Course: Biostatistics 200B Winter 2000, MWF 9-10am

Instructor: R.I. Jennrich

This course will integrate the use of specific statistical software (SAS) and the analysis of specific data sets with the development of statistical and data analytic theory for regression, analysis of variance, nonlinear regression, and applications to maximum likelihood analysis and robust methods.

It will cover Chapter 3 and the second half of my book An Introduction to Computational Statistics: Regression Analysis.

Applying simple linear regression. Model formulation. Weighted regression. Diagnostic plots. Searching for linearizing, stabilizing, and normalizing transforms. Using weights and transforms.

The general linear model. The multivariate normal distribution. Projection. The chi-squared theorem. The fundamental F-test. Generalized least squares and weighted regression.

Analysis of variance and covariance. The one way model. Dummy variables. Software (PROC GLM). The two way additive model. General two way and higher order models. Analysis of covariance. Transformations and model diagnostics.

Nonlinear regression. Nonlinear least squares. The Gauss-Newton algorithm. Software (PROC NLIN). Analysis of growth curves. Statistical inference.

Maximum likelihood analysis and robust estimation. Iterative reweighting. Exponential family maximum likelihood. Robust regression.