Dissertation Defense Announcement
The College of Arts and Sciences announces the final dissertation defense of
for the degree of Doctor of Philosophy
August 2, 2016 at 1:00 pm in 231, Dunn Hall
Major Advisor: E. O. George, PhD
Exact nonparametric Test for Clustered Binary Data
ABSTRACT: It is important to make minimal assumptions on the distribution of the data ,such as coming from a normal or chi-squared distribution in the computation of the p-value . Classical statistical tests such as likelihood ratio test rely on large sample approximate distributions of the test statistic to make inferences about a contingency table. These asymptotic distributions are known to be inaccurate when sample size is very small, or when data is unbalanced or sparse. The goal of this dissertation is to develop and implement exact conditional and exact nonparametric testing procedures for sparse ,unbalanced correlated clustered binary data.