTM
Technical Reports

Scott's CV
Contact Scott

Links
UW Biostatistics
Scott's UW Faculty page
RCTdesign.org
uwIntroStats.org

Online Courses at www.EmersonStatistics.com

Scott is in the process of re-packaging materials (syllabi, handouts, presentations, homeworks, exams, etc.) into "Online Courses". It is envisioned that a user could review these materials to gain mastery of the topics, much as one would taking a formal course. The lectures have been modularized to a greater extent than would be in a typical university course.

Introductory Applied Biostatistics. This course provides an introduction to the statistical analysis of data. The major objectives of this course are to explore the ways in which statistical methods can be used to address scientific questions, to present simple data analysis methods, and to teach a general approach to a data analysis problem. To those ends, this course will stress the general abstraction of descriptive and inferential statistics to address a scientific question. We primarily address methods for the setting of one response variable and one grouping variable. This includes one and two sample problems, one way analysis of variance, and simple regression. Late in the course we address stratified analyses.Topics covered will include definition of common descriptive techniques, estimation and testing for continuous, discrete, and censored response variables in parametric models, and semiparametric and nonparametric alternatives to those tests, including Monte Carlo methods. Emphasis will be placed on the similarity among the various forms of analyses. The materials for this course are primarily re-packaging of the handouts and lectures from Biost 514 / 517 at the University of Washington, especially as taught in fall 2012.

Introductory Applied Regression. This course builds on the material covered in Introductory Applied Biostatistics, and thus it is assumed that a student is familiar with the basic principles of descriptive and inferential statistics related to means, proportions, and survival estimates in one and two sample problems, as well as the descriptive (and some inferential) statistics related to simple linear regression and correlation. It is also assumed that you have some familiarity with the most basic aspects of confounding and effect modification. This course focuses on multiple regression inference on means (linear regression), odds (logistic regression), rates (Poisson regression), and hazards (proportional hazards regression). This course will again stress the general abstraction of descriptive and inferential statistics to address a scientific question. Emphasis will be placed on the similarity among the various forms of regression analyses. The materials for this course are primarily re-packaging of the handouts and lectures from Biost 515 / 518 at the University of Washington, especially as taught in winter 2015.

Theory of Linear Predictors. The primary goal of this course is to develop the theory that is the basis for the analysis of data using regression models that involve linear predictors. Emphasis will be placed on that theory which is crucial to the application of linear regression analysis to a dataset and the theory that generalizes to other forms of regression. That is, we are at all times interested in the theoretical results that allow the greatest generalization of regression analysis. Hence, although we will typically develop theory based on strong parametric assumptions (i.e., when pretending that the regression models accurately model the data), we will ultimately want to be able to describe the behavior of those parametrically derived estimates in other settings (i.e., to be able to use theory to investigate the extent to which a regression method might be able to model a scientific question in a more distribution-free manner). Less attention will be paid to the theoretical results that are generally only of interest when programming software to perform analyses. It is assumed that students are familiar with matrix algebra and the basic properties of the multivariate normal distribution. The materials for this course are primarily re-packaging of the handouts and lectures from Biost / Stat 533 at the University of Washington, especially as taught in spring 2014.

Mathematical Statistics. This course covers the mathematical theory behind standard statistical inferential techniques. It is targeted to graduate students majoring in statistics, biostatistics, and other disciplines requiring an understanding of statistical theory. The course starts with a review of the probability theory that is the basis for that inference, especially laws of large numbers, asymptotic central limit theorems, Slutsky's theorem, the continuous mappiing theorem, and the delta method. We will then cover both frequentist and Bayesian methods of point estimation, interval estimation, and hypothesis testing. In particular, we consider method of moments, maximum likelihood, Cramer-Rao, Rao-Blackwell, Lehmann-Scheffe, Neyman-Pearson, Karlin-Rubin, and decision theory. The materials for this course are primarily re-packaging of the handouts and lectures from Stat 512 at the University of Washington, especially as taught in fall 2015, and from Stat 513 at the University of Washington, especially as taught in spring 2016.

Case Studies in Data Analysis. This course provides a practicum in the statistical analysis of data. Emphasis is placed on the analysis of data to answer scientific questions. Thus the major objectives of this course are to explore the ways in which various statistical methods can be used to address scientific questions, gain experience in the preparation of complete statistical analysis plans, develop a general approach to a data analysis problem, gain experience in the generation of tables and figures that communicate suitable description of the data, and gain experience in the presentation of statistical inference to an applied audience. The materials for this course are primarily re-packaging of the handouts and lectures from Biost / Stat 579 at the University of Washington, especially as taught in fall 2011, however, data projects used in many different applied biostatistics classes are also used.

 
 
©2005-17 Scott S. Emerson and emersonstatistics.com
email questions or comments about this site