Measurement Error in Generalized Linear Models

This research focused on the measurement error in the generalized linear models. We show the asymptotic statistical properties of the proposed estimations for the covariates in the presentence of measurement error. These estimators are based on conditional expectation and take different forms according to the measurement error density function. Our simulation study examines the sampling bias of parameter estimation in generalized linear models, including logistic and Poisson regression models. Our real data study compares estimators from measurement error models to naive estimators where the measurement errors have been ignored. We used data from the Framingham Heart Study and data from the study of diet (fat intake) and coronary heart disease.