论文部分内容阅读
目的纵向观测数据是按时间顺序对个体的某一变量进行多次观测获得的资料。本文利用广义线性模型对纵向计数数据进行了分析，充分考虑重复观测间的相关性。方法采用Ｚｅｇｅｒ和Ｌｉａｎｇ提出的广义估计方程，在拟合对数广义线性模型的同时，引入偏离参数，讨论三种协方差矩阵的结构。结果同时获得回归参数、相关参数、偏离参数的估计，完成了较为实用的运行程序，并进行了实例讨论。结论医学研究和临床试验中经常接触到纵向观测数据，对这类数据需采用特殊的方法进行分析处理
Longitudinal observational data of the purpose are data obtained by observing a certain variable of an individual in chronological order. This paper uses the generalized linear model to analyze the longitudinal counting data and fully considers the correlation between repeated observations. Methods The generalized estimating equations proposed by Zeger and Liang were used to fit the logarithmic generalized linear model. At the same time, the deviation parameters were introduced to discuss the structure of three kinds of covariance matrices. At the same time, the regression parameters, related parameters and deviation parameters were also obtained. A more practical operating procedure was completed and examples were discussed. Conclusion Long-term observation data are often exposed in medical research and clinical trials, and special methods are needed for analysis and processing of such data.