Authors Note: This article is intended to provide a wide-scope overview of the concepts of data validation for all who utilize analytical laboratory data. The author has significant experience providing third-party data validation services. The article was intended to be beneficial for those who have utilized data validation as well as for those who are unfamiliar with data validation.

What is Data Validation?

Data Validation is a higher-level scientific-based process that determines the technical usability of the analytical data. In the scientific context, the term "data validation" is commonly used when referring to a process during which a body of data is subjected to a comparison with a set of acceptance criteria. The acceptance criteria are designed in a manner that will enable the data user to know if the set of data will meet the intended purpose.

Data Validation generally can be defined to mean: A systematic process that compares a body of data to the requirements in a set of documented acceptance criteria.

Who Uses the Data Validation Process?

Governmental and regulatory bodies have frequently used a formal process of data validation with environmental protection responsibility. Many governmental projects that involve environmental cleanup, air and water quality monitoring, and risk assessment are required to utilize the data validation process.

Other users of the data validation process are those who rely on analytical data to provide a reliable basis for decision making, monitoring of manufacturing processes, research, and other non-regulatory activities. For them, the driving force is data usability of the data, not environmental regulatory compliance. The usability of the data can be determined by objectively reviewing the analytical data and associated analytical laboratory documentation and comparing it to the data quality objectives and requirements. Thus, by knowing the degree and extent of the usability of the data, the data can serve its intended purpose.

Who Performs Data Validation?

Data validation can be performed by several different entities. Commonly, the analytical laboratory providing the analytical data may do a limited scope data validation. The laboratory's role is generally limited to assigning data qualifiers based on events relating to the analysis of the samples and the requirements in the methods and procedures that were utilized by the laboratory. In addition, the laboratory may follow other applicable contract requirements for data reporting and reporting.

Validation of data by a party independent of the laboratory is most often the more desirable option. A full scope data validation process performed by such a third party for the data user ensures an unbiased and technically sound outcome.

The ultimate data user also can undertake the data validation process.

What are the Benefits of Data That Have Been Validated?

Data that have undergone the data validation process will subsequently have a known degree of compliance to the applicable quality acceptance criteria. Data qualifiers (usually coded footnotes) will have been assigned, when appropriate, and the usability of the data can be determined. This will benefit the regulators, decision makers, and all who depend on data of known quality.

What is Data Verification and How Does it Differ From Data Validation?

The US Environmental Protection Agency (EPA) also uses a process called Data Verification that complements the Data Validation process. Data Verification focuses on the aspects relevant to sample collection criteria that are specified in a standard or contract. This is a systematic and mechanical determination of whether the samples were collected and handled in accordance with contractual requirements. If Data Verification is done, it is performed before Data Validation.

What is Data Assessment?

This process is makes use of statistical tools and principles to determine if data meet the documented requirements of their intended use. Statistical tests that identify data outliers, distributions, and biases are employed during Data Assessment to objectively determine the conformance to the requirements.

Summary

Data Verification, Data Validation, and Data Assessment are related processes that are used to measure, monitor, and document the degree of compliance to stated quality requirements of analytical data. While all three processes require technical competence, Data Validation requires extensive laboratory systems knowledge and the ability to apply consistent professional judgment during the process.