配对t检验与完全随机设计t检验统计效率的判别分析

The Discriminatory Analysis of the Statistical Efficiency Between the Paired t-test and the Completely Random Design t-test

目的 :配对设计可以提高实验效率 ,本文对定量资料配对设计与完全随机设计 (成组设计 )的统计效率进行了对比分析 ,根据t界值表计算出两种设计统计效率的判别标准。方法 :理论推导与模拟数据计算相结合。结果 :(S21 +S22 ) S2d 的比值对统计结果有影响。结论 :(S21 +S22 ) S2d 时 ,配对设计的统计效率高 ,(S21 +S22 ) S2d t2α ,2v时 。

Objective The paired design may improve experiment efficiency. This paper aims to conduct a comparative analysis of the statistical efficiency between the paired design and the completely random design of the quantitative data. In accordance with t-critical value to compute the discriminatory standard of the two designs. Methods The combination of theoretical deduction and simulated data computation. Results The ratio of (S 2 1+S 2 2)/S 2 d has had some influence upon the statistical results. Concl...