Information Collection Is No Response Error Of The Theoretical Analysis

In today's information society, the considerable outbreak and rapid spread of information make people have to discern the valuable information in the vast sea of information, and transform the information into kinetic energy for survival and development. With the process of informationization, information collection has become an essential access to reliable information and important work for all walks of life. In information collection, one often encounters a thorny problem which is no response, it directly influences the precision of statistical inference, and it is a key factor in the control of information quality. However, there are few theoretical research and problem-solving strategies in related bibliographies. Therefore, this thesis first bravely proposes the approaches to the theory of probability and statistics applying to the process of the information collected in the non-response, to improve the reliability of the information collection and accuracy.Starting from how to control this non-response error, this thesis elaborates on non-response's concepts, causes and research background at home and abroad. Through the establishment of models, this thesis makes a specific analysis of the statistical influence of non-response, and from two aspects of prevention and remedy puts forward the methods to deal with non-response. On the unit non-response and item non-response, this thesis introduces respectively prevention methods for improving response rate and remedial methods for reducing errors. In terms of the quantitative difference between responders and non-responders, this thesis makes a detailed and systematical research on the weighted adjustment methods and the substitution methods. Meanwhile, based on the routine double sampling theory, the thesis has proposed a new double sampling method, this new method is the combination of regression analysis and double sampling, which can solve the non-response, improve the estimation accuracy and save the information collection costs.