666 NONPARAMETRIC STATISTICAL METHODS

DEPARTMENT OF STATIS TICS
COLLEGE OF ARTS & SCIENCES
330.972.6886 (TELEPHONE)
Course Description
3470:666
Prerequisite:
Nonparametric Statistical Methods
4 credits
3470:461/561 Applied Statistics or equivalent.
Course Description: Theory and practice using techniques requiring less restrictive
assumptions. Nonparametric analogues to t- and F-tests, ANOVA, regression and
correlation. Computer applications.
Course Topics:
Introduction
Parametric vs. nonparametric procedures
Measurement scales
Review of hypothesis testing
Power, efficiency
Advantages and disadvantages of nonparametric statistics
Analysis of Categorical
Binomial test
Chi-square test for one-way layout
Two-way contingency tables
Independence hypothesis
Homogeneity hypothesis
Fisher’s exact test
Goodness-of-Fit Tests
Chi-square tests
One-sample Kolmogorov-Smirnov test
Assessing Normality
One-Sample Tests of Location
Sign test
Introduction to ranking
Wilcoxon signed-rank test
Two-Sample Tests of Location
Mann-Whitney U-test (Wilcoxon rank sum test)
Median test
Assessing Homogeneity of Variance
Two-Sample Tests of Dispersion (Scale)
Ansari-Bradley test
Moses test
Jackknife procedure
Procedures for Paired Data
Sign test
Wilcoxon signed-rank test
Nonparametric Analogues of One-Way ANOVA
Median test
Kruskal-Wallis test
Multiple comparisons
Nonparametric Analogues of Randomized Block Designs and Two-Way ANOVA
Friedman’s test
Multiple comparisons
Mack-Skillings test
Other tests for two-way layout
Nonparametric Simple Linear Regression and Correlation
Estimation
Hypothesis testing
Spearman’s Rho (р)-rank correlation
Fall 2014