What Are the Different Kinds of Quality Control Techniques?

Data quality control technology refers to adopting certain technical measures to make the data meet the relevant quality requirements in the acquisition, storage, and transmission. ^[1]

Data quality control mainly includes two types: real-time data quality control and delayed data quality control. Real-time data quality control is the basis of delayed data quality control. Good real-time data quality control can alleviate the work of delayed data quality control. The data controlled by the two types of data quality control are different, so the data quality control methods adopted are also different. However, the two types of data quality control processes are basically the same.

Research on data quality control methods is the focus of data quality control. The quality of control methods directly affects the data quality. At present, the methods often used in data quality control include extreme value control, Rheinda test, Dixon test, Grubbs test, and Cochran test.

According to the actual observations of the ocean station observation platform, the data quality control process can be divided into data collection and transmission, data conversion, selection and processing of data quality control methods according to the characteristics of the collected data, and display and storage of data quality control results, etc. .

1. Data acquisition and transmission

Data collection and transmission are the basis of data quality control. Data collection is mainly reading real-time data or time-delayed data from observation instruments. The current data collection is generally based on observation instruments automatically recording observation data, which can eliminate some reading errors caused by humans. During the data transmission process, some transmission errors may be caused, and non-code errors are generally caused. This requires that the data should be non-code checked when receiving data to ensure the unnecessary post-processing of data.

Data transformation

The data transmitted is not necessarily the data form or data dimension as we know it, which requires data conversion. The converted data can more clearly reflect the current state and trends of the marine environment.

3. Select the appropriate data quality control method for data quality control Selecting the data quality control method is the core of the ocean station data quality control process. In order to select a suitable quality control method, the characteristics of the data itself must be analyzed first. Generally, there are data normality judgment, data error normality judgment, the size of the data volume, and even data packet processing. At present, the commonly used data quality control methods include extreme value test, consistency judgment, incremental judgment, 3 method test, Grubbs test, and Dixon test.

4. Display and storage of data quality control results

Data quality control results are generally analyzed by scatter plots, fitting plots, and dot-line plots to analyze the current state and trends of the marine environment. The purpose of data quality control is not only to show the current state of the marine environment, but also to store the results of data quality control in order to understand the laws of the ocean through data accumulation.

The above 4 stages are a brief summary of the real-time data quality control process. Delayed data quality control is a more in-depth data processing process based on real-time data quality control, but the basic data quality control process is only the above 3 , 4 two processes. ^[1]