Data Quality Assessment Research on Total Retail Sales of Consumer Goods Based on Multi-dimensional Evaluation Methods
|Keywords||total retail sales of consumer goods multi-dimensional evaluationmethod data quality consumer satisfaction investigation structuralequal model|
In order to overcome the negative influences on the export trade of China causedby the global financial crisis of2008and maintain the rapid growing trend, ourcountry had adopted a series of policies to stimulate domestic demand. Total retailsales of consumer goods, the indicator for the final consumption, has gained broaderattention of the society. It is not only the important indicator to reflect the effect ofthe policies to stimulate domestic demand and to develop new policies and plans, butalso the key reference indicator for the enterprises to formulate production plan andpersonal investors’ decisions, which require high quality statistics of total retail salesof consumer goods. Thus, it is of great significance to assess the quality of the totalretail sales of consumer goods statistics based on the multi-dimensional evaluationmethods.Firstly, a research was done on the multi-dimensional evaluation methodsadopted by the three main international organizations, IMF, ESS and OECD. Thispaper mainly explained their data quality assessment frameworks and procedures, andanalyzed their advantages and disadvantages respectively. Then, comparison analysiswas done among these multi-dimensional evaluation methods.Secondly, an assessment framework for total retail sales of consumer goods wasestablished. This paper combined the data quality assessment and managementframeworks of3international organizations, selected7quality dimensions, defined3quality indicators for each quality dimension and established the framework. Then,the customer satisfaction investigation approach and structural equal model werechosen as the methods for data collection and data processing.Thirdly, an empirical study was done on the total retail sales of consumer goods.The structural equation model was used to fit the statistics got from investigation.With a good model fit, the standard loadings of each path were regarded as theweights. The score of each dimension and the total were got by calculating theweighted average of the original data.Finally, combined with the empirical results and the present situation of the totalretail sales of consumer goods, suggestions of3aspects which include setting up thestatistical quality assessment system, increasing the transparency of total retail salesof consumer goods and promoting the statistical reform, were proposed to improve the quality of total retail sales of consumer goods.