Experimental Research On The Preventive Measures For Cell Non-response

Non-response error is of non-sample error with high occurrence frequency and high control difficulty. It has a great harm for estimator and by thus influence the quality of investigation data. The major measurement taken for controlling non-response error is to reduce the non-response rate. In this paper, the feasible measurements for reducing the cell non-response rate from the perspective of prevention are discussed. The Chengguan District of Lanzhou is selected as a sample district and the experimental survey is applied to carry the confirmation study on the measurements of selecting the proper survey time, survey method, surveyor and the survey times.In Chapter 1, there is an overview of document on the non-response control measurement. It is found that academia focuses on the study of the pose remedial measure and the study of preventive measures for project non-response; while there are few studies on the preventive measures for cell non-response concerning the embody, normalization and practical operation feasibility, especially in china, there are few demonstrations combing the test in method study. Based on these above, the cell non-response preventive experimental study shall be raised.Chapter 2 focuses on the experiment design. Based on the experiment purpose, the survey time, survey method and surveyor shall be regarded as three factors influencing the cell response rate. The multi-factor experiment shall be selected and the experiment shall be carried repeatedly. The whole sample district shall be divided to 19 blocks; a block shall be a combination of different factors on different levels. The PPS three-stage stratified random sampling method shall be applied to select sample form every block.Chapter 3 is the key part of this paper. The survey result shall be analyzed. Firstly the representation of the experiment sample is checked, and the Logistic regression mould shall be applied to find out that the survey time, survey method and surveyor have a significant correlation with the cell response situation; while the survey times has not a notable influence on the cell response situation. Then, the survey questionnaire shall be analyzed. It is found that for different survey objects, the distribution of leisure time in a day, the more welcomed survey method and the more welcomed surveyors are featured with different central tendency. Based on these above, it is raised that in survey for survey objects with different features, the survey time on different slots of a day, different survey methods and different surveyors (man or woman) shall be designed. By thus, the cell unresponseed rate can be reduced. Therefore, in survey design to select proper survey time, survey method and surveyor is the feasible measure to prevent the no responses for cell.Lastly the distribution feature and rejection reasons of survey rejecters are analyzed. The last chapter , the meaning of our organized excremental study is evaluated , and the conclusions and the reveals from experiment and the survey process shall be summarized.