基于结构匹配性视角的数据质量评估方法  

Data Quality Assessment Method From the Perspective of Structure Matching

作　　者：陶春海[1,2] 钟桂珍 Tao Chunhai;Zhong Guizhen(School of Statistics,Jiangxi University of Finance and Economics,Nanchang 330000,China;Key Laboratory of Financial Data Science,Jiangxi University of Finance and Economics,Nanchang 330000,China)

机构地区：[1]江西财经大学统计学院,南昌330000 [2]江西财经大学财经数据科学重点实验室,南昌330000

出　　处：《统计与决策》2023年第21期17-22,共6页Statistics & Decision

基　　金：国家社会科学基金资助项目(21&ZD150,20ATJ003);江西省研究生创新专项资金项目(YC2022-B148)。

摘　　要：数据是国家各级部门制定相关政策的重要依据,开展数据质量评估方法研究有利于提高数据质量评估的精度,从中把握事物的内在规律,进而提高决策效率。鉴于此,文章根据贫困县调研数据中家庭成员年龄结构和劳动能力结构与收入之间的匹配性关系,提出基于结构匹配性视角的数据质量评估方法,以2020年S省Z贫困县的实地调研数据为样本,验证该方法的有效性。研究结果表明:提出的结构匹配性视角下的数据质量评估方法确实有效,能够从不同方面、角度对数据质量进行评估,评估结果显示,样本数据异常值少,模型的经济学意义、统计学意义一致,抽样分布与原始数据分布一致,贫困县调研数据质量较高。该方法对于建立健全数据质量评估体系、拓展结构匹配性视角数据质量评估方法的应用范围、加强对数据质量评估方法的研究具有重要的理论和实践意义。Data is an important basis for national departments at all levels to formulate relevant policies.Carrying out re-search on data quality assessment methods is conducive to improving the accuracy of data quality assessment,grasping the inher-ent laws of things,and improving decision-making efficiency.In view of this,this paper proposes a data quality assessment meth-od based on the perspective of structural matching according to the matching relationship between the age of family members and the structure of labor capacity and income in the survey data of poverty-stricken counties.Field survey data of Z poverty-stricken county in S Province in 2020 are taken as samples to verify the effectiveness of the method.The research results show that the data quality assessment method from the perspective of structural matching proposed in the paper is indeed effective,and can evaluate the data quality from different aspects and angles.The evaluation results are shown as below:The sample data has few outliers,and the model has economic and statistical significance,the sampling distribution is consistent with the original data distribution,and the quality of the survey data in poor counties is good.This method has important theoretical and practical significance for es-tablishing and improving the data quality assessment system,expanding the application scope of the data quality assessment meth-od from the perspective of structural matching,and strengthening the research on the data quality assessment method.

关 键 词：BENFORD法则 异常值检测 模型一致性 Bootstrap抽样 

分 类 号：F224[经济管理—国民经济]