• Topic: Statistical Inference of Association Parameters for Stratified Bilateral Correlated Data

• Speaker: Dr. Ma Chang Xing (Associate professor in Department of Biostatistics, University at Buffalo)

• Time:14:00 - 15:00,  December 28 (Friday)

• Venue: T2-202
  

In clinical researches such as ophthalmologic (or otolaryngologic) studies, bilateral correlated data naturally arise when investigators collect information from paired organs (or body parts). Since the measurements from such paired observations are usually highly correlated, appropriate data analysis requires accounting for the intra-class correlation. Specific statistical methodology dealing with bilateral correlated data is a well-studied topic in the past several decades, both in terms of developing new analysis and efficient algorithm for computations. In some analyses, the center-effect or confounding-effect could lead to imbalance with treatment arms, making it necessary to adjust for stratification/confounding factors in the data analysis. Therefore, either ignoring the intra-class correlation or confounding effect may lead to biased results.  In this dissertation, we first derive several approaches for testing homogeneity of risk difference for stratified bilateral correlated data under the assumption of equal correlation. Further, we propose several procedures for testing equality of difference of two proportions in a stratified bilateral design. Then, we extend the methodologies to construct a variety of confidence intervals (CIs) for estimating difference in the proportions of responders. In addition, we propose three approaches for testing homogeneity of odds ratio for stratified bilateral correlated data under the assumption of equal correlation. The performance of the proposed test methods and CI estimations is evaluated by Monte Carlo simulations. We use real data to illustrate the practical implementation of the proposed methodologies as well.