| Data is the country’s basic strategic resource,permeates all aspects of life,and has an important impact on production,circulation,and distribution.As a kind of data,statistical data plays a very important role both at the macro level and at the micro level.The quality of statistical data is directly related to the correctness of the country’s macroeconomic policy direction,and affects the accuracy of important corporate decision-making and the authenticity and reliability of academic research results.As an important part of statistical data,poverty alleviation statistics are an important basis for formulating poverty alleviation policies.The quality of their statistical data is directly related to the direction of poverty alleviation policy formulation.In order to better measure the quality of my country’s statistical data and improve the statistical data system,this thesis takes the statistics of poverty alleviation as an example,focusing on the main line of "statistical data quality evaluation".Firstly,explain the theory of total quality management,the connotation of statistical data quality,influencing factors,and the characteristics of the evaluation content;secondly,the premise of analyzing the current status of statistical data quality and existing problems in China,according to the general principles of statistical data quality evaluation,a statistical data quality evaluation system is constructed from two aspects:evaluation index system and evaluation method;again,on the basis of the statistical data quality evaluation index system,data user satisfaction is introduced,and the user satisfaction Multi-dimensional statistical data quality analysis from the perspective of multi-dimensional statistical data,discusses the influencing factors and action paths of user satisfaction with statistical data quality in China,and reveals the key indicators of statistical data quality in China;further based on key indicators,consider the quantification of indicator data,and select accuracy This key indicator uses the statistics of poverty alleviation as an example to carry out data quality research.It mainly takes the statistics of four important poverty alleviation indicators across the country and poverty-stricken areas from 2013 to 2018 as the research object,and uses Benford’s rule to determine the authenticity of the first two digits.Test,and build a principal component panel regression model to test the logical matching between indicators,and further detect the time and area of the problem data based on the model residuals;finally,combining the above empirical analysis results,propose policy recommendations to improve the quality of China’s statistical data.According to the research results of multidimensional statistical data quality from the perspective of user satisfaction,"data openness","data timeliness","data comparability","data authenticity",and "speed of data acquisition" directly affect research user satisfaction,In line with the research hypothesis.Data comparability and data accuracy have a greater impact,while the impact of data openness,data acquisition speed,and data timeliness is relatively weak.Judging from the results of the research on the quality of poverty alleviation statistical data: the c2goodness-of-fit test results of Benford’s law show that the poverty incidence rate,the per capita disposable income of rural residents,and the per capita consumption expenditure of rural residents fail the test.Except for the national rural per capita disposable income,the second digits of other indicators have passed the statistical distribution test at the 0.05 significance level;the poverty incidence data of poverty-stricken areas from 2013 to 2018 has a good relationship with the other three indicators.Logical matching;except for the data of individual regions such as Hebei,Shaanxi,Hainan,and Tibet that failed the residual test,the quality of the data in other regions is relatively good.Based on this,this article believes that the quality of my country’s statistical data is relatively good,but the quality of statistical data should be continuously improved in terms of comparability and accuracy. |