随机效应模型广泛应用于刻画重复测量数据的特征,Banerjee和Frees^[1]用Cook距离,Lesaffre和Verbeke^[2]用影响曲率分别对线性随机效应模型进行了分析。本文利用影响曲率对具有AR(1)误差的非线性随机效应模型中的自相关系数扰动进行了分析,得到了影响曲率的表达式,并且利用血浆药物渗透数据(Davidian和Gillinan^[3])来说明分析方法的应用。
Random effects models are widely used to model the characteristics of repeated measurement data. Banerjee & Frees^[1] and Lesaffre & Verbeke^[2] respectively used Cook displacemnet and influence curvature to analyse linear random effects models. This paper uses influence curvature to study perturbation diagnostics of autocorrelation coefficients in nonlinear models with AR(1) errors. The plasma infiltration data (Daridian and Gillinan^[3]) is used to illustrate our results.