统计数据质量评估方法研究述评  

Reviews on Assessment Methods Quality of Statistics

作　　者：许涤龙[1,2] 叶少波[2] 

机构地区：[1]湖南大学社会经济统计研究中心,湖南长沙410079 [2]湖南大学金融与统计学院,湖南长沙410079

出　　处：《统计与信息论坛》2011年第7期3-14,共12页Journal of Statistics and Information

基　　金：国家社会科学基金重点项目<基于国际可比性的政府统计数据质量研究>(08ATJ001);国家统计局统计科学研究所研究基地项目<政府统计数据质量评估方法及其应用研究>(2009001)

摘　　要：数据质量评估是统计数据质量管理的重要环节。统计数据质量的评估方法有逻辑关系检验法、计量模型分析法、核算数据重估法、统计分布检验法、调查误差评估法以及多维评估法六个类别,在详细讨论其各自的评估过程、适用性及其优缺点的基础上,按照评估维度和评估方式对各种统计数据质量的评估方法进行再归类,指出评估方法的进一步研究应该围绕计量模型分析法、核算数据重估法中的物量指数重估法、调查误差评估法和多维评估法等方向展开。Data quality assessment is an important link of the quality of statistical data management. The paper classifies the assessment methods of statistics quality into the following six categories, namely, logical relationship investigation methods, econometrical model analysis methods, accounting data revaluation methods~ statistical distribution investigation methods, survey errors assessment methods and multi- dimension assessment methods of data quality. After that, it respectively discussed their assessment procedures, applicability and its merits and shortcomings in detail. Based on this, this paper re --categorized the above mentioned assessment methods according to the assessment dimension and the assessment approach, and as well pointed out that the further study on assessment methods of statistics quality should take the following four aspects as centers, that is, econometrical model analvsis Asnort.revaluation according to the volume index aspect in accounting data revaluation methods, survey errors assessment aspect and multi--dimension assessment of data quality aspect

关 键 词：数据质量 评估方法 一维评估法 多维评估法 发展方向 

分 类 号：C811[社会学—统计学]