Truncation of the Bechhofer-Kiefer-Sobel sequential procedure for selecting the normal population which has the largest mean

Abstract

We study the effects of truncation on the performance of an open vector-at-a-time sequential sampling procedure C) proposed by Bechhofer, Kiefer and Sobel (1968), for selecting thenormal population which has the largest mean, when the variances of the populations are known andequal. The performance of a truncated version (#R ), of 9L is compared to that of 9L. The performance characteristics studied include the achieved probability of a correct selection (P{CS}), theexpected number of vector-observations (n) to terminate sampling, and the variance of n; all of thesequantities are estimated using Monte Carlo sampling. Both 9L and 9L guarantee the 15j.15specified P{CS}. It is shown that 9L is far superior to 9^ in terms of E{n} and Var{n}, particularly when the population means are equal. The implications of this finding in terms of 1067