The Research On The Approach Of Calibration Estimator

The primary research on this paper is that on the basis of R language and environment we conclude the approach of calibration estimator in sampling survey from two aspects: the lengthways and the transverse. Firstly, in chapter 1, summarizing the background of choosing this topic, the research meaning, the study of the theory, the development of the software and the practice at present. Then giving some basic concepts. Secondly, introducing the development of calibration estimator from the lengthways and the transverse aspects in chapter 2 and chapter 3. In chapter 3, giving three classical estimators: poststratification estimator, raking, generalized regression estimator. Then presenting calibration estimator and model-calibration approach. In chapter 4, introducing the application of calibration estimator in two-phase sampling, non-response and sampling rotation. The characteristic of this paper is that on the basis of theory research, using R language and environment to link the approach of calibration estimation and the application of actual data. At last, sum up in chapter 4, and point out some shortcomings and the prospects in the future.