Statistics and Its Interface

Volume 11 (2018)

Number 3

Response-adaptive randomization using power function of hypothesis testing

Pages: 455 – 462

DOI: http://dx.doi.org/10.4310/SII.2018.v11.n3.a8

Authors

Zhongqiang Liu (School of Mathematics and Information Science, Henan Polytechnic University, Jiaozuo, China; and School of Mathematical Sciences, Zhejiang University, Hangzhou, China)

Xuefeng Han (School of Mathematics and Information Science, Henan Polytechnic University, Jiaozuo, China)

Abstract

Response-adaptive randomization (RAR) procedures have received extensive attention in clinical trials due to their considerations of ethics and efficiency. In the framework of RAR procedures, various target allocation proportions have been proposed and studied in literature. In this paper, we develop a family of RAR procedures using power function of corresponding hypothesis testing (RAR-P), and obtain the asymptotic properties under widely satisfied conditions. The proposed procedures are: (i) easy to understand and implement; (ii) applicable in more situations (continuous and discrete responses); (iii) more ethical than classical RAR procedures under some situations; and (iv) capable of monitoring efficiency and re-estimating sample size in clinical trials. Finally, we investigate performance of the proposed procedures through simulation studies.

Keywords

power function, doubly adaptive biased coin design, allocation proportion function, efficiency and ethics

2010 Mathematics Subject Classification

Primary 62L05, 62L20. Secondary 62K99.

Full Text (PDF format)

The authors’ research work was supported in part by the National Natural Science Foundation of China (Nos. 11701152 and 11731012), the Zhejiang Provincial Natural Science Foundation (No. LY17A010016), the Doctoral Research Fund of Henan Polytechnic University (No. 672103/001/147), and the China Postdoctoral Science Foundation (No. 2017M620240).

Received 6 April 2017