How To Achieve Perfect Data Quality Control – 3 Tips To Improve Data Quality


The wave of economic globalization has increased the development of the national economy. In today’s big data era, data quality is the top problem we face. In the face of data quality control, what aspects should we look for problems and strive for breakthroughs? With the reform of China’s economic system and the new situation of economic globalization, the quality control of data has long become the core content of statistical work, and the quality of statistical data is also the most fair evaluation of the whole statistical work. From all aspects, there are still some problems in the quality control of statistical data in China, which need to be solved by effective control measures.

What Are The Problems In The Quality Control Of Statistical Data In China?

1. Absence of quality control specifications
Due to the absence of quality control norms, it is difficult for some statistical institutions to achieve effective communication with users and society. Many data users and all sectors of society have no relevant understanding of the complexity of statistical work and inevitable statistical errors, and lack understanding of the control methods, degree and results of statistical data quality. Therefore, even though China’s statistics has always attached great importance to the quality of statistical data and done a lot of work for this purpose, the users of statistical data and all sectors of society are still very dissatisfied with the quality of statistical data.


2. One sided statistical data quality control
The experience and measures that have achieved great success in the whole process control of statistical data are often concentrated in several project areas with narrow application scope. The quality of most statistical data is lack of whole process control, and there are omissions in many links. The quality control requirements for the investigation link are high, but less consideration is given to the data demand in the design link, and the relevance of the data can not be controlled. In terms of personnel, the internal personnel control of the statistical system is relatively in place, but the grass-roots personnel outside the statistical system who account for the statistical staff are lack of control.


3. The quality control system of statistical data is not perfect
The quality control technology of statistical data is not widely used in statistical work and research practice, and the regular data quality control also tends to post analysis and evaluation in actual operation. The prior work is not in place, and the application of error model is also very few. In the quality evaluation of statistical data, the application of the evaluation results or conclusions to the design and improvement of the next survey plan is not fully reflected, and the afterwards quality control technology and data quality control organization activities can not achieve a good combination. Although the data quality control evaluation method has been formulated, it lacks specific support measures, such as unclear division of department responsibilities, unclear authorities and obligations, and non-standard application of data quality control evaluation method.


How To Achieve Perfect Statistical Data Quality Control?

In view of the above problems, we put forward the following improvement measures, hoping that the existing problems can be improved accordingly, so that the statistical data can play its important value in the fierce market competition.


1. Formulate clear quality control specifications
We should speed up the establishment of independent social intermediaries engaged in the quality evaluation of statistical data and ensure that the quality evaluation of statistical data is impartial and independent. Formulate complete statistical data quality assessment and evaluation standards, and clarify the error range of various statistical data, so that there are clear reference standards for the assessment and evaluation of statistical data quality. Make detailed provisions on the scope and focus of statistical data quality evaluation indicators, formulate relevant evaluation and supervision methods, and communicate with the public and data users.


2. Establish a scientific statistical system
We should further improve and improve various census systems, establish and improve relevant laws and regulations, and make reasonable arrangements for the items, sequence, time and cycle of various censuses, so as to ensure a good connection between various censuses and between censuses and annual reports. Merge the current census according to the nature, so as to facilitate the classified management according to the characteristics of all kinds of census. We should adapt to the diversified needs of the market, combine various forms of investigation means, learn from each other’s strengths and complement each other’s weaknesses, and verify each other, so as to ensure the accuracy of statistical data.


3. Apply SPC statistical data quality control system
The quality of statistical data is guaranteed through the process of data collection, analysis, processing and evaluation. Therefore, the method of post inspection of statistical data should be changed, the quality control of data collection, analysis, processing and evaluation should be strengthened, and the whole process control should be implemented. SPC statistical process control system can quickly collect and analyze process data to help enterprises achieve the purpose of process data quality control. Applying infinite SPC statistical process control system for quality control can not only improve the quality of statistical data, reduce the number of unqualified products, meet customer needs, but also reduce statistical costs and save resources for enterprises.