Calibrating on principal components in the presence of multiple auxiliary variables for nonresponse adjustment

Nonresponse is a major impediment to valid inference in sample surveys. In the nonresponse scenario, driver of successful estimation efficient use available auxiliary information. As electronic devices provide considerable data storage capacities, at stage it natural for survey statisticians face large datasets variables. It unwise all as doing so may lead poor estimators, especially if some variables are strongly correlated. Furthermore, selecting subset not be best alternative given issues related selection criteria. this paper, we propose reducing dimensions original set by using principal components. The components place evaluated via two calibration approaches, linear no explicit response model and propensity known model. For latter, based on their canonical correlation with results simulation studies suggest that appropriate, offers great advantage computational burden. Keywords: Calibration, Nonresponse, Principal components, Weighting