论文部分内容阅读
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 ,完全随机设计的统计效率高
OBJECTIVE: The pairing design can improve the experimental efficiency. In this paper, the statistical efficiency of the quantitative data pairing design and the completely random design (group design) are compared and analyzed. The discriminant criteria of the two design statistical efficiencies are calculated based on the t-cut table. Methods: The combination of theoretical derivation and simulation data calculation. Results: The ratio of (S21 + S22) S2d had statistical significance. Conclusion: (S21 + S22) S2d, the statistical efficiency of the paired design is high, (S21 + S22) S2d t2α, 2v, the statistical efficiency of the completely randomized design is high