论文部分内容阅读
探讨现行文献中关于交叉试验资料比较阶段差别间的秩和检验方法的合理性。方法：先将指标的效应进行分解，然后按照秩和检验的方法把阶段间的差别用试验设计的原理进行分析。结果：反映阶段间差别的测试顺序的效应已经被完全抵消。结论：现行的交叉试验的秩和检验无法比较阶段间的差别，应予以改进。
To explore the rationality of the rank sum test method used in the current literature for the comparison of differences in the cross-test data. Methods: First decompose the effect of indicators, and then use the method of rank sum test to analyze the differences between the stages using the principle of experimental design. Results: The effect of the test sequence reflecting the differences between stages has been completely offset. Conclusion: The current rank-sum test of crossover tests cannot compare the differences between stages and should be improved.