Download PDF

3. Data Life Cycle

3.4 Quality

Whether it is a scientific advice, a discovery, an analysis or a report, its credibility is directly linked to the quality of the data used to produce it. Quality management is therefore a key component of the science data management process.

3.4.1 QA/QC


Quality assurance refers to actions beyond tests and controls, and aims at identifying and preventing non conformity problems. QA also includes training and audits.

Real-time data acquisition involves specific procedures due to the fact that data is being used as soon as it is produced. This presents a level of risk for users who could misinterpret data or could not recognize inaccuracies. As an example, the Integrated Ocean Observing System – IOOS has developed a series of automated procedures for real-time ocean data control (QARTOD: Quality Assurance for Real-Time Oceanographic Data).1


Quality control generally refers to activities aiming at products & services quality verification by looking at problems and abnormal elements. In the area of data management, data is examined in order to identify errors, missing data or acquisition problems (see: section 3.4.4 Quality Flagging).


Quality management represents a higher level of management e.g. organizational or institutional. ISO standard 9001 describes a quality management model including procedures, management of responsibilities, resources and services, as well as management of quality assessment, analysis and improvement.

Statistics Canada − an organization based on data – defines its quality management framework and the quality of its information products in terms of their fitness for use by clients. It is a multidimensional concept including the relevance of the information to users' needs, as well as its accuracy, timeliness, accessibility, interpretability and coherence.2