Nonparametric and Semiparametric Methods of Data Analysis

This elective course dealing primarily with nonparametric methods follows on the one-year general linear model (GLM) sequence with a variety of statistical methods that may be used when the assumptions of the classical model cannot be met. Randomization methods are introduced and linked to concepts of resampling and bootstrapping. In addition, one-, two-, and c-sample methods are covered, along with nonparametric procedures for association, regression, and goodness-of-fit. Where appropriate, extensions are made to nonparametric procedures used in survival data analysis. Semi parametric methods of data analysis are introduced, with emphasis on survival data. Cox regression and proportional hazards modeling is described. The course will incorporate individual and team project reports and brief oral presentations to enhance student teamwork and communication skills.