A0937
Title: An improved MC-SIMEX method
Authors: Lili Yu - Georgia Southern University (United States) [presenting]
Varadan Sevilimedu - Memorial Sloan Kettering Cancer Center (United States)
Abstract: The problem of misclassification in covariates is ubiquitous in medical data and often leads to biased estimates. The misclassification simulation extrapolation method (MC-SIMEX) is a popular approach to correct this bias. However, it utilizes an approximated extrapolation function derived from the simulated data, which not only reduces the reliability of the estimator but also increases computation time. The aim is to propose an improved MC-SIMEX estimator for generalized linear models, in which the closed-form exact extrapolation function is derived, thus addressing the challenges in the original MC-SIMEX method. Simulations demonstrate that the newly proposed method outperforms the original MC-SIMEX approach in terms of bias correction. Additionally, a real data example is provided to illustrate the effectiveness of the proposed method.
