统计研究 ›› 2009, Vol. 26 ›› Issue (1): 32-39.

• 论文 • 上一篇    下一篇

统计数据准确性评估:方法分类及适用性分析

王华 金勇进   

  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2009-01-15 发布日期:2009-01-15

Statistical Data Accuracy Assessment: Methods Classification and Applicability

Wang Hua&Jing Yongjin   

  • Received:1900-01-01 Revised:1900-01-01 Online:2009-01-15 Published:2009-01-15

摘要: 对于统计数据准确性评估方法的归纳分类,可以从辅助资料信息的来源及表现形态、包含目标特征真值关键信息的参照标准的构造方法、实际统计数值与参照标准之间的比较逻辑等方面加以考察。针对总量统计数据的准确性评估,主要在纵向时间维度上开展,根据统计指标与其相关联指标变动趋势的偏离程度加以评判;针对个体或分类统计数据的准确性评估则主要在横向空间维度上开展,对数据的统计分布形态加以检验,或者利用重复调查或随机实验方法对事先假定的误差参数进行估计。不同方法具有不同的适用条件,在实际应用中,应基于所掌握辅助资料的详实程度来选择评估方法,并对导致评估方法发生误判的因素加以细致分析和排除,以确保评估结论的说服力和可信度。

关键词: 统计数据准确性, 评估方法, 适用性

Abstract: For generalizing and classifying the assessment methods of statistical data accuracy, some factors should be reviewed such as source and form of the assistant information, conformation of contrast criterion containing key information of objective characters’ real values, and logistic for comparing factual statistical data with contrast criterion. The assessment of gross statistical data accuracy has mostly been processed along the longitudinal way, the conclusion of which can be obtained according with actual statistics’ trend departure from correlate indices. Meanwhile, the assessment of unit or section statistical data accuracy has mostly been processed along the landscape orientation, for which probability distribution of data would be tested, or error parameters be estimated based on re-interview or random experiment. In addition, there’ve been different applicability conditions for different assessment methods. In order to insure reliability of assessment conclusions, abundance degree of assistant information should be considered, and those factors inducing miscarriage of justice be checked and removed.

 

Key words: statistical data accuracy, assessment methods, applicability