Evaluating the Efficiency of Treatment Comparison in Crossover Design by Allocating Subjects Based On Ranked Auxiliary Variable
Term of Award
Doctor of Public Health in Biostatistics (Dr.P.H.)
Document Type and Release Option
Dissertation (restricted to Georgia Southern)
Copyright Statement / License for Reuse
Creative Commons Attribution 4.0 International
Department of Biostatistics (COPH)
Committee Member 1
Committee Member 2
Committee Member 3
Committee Member 3 Email
“Randomized control design is the gold standard” in the design of experiments which focus on treatment comparisons in health related fields. The validity of statistical inference depends heavily on proper randomization processes. Several approaches for treatment allocation have been studied to ensure the validity of statistical inference, such as complete randomization, stratification, block randomization and minimization. However, even with proper randomization, we could have unbalanced characteristics of subjects among treatment groups. In addition, for those studies on chronic disease, crossover designs using Latin squares provide a solution in the experimental design stage to balance the characteristics of subjects among the treatment groups. In this dissertation, we introduce a method based on ranked auxiliary variables for treatment allocation, which is inspired by ranked set sampling (RSS), for crossover designs using Latin squares. We also evaluate the improvement in efficiency of the proposed method. Our simulation study reveals that the proposed method provides a more powerful test compared to simple randomization under equivalent sample sizes. This will translate to a reduction in the number of replicates needed in the crossover design using Latin squares.
Huang, Yisong, "Evaluating the Efficiency of Treatment Comparison in Crossover Design by Allocating Subjects Based On Ranked Auxiliary Variable" (2016). Electronic Theses and Dissertations. 1450.
Research Data and Supplementary Material