论文部分内容阅读
以日常护理需求为实例 ,探讨了应用事件数回归模型所应当注意的事项 ,着重考察 :(1)当事件数变量存在大量 0值时 ,如何在建模时加以处理 ;(2 )如何处理零堆积泊松回归模型中的右屏蔽效应 ;(3)并比较不同统计模型对相同变量的边际效应的影响。最后简单地讨论了零堆积负二项回归模型对我们理解老年人日常护理需求模型的涵义。
Taking the daily nursing needs as an example, we discussed the matters that should be paid attention to when applying the event number regression model, focusing on the inspection: (1) When there are a large number of zeros in the event number variable, how to deal with it in the modeling; (2) How to deal with zero Accumulate the right masking effect in the Poisson regression model; (3) Compare the effect of different statistical models on the marginal effect of the same variable. Finally, we briefly discuss the implications of the zero-stack negative binomial regression model for our understanding of the model of daily care needs of the elderly.