用累积比数模型分析有序数据是有效的,但也存在两方面的问题:其一,数据模型在“有序”的表现上是模糊的;其二,无法对各相邻的类之间作显著性检验。但对这两方面的讨论有时是需要的,比较常见的问题是想要了解在解释变量 X 的作用下,反应变量 Y 在不同相邻类中的变化情况,即所谓的“关节点”问题,如冠心病研究,危险因素既可能影响到发病与不 Using the cumulative ratio model to analyze ordered data is effective, but there are also two problems: First, the data model is fuzzy in the performance of the “ordered”; Second, it can not be the adjacent class Significant test between. However, it is sometimes necessary to discuss these two aspects. The more common problem is to understand how the response variable Y changes in different adjacent classes under the influence of the explanatory variable X, which is the so-called “joint point”. Problems, such as coronary heart disease research, risk factors may affect both the onset and the