The Data Analysis Of GEE And QIF Used In The Asthma Data With Missing Data

Longitudinal data is different for each individual observation time obtained bythe cross-section and time series data together. The greatest feature of longitudinaldata is that it cross-section data and time series data together, both of the individualtrends over time, but also of the overall trend. However, longitudinal data for the sameindividual at different times of the repeated observations, but also between individualsbut also there are some differences, which led to the overall integrity of thelongitudinal data is difficult to guarantee. Chen and Little (1999)proposed aWald-type test based on the generalized estimating equation,Calculated and comparedunder different forms of missing data subsets of the parameters of the sample estimateof the difference. Qu and Peter (2002) proposed a more unified generalised-score-typetest, this test method is based on the quadratic inference function is proposed, whichavoids the lack of mechanisms for each parameter estimation. Based on the specificdata is simulated to verify the two methods are asymptotically the same.