本研究主要探討（SRS,SYS）二階段抽樣設計的理論，並以模擬分析
找出最適之樣本配置且與（SRS,SRS）二階段抽樣設計比較其優劣。
其本研究內容摘要如下： 1.求出母體變量總和估計量及變量平均數估
計量之期望值與變異 數。 2.找出母體變量平均數估計量變異數
之不偏估計量。 3.比較（SRS,SYS）二階段抽樣法與（SRS,SRS）二階
段抽樣法之 變異數大小。 4.藉由理論和模擬分析，考慮在成本
限制下和無成本限制下， （SRS,SYS）二階段抽樣法樣本數之最適配
置。 5.藉由模擬分析，比較（SRS,SYS）二階段抽樣法與（SRS,SRS）
二階段抽樣法變異數之差異，並找出使變異數為相對極小時的 樣本
大小。

The objects of this study is to investigate the theory of
(SRS,SYS)two-stage sampling design, to find the optimal
allocation of sampling units, then to compare its with (SRS,SRS)
two-stage sampling design and to determine which one is better.
The summaries of this thesis are as follows: 1. To calculate
the expectation and variance of the estimated population
mean and total. 2. To find the unbiased estimator of variance
of the estimated population mean. 3. To compare variance
of (SRS,SYS)two-stage sampling with that of (SRS,SRS)two-
stage sampling by verification. 4. To consider the optimal
allocation of sampling units under the limitations of cost
and no cost by verification and simulation. 5. By the analysis
of simulation, to compare the differences between variance
of (SRS,SYS)two-stage sampling and that of (SRS,SRS) two-
stage sampling, and to find the number of sampling units which
makes the variance relatively minimum.