考虑纵向数据的线性误差模型,其中协变量含有测量误差.使用2次函数推断方法得到回归参数的估计,证明所得到的估计渐近地服从正态分布;对参数的假设检验问题,证明所得统计量渐近地服从χ^2分布,并通过数值模拟讨论方法的有限样本性质.最后,该方法被用于1组艾滋病数据的实证分析中.
A linear model for longitudinal data with continuous responses and error-prone covariates via quadratic inference functions methods is considered. Asymptotic normality of the parameter estimators is established by quadratic inference functions. In order to testing interested parameter,the statistic that proposed asymptotically follows a chisquared distribution. The finite-sample properties of the procedures are studied through Monte Carlo simulations. At last,an application to a longitudinal study is used to illustrate the procedure developed here.