Analysis Of Marginalized Transition Model For Binary Longitudinal Data With Dropout

There always exsists a lot of missing data in the statistical survey,many reasons account for the dropout. Different kinds of missing data will have different effects on the statistical analysis. To improve the quality of the survey data, on the one hand, effective measures should be taken to reduce the missing data and improve response rate in survey. On the other hand, when there is missing data, we should take some action to deal with incomplete data set and reduce the impact that due to missing data. This paper uses the marginalized transiton model to analyze the longitudinal binary data, fully considered the relationship between response variables and the covariates, and relationship among response variables with the same individual, and give one general analysis method.