IMPROVED APPROXIMATION FOR ESTIMATION FOLLOWING CLOSED SEQUENTIAL-TESTS
成果类型：
Article
署名作者：
KIM, KM
刊物名称：
BIOMETRIKA
ISSN/ISSBN：
0006-3444
发表日期：
1988
页码：
121128
关键词：
摘要：
Siegmund (1978) has developed a procedure for estimation following a closed sequential test, known as the repeated significance tests. Here a more elaborate approximation is proposed for a wider class of closed sequential tests. The major goal is to account for the excess over boundary in addition to the stopping time when the sequential test is terminated. Some hypothetical examples are used for comparing the two approximations. To investigate the properties of these procedures further, simulation results are presented in terms of estimated mean squared error, bias and variance for point estimators, and in terms of coverage probability and width for confidence intervals.