3.0

Average Course Rating

This is a first course in nonparametric statistics. Some basic knowledge of elementary statistics will be assumed, for example, you should know the structure and concepts behind a hypothesis test and the meaning of a p-value. You will need a solid understanding of probability, especially the notions of ordered statistics and exchangeability. Single sample statistics: nonparametric confidence interval for the median, estimation of population CDF and percentiles, independent sample tests: two-sample permutation test, Wilcoxon rank-sum test (small and large sample versions, with tie adjustments), Kruskal-Wallis ANOVA test, paired and blocked designs: Wilcoxon signed-rank test (small and large sample versions, with tie adjustments), Friedman's randomized block ANOVA test, permutation test for correlation and slope, Spearman's rank correlation, permutation test for contingency tables, Fisher's exact test for 2x2 contingency tables. Nonparametric bootstrapping: the basic bootstrap method, bootstrap estimate for MSE, bootstrap variance and bias. Other topics as time permits.