NIST Interactive Data Evaluation Assessment Tool

Ceramics WebBook: Property Data

Data evaluation levels provide a means of classifying property data according to how the data were acquired, processed, and reported. The present protocol is based on the seven categories of acceptable data that are discussed in the NIST Recommended Practice Guide SP 960-11. These categories are used in both the NIST Structural Ceramics Database (SCD) and the NIST Database on High Temperature Superconductors (HTS). The protocol provides an interactive, step-by-step procedure in the manner of a decision tree. Based on the user's responses to a series of questions, the protocol determines the appropriate data evaluation level for a given set of data.

Reliability may be the most fundamental concern encountered in any application of materials property data. What constitutes reliable data, however, depends on the purpose of the data set and how it will be used. To establish a degree of reliability in the context of “purpose” and “use” requires both quantitative and non-quantitative criteria. Details of material composition, chemical phase, microstructure, and methodology, in particular, often play a critical role in assessing reliability. Consequently, NIST Recommended Practice Guide (RPG) SP 960-11 recommends the use of a data quality indicator called the “data evaluation level” and provides a protocol for determining its value.

In essence, a data evaluation level summarizes the status of the evaluation process that has been applied to the set of data. At the base of the evaluation paradigm, there are two counterbalancing mandates: (1) ensure the reliability of the data; and (2) do not reject data without cause.

In the NIST protocol, there are seven classes of accepted data. The seven corresponding data evaluation levels are:

• Certified
  • standard reference values, usually established by recognized national standards laboratories
• Validated
  • values confirmed or supported by means of correlations and model calculations
• Qualified
  • the broadest category
  • data satisfy at least the minimum acceptance criteria
  • data are generally compatible with other related values in the database
• Commercial
  • manufacturer's data
• Typical
  • usually values or ranges of values derived from surveys or broad reviews
• Research
  • preliminary values from work in progress
• Unevaluated
  • all other accepted data
  • commonly, the measurement method is unknown or described insufficiently
  • commonly, useful supplementary data.

The present program has been developed in the hypertext markup language (HTML) to provide a convenient interactive tool by which the data evaluation level of any set of property data may be assessed using the same protocol that is applied to the data in the SCD and the HTS.