Journal of the Korean Data and Information Science Society
- Volume 17 Issue 1
- /
- Pages.245-257
- /
- 2006
- /
- 1598-9402(pISSN)
A Comparative Study of Assessing Average Bioequivalence in $2{\times}2$ Crossover Design with Missing Observations
-
Park, Sang-Gue
(Department of Statistics, Chung-Ang University)
;
- Choi, Ji-Yun (Department of Statistics, Chung-Ang University)
- Published : 2006.02.28
Abstract
A modified Anderson and Hauck(1983) test for analyzing a two-sequence two-period crossover design in bioequivalence trials is proposed when some observations at the second period are missing. It is based on the maximum likelihood estimators of average bioequivalence model and designed for handling missing at random(MAR) situation. The performance of the proposed test is compared to other tests using Monte Carlo simulations.
Keywords
- Average bioequivalence;
- Incomplete crossover design;
- Maximum likelihood estimation;
- Missing at random