在母體為常態分配下，假設我們是以母體平均數大小為衡量母體好壞的準則，而母體變異數則有可能是已知或是未知的狀況。在母體變異數已知或未知但知比例關係的情況，已有不同研究提出方法來解決，因此本文章研究的是在母體變異數未知且可能不等之情況下，利用兩階段抽樣方法及子集合擇優法，完成從k個獨立常態母體中選取出最佳的t個之擇優選取程序。

This paper gives a procedure for selecting the t best out of k normal populations with unknown and unequal variances. The method is based on a two-sample procedure which proposed by Stein (1945) and subset selection proposed by Gupta (1965). And, comparisons of simulations for setting different k, t, no are made.

Bechhofer, R. E. (1954). A single-sample multiple decision procedure for ranking means of normal populations with known variances. The Annals of Mathematical Statistics, 25, 16-39.
Bechhofer, R.E., Dunnett, C. W. & Sobel, M. (1954). A two-sample multiple decision procedure for ranking means of normal populations with a common unknown variance. Biometrika, 7, 170-176.
Dudewicz, E. J. & Dalal, S. R. (1975). Allocation of observations in ranking and selection with unequal variances. The Annals of Mathematical Statistic, 37, 28-78. Gupta, S. S. (1965). On some multiple decision (Selection and Ranking) rules. Technometrics, 7, 225-245.
Ofosu, J. B. (1973). A two-sample procedure for selecting the population with the largest mean from several normal populations with unknown variances. Biometrika, 8, 117-124.
Stein, C. M. (1945). A two-sample test for a linear hypothesis whose power is independent of the variance. The Annals of Mathematical Statistics, 16, 243-258.
Wen, M. J. (1991). Single-stage multiple comparison procedure under heteroscedasticity. Ph. D. dissertation, department of statistics, university of Georgia, Athens, GA. Chapter 4, 33-39.