|
Statistical data are not only the core products of statistics, but also the "barometer" of the macroeconomic performance, and an important basis for scientific decision-making of the government, enterprises and the public. Quality of statistical data has been widespread concern, which is one of the focus issues of current Chinese government and the public.This article based on economic theory and statistical analysis methods, using statistical diagnostic principles from the correlation and matching degree between China’s major economic indicators, select7indicators that is different data acquisition channels from GDP to analyze the statistical data quality in-depth,7indicators include the fixed asset investment, electricity consumption, number of employments, freight turnover and other factor inputs indicators to constitute GDP, as well as the impact indicators of socio-economic change in GDP, such as tax revenue, total final consumption expenditures, unemployment rate and other factors. First, using the Benford’s Law to conduct a preliminary examination on the accuracy of the monitoring indicators; Then, from the cross-sectional data, using the idea of relative measure of the quality of data, constructed the dependence indicators reflected the quality of2011GDP data; Finally, with the correlation coefficient weighting method and the variation coefficient weighting method to measure the quality of GDP data of Chinese provinces (cities, districts) respectively, and to rank the quality credibility of data according to the comprehensive dependence results.The empirical analysis shows that, two different weighting methods calculate a similarity rank result of comprehensive dependence, and the eastern region get the highest data quality credibility, Among them, Jiangsu, Guangdong and Zhejiang’s data quality credibility is the best, Tianjin, Shanxi and Guangxi is the worst. The quality measure method of data dependence is both the core and an important innovation of this paper.
|