A Class of Regression-type Estimators for Population Mean Utilizing Transformed Variable with Missing Data with an Application to COVID-19

The COVID-19 pandemic has negatively afflicted people’s lives nowadays and impacted human health and economies around the world. Knowledge of COVID-19 data in advance benefits policies and planning for all countries around the world. However, some COVID-19 data may be missing. A class of regression-type estimators utilizing variable transformation has been proposed when the study variable is missing under simple random sampling without replacement. Both the bias and mean square error of the proposed estimators are studied. The performances of the proposed estimators are investigated via simulation studies and an application to COVID-19 data. The results from the application to COVID-19 data show that the proposed estimators yield the highest efficiency. Compared to the mean imputation method it has at least 6.59 times higher efficiency.