Likelihood Inference in the Random Effects Logistic Regression Model with Response Misclassification and Covariate Subject to Measurement Error
Generalized linear mixed models (GLMMs) are common methods for the analysis of clustered data. In many longitudinal and hierarchical epidemiological frameworks, accurate measurements of variables are invalid or expensive to be obtained and there might be situations that both the response and covariate variables are likely to be mismeasured. Insensitivity of errors in either covariate or response variable is, not always plausible. With nonlinear regression models for the outcome process, classification errors for binary responses and measurement error in covariates basically needs to be accounted for in order to make conclusive inferences. In this article, we provide an approach to simultaneously adjust for non-differential misclassification in the correlated binary response and classical measurement error in the covariates, using the multivariate Gauss-Hermite quadrature technique for the approximation of the likelihood function. Simulation studies are then conducted to inform the effects of correcting for measurement error and misclassification on the estimation of regression parameters. The application of the multivariate Gauss-Hermite quadrature method in the conjunction of measurement error and misclassification problems is further highlighted with real-world data based on a multilevel study of contraceptive methods used by women in Bangladesh.
- حق عضویت دریافتی صرف حمایت از نشریات عضو و نگهداری، تکمیل و توسعه مگیران میشود.
- پرداخت حق اشتراک و دانلود مقالات اجازه بازنشر آن در سایر رسانههای چاپی و دیجیتال را به کاربر نمیدهد.