MODEL BASED PREDICTION OF FINITE POPULATION TOTAL UNDER SUPER POPULATION MODEL

Abstract

In the present paper, we have proposed a predictor of finite population total under super population model when the study variate is subjected to measurement error. The results of the simulation study revealed that the percent loss in precision of the predictor varied between 26% and 40% depending upon the sample size and ratio of variance of measurement error to model error variance. The robustness of the predictor has also been examined when the assumed model deviates with measurement error in y.