Easy and Not-so-easy Statistical Methods of Uncertainty Analyses

This two-day workshop will focus on methods for evaluating and reporting measurement uncertainties based on experience with NIST calibration and measurement services. On the first day, we will develop a toolbox of statistical techniques for uncertainty analysis; explore the role of measurement assurance in developing and validating uncertainty statements; and gain an understanding of how to structure measurement protocols in a manner that will lead directly to uncertainty measures. On the second day, we will examine the history and requirements of the ISO Guide to the Expression of Uncertainty in Measurement and its implications. The afternoon will be devoted to case studies where NIST staff will outlilne procedures for creating uncertainty budgets for mass, dimensional, and electrical measurements.

Supporting subject matter will include discussions on:

1. How to extract pertinent information from calibration designs
2. Uncertainties based on check standard histories
3. Relationship between uncertainty - repeatability - reproducibility
4. Strengths and pitfalls of propagation of error

Time and Place:

February 2 and 3, 1998
Measurement Science Conference (MSC)
Pasadena, California

Developer:

Carroll Croarkin
Rm 353 Bldg. 820
NIST
Gaithersburg, MD 208999
Email: carroll.croarkin@NIST.gov

Sponsers:

NIST and Measurement Science Conference (MSC)

MSC Contact:

John Gerhard
Email: john.f.gerhard@boeing.com