The Research Of Reliability Assessment Methods Based On GDP Statistics With Logical Relationship

Statistical data, as the "barometer" of measuring the status of social andeconomic development, is an important basis for the government to proceedmacroscopic guide. The reliability of statistical data is an indispensable factor of thequality of statistical data, also is the scientific and effective criteria of the nationalpolicy-making. The reliability of statistical data comprehensively embody theobjectivity of statistical data to describe socio-economic development situation,effectively reflect the degree of deviation of statistical data to actual development ofthe national economy.GDP, as the main statistical indicators to effectively measure a country’seconomic development, has been much attracting an intense attention in the social andeconomic development. The reliability of GDP statistics has also been caughtworldwide attention at home and abroad.There are interdependent and interrelated relationship between a variety ofindicators reflecting the socio-economic phenomenon, under certain technicalconditions and the scale of production, the relationship between the various indicatorsare relatively stable. The assessment method of logical relationship is on the basis ofaccounting rules of statistics in science, with same measure of statistical indicators asthe reference standard, to assess the phenomenon against the specific logical rules orrelationship between the indicators.The reliability evaluation method of GDP statistics based on the logicalrelationship uses the interdependent and interrelated logical relationship betweenstatistical indicators of GDP and their relevant indicators, through the use of scientificdata assessment methods to effectively measure if the statistical indicators’ value of GDP is contrary to the objective laws of society, and then identify the reliability ofdata or whether there is a distortion.Based on the research on the reliability assessment method of GDP statistics onthe basis of logical relationship, through to filter statistical indicators on therelationship of GDP, this paper selected relevant indicators with reliable data sourcesto build the reliability evaluation index system of GDP statistics, and used correlationcoefficient analysis to excluded strong correlation indicators, determined the finalindex system. As to non-stationary long-run equilibrium relationship between the timeseries of statistical indicators data, this paper constructed a vector error correctionmodel to assess the reliability of country’s GDP statistics, and verified the validity offeasibility and assessment result of model evaluation method with proportionalrelationship analysis. Further, this paper used price index comparative analysis andproduction function analysis based on the logical relationship to assess the reliabilityof statistical data of regional GDP of Sichuan Province.The results showed that the country’s GDP statistics are more reliable, and onlyin a particular year there was a certain degree of inflated or underestimated from1993to2010, the average relative error of the national GDP growth rate was only2.42%.Sichuan Province’s GDP statistics in the same was relatively reliable, there was acertain degree of inflated since1997, but inflated amplitude is limited.Based on the reliability evaluation method of GDP statistics of logicalrelationship, this paper assessed and researched the reliability of our country and theregion’s GDP statistics, from the perspective of empirical analysis of the reliability ofGDP statistics, explored a scientific and effective technical determination means.