农业技术

抽象的

Calibration Estimator of Regression Coefficient Using Multi-Auxiliary Variables

Vandita Kumari, Kaustav Aditya

Regression coefficients computed using ordinary least square technique assume that the observations are independent and identically distributed. These assumptions are questionable for the data that are collected using complex survey design. The sampling design information must be incorporated in estimating the regression coefficients from survey data using the sampling weights.An efficient estimator of regression coefficient has been developed by extending the calibration method with multiauxiliary variables that are related to the study variable.The estimators of variance of the proposed calibration estimator have also developed using Taylor series linearization technique and the bootstrap method. The results based on empirical studies using both simulated as well as real datasets show that the proposed calibration estimator performs better than the existing
estimator. In addition, both proposed methods of variance estimation for the calibration estimator perform adequately.