论文部分内容阅读
计数资料在病虫测报工作中占有很大比重。目前多用列联表 x~2检验法进行统计检验。但 x~2检验不适用于分析有序性列联表(等级分组资料),如害虫发育进度分一龄、二龄,…;病害发主程度分一级、二级,…等。这些资料列联表多属有序性者。此外 x~2检验对多维列联表分析有困难。再加上 x~2检验必须满足它的使用条件,如当实测频数小于5时不能使用 x~2变量。因此 x~2检验远不能满足我们病虫测报日常工作的需要。在此笔者提出一
Counting data occupy a large proportion in pest forecasting work. At present, the use of contingency table x ~ 2 test for statistical tests. However, the x ~ 2 test is not suitable for the analysis of ordered contingency tables (hierarchical grouping data). For example, the progress of pests is divided into one age, second age, ...; Collective tables of these materials are mostly orderly. In addition x ~ 2 test multidimensional contingency table analysis difficult. Plus x ~ 2 test must meet its conditions of use, such as when the measured frequency is less than 5 can not use x ~ 2 variables. Therefore, x ~ 2 test far can not meet the needs of our pest forecast daily work. In this author put forward one