Carroll Croarkin, Raghu Kacker
The Statistical Engineering Division supports the development of international standards, particularly those that impact measurement science, via participation on ISO Technical Committee 69 on Statistical Methods and Sub-committee 6 on Measurement Methods. Clients for TC 69 standards are US industry and other technical committees of ISO which produce standards with statistical content.
Documents under development in SC 6 include ISO 11843 on Capability of Detection that outlines statistical methods for quantifying the smallest amout of a trace element which can be detected by an instrument. ISO DIS 10576 on Specification Limits outlines methods for the producer to set specification limts, and methods for the consumer to test product for conformance. A document on Proficiency Testing provides the statistical underpinnings for documents being developed within international organizations, such as CASCO, that are interested in laboratory accreditation.
Requirements of the ISO 17025 and the 9000 series have added impetus to the need for international standards on uncertainty analysis as engineers in metrology and test laboratories find themselves in need of guidance on data collection and analysis methods that are consistent with the ISO Guide to Expression of the Uncertainty of Measurement. Croarkin is convenor of working group, ISO/TC69/SC6/WG7, which has two documents in draft (see http://www.itl.nist.gov/div898/carroll/sc6wg7.htm) on statistical approaches to uncertainty analysis.
One document is intended for test laboratories that do not have the resources or technical expertise to collect and analyze experimental data. This document will be consistent with the EURACHEM Guide to Quantifying Uncertainty in Analytical Measurement in recommeding that repeatability and reproducibility measures from interlaboratory tests be used for computing uncertainties of test results even where the reporting laboratory does not participate in the interlaboratory test. Dr. Steve Ellison (UK) is working with SC 6 on this approach that requires the reporting laboratory to establish a link to the interlaboratory test results via a series of experiments.
A second document is intended for metrology and scientific laboratories that are capable of collecting data to evaluate both short-term and long-term sources of error in the measurement process and have the capability of performing statistical analyses. The two documents will be brought to the committee draft stage in the spring of 2000.
Date created: 7/20/2001