Industrial Statistics STAT 498A



Spring Quarter 2009 Course Announcement

STAT 425: Introduction to Nonparametric Statistics

Instructor: Fritz Scholz

1 Coverage: This course focuses on nonparametric and/or distribution free statistical methods. The coverage follows the text by Lehmann with chapter headings:

1. Rank tests for comparing two treatments,

2. Comparing two treatments or attributes in a population model,

3. Blocked comparison for two treatments,

4. Paired comparison in a population model and the one-sample problem,

5. The comparison of more than two treatments,

6. Randomized complete blocks,

7. Tests of randomness and independence.

The k-sample Anderson-Darling test and its extension will be added in the discussion of chapters 5 and 6. The statistical platform R will be used extensively to carry out these tests and explore their power and large sample approximation properties.

While the above addresses mainly rank based methods we will also discuss nonparametric confidence bounds for population quantiles based on order statistics.

Prerequisites: A basic introductory course of probability and statistics (e.g., Stat 342 or Stat 390 or my consent) covering testing (type I and type II error, significance level, p-value, power), confidence intervals, estimation, the essence of the central limit theorem and knowledge or exposure to the statistical analysis platform R. If deemed appropriate a short review of R will be given.

Text: Nonparametrics: Statistical Methods Based on Ranks (Paperback, 2006)

by Erich L. Lehmann, Springer Verlag (required).

A First Course in Statistical Programming with R

by W. John Braun and Duncan J. Murdoch, Cambridge University Press 2007 (optional).

Lecture slides and other materials will be provided in pdf file format.

Grade: Based on homework only.

Time & Place: Tuesday, Thursday 9:00-10:20 in Padelford C-301

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download