Case deletion model and related diagnostic for linear mixed measurement error models

Case deletion model and related diagnostic for linear mixed measurement error models

In this paper, we present case deletion model for linear mixed measurement error models using the corrected likelihood of Nakamura (1990). Furthermore, generalized Cooks distance is constructed as a tool for influence diagnostics. It is found that it can be written in terms of studentized residuals of model, error contrast matrix and the inverse of the response varicovariance matrix. Our influence diagnostic is illustrated through a real (i ata. Sët.