Comparison of Several Imputation Techniques for Log Logistic Model with Covariate and Interval Censored Data

Abstract

The main purpose of this study is to compare the performance of midpoint, right, and left imputation techniques for log logistic model with covariate and censored data. The maximum likelihood estimation method (MLE) is used to check the efficiency of imputation techniques by estimating the parameters. The performance of the estimates is evaluated based on their bias, standard error (SE), and root mean square error (RMSE) at different sample sizes, censoring proportions, and interval widths via a simulation study. Based on the results of the simulation study, the right imputation had the best overall performance. Finally, the proposed model is fitted to the real breast cancer data. The findings suggest that the log logistic model fits the breast cancer data well and the covariate of treatment significantly affects the time to cosmetic deterioration of the breast cancer patients.