网络调查中的非抽样误差——以开放式Web调查为例  

Non-Sampling Errors in Online Survey——Taking Open Web Survey as An Example

作　　者：张华 ZHANG Hua(School of Statistics,Shanxi University of Finance and Economics,Taiyuan 030006,China)

机构地区：[1]山西财经大学统计学院,山西太原030006

出　　处：《统计学报》2022年第5期70-82,共13页Journal of Statistics

摘　　要：以开放式Web调查为例分析了网络调查非抽样误差的来源与分类,根据开放式Web调查的样本选择机制说明了非抽样误差的分类,在此基础上给出各类非抽样误差的控制措施与开放式Web调查中的参数估计。研究表明,开放式Web调查的非抽样误差可分为覆盖误差、调查者误差和被调查者误差,相应控制措施为明确目标总体,问卷设计兼顾规范化、技巧化与方法化,善于运用随机化的问答技术和数据质量识别机制。This paper took open Web survey as an example to investigate the origins and categorization of non-sampling errors in online surveys.The categorization of non-sampling errors was presented according to the sample selection mechanism of open Web survey.On this basis,the control measures of various types of non-sampling errors and the parameter estimation in open Web survey were proposed.The results showed that,the non-sampling errors of open Web survey included coverage errors,investigator errors,and respondent errors.And relevant control measures covered clarifying the overall goal,optimizing questionnaire design,reasonably applying randomized question-answering technology,and constructing data quality recognition system.

关 键 词：开放式Web调查 样本选择机制 非抽样误差 误差控制 参数估计 

分 类 号：F222[经济管理—国民经济]