Topics in ordinal logistic regression and its applications
MetadataShow full item record
Sample size calculation methods for ordinal logistic regression are proposed to test statistical hypotheses. The author was motivated to do this work by the need for statistical analysis of the red imported ﬁre ants data. The proposed methods use the concept of approximation by the moment-generating function. Some correction methods are also suggested. When a prior data set is available, an empirical method is explored. Application of the proposed methodology to the ﬁre ant mating ﬂight data is demonstrated. The proposed sample size and power calculation methods are applied in the hypothesis testing problems. Simulation studies are also conducted to illustrate their performance and to compare them with existing methods.
Kim, Hyun Sun (2004). Topics in ordinal logistic regression and its applications. Doctoral dissertation, Texas A&M University. Texas A&M University. Available electronically from