Sample Size Determination for Two-stage Equivalence Test

• Main Authors: Shang-bo Ye, 葉尚柏
• Other Authors: Mi-Chia Ma
• Format: Others
• Language: en_US
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/87773637833284788175

碩士 === 國立成功大學 === 統計學系碩博士班 === 97 === In clinical research, several experimential treaments were considered and compared to a standard teratment in the pilot study. If an experimental treatment is the best in the pilot study, then it will be confirmed whether it is better than the standard control or it is equivalent to the standard control. The general sample size determination of noninferiority test or equivalent test had been studied. This paper proposed a two-stage design which selects the best of several experimental treatments and tests whether it is equivalent to a standard control or better than a standard control. The design allows early termination with acceptance of the null hypothesis for noninferiority test or equivalent test. By minimizing expected total sample size for fixed significance level and power, optimal sample size and cut-off parameters were obtained.