统计研究 ›› 2009, Vol. 26 ›› Issue (3): 91-96.

• 论文 • 上一篇    下一篇

基于经典计量模型的统计数据质量评估方法

刘洪 黄燕
  

  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2009-03-15 发布日期:2009-03-15

Based on the Classical Econometric Model of the Quality of Statistical Data Evaluation Methods

Liu Hong Huang Yan   

  • Received:1900-01-01 Revised:1900-01-01 Online:2009-03-15 Published:2009-03-15

摘要: 本文以经济理论为基础,从整个经济系统出发,利用研究对象的相关影响因素构造计量模型,在既定模型下,运用异常值的检验方法及统计诊断原理进行数据质量的定量评估。通过选择合适的模型对考察对象的变化规律进行模拟,找出异常数据(离群值),判断异常数据是否显著异常,对异常数据进行多方查证和原因分析来进一步判断数据的质量,并对我国的统计数据质量进行了实证分析。

关键词: 统计数据质量, 经典计量模型法, 异常数据, 实证分析

Abstract: Based on economic theory and the entire economic system, this paper uses related factors of the research subjects to build econometric model. And depend on the established model, this paper uses the test of abnormal values and the principles of statistical diagnose to evaluate the quality of data quantitatively. By choosing the appropriate model, the paper simulates the changes of inspected objects, identifies exceptional data (outliers), judges its significance and investigates the causes. Finally, the paper empirically analyses the quality of our country’s statistical data.

 

Key words: the quality of statistical data, classical econometric model, exceptional data, empirical analysis