SED navigation bar go to SED home page go to SED projects page go to NIST home page SED Home Page SED Contacts SED Projects SED Products and Publications Search SED Pages

Data Gallery: Wise Mercury Mass Fraction (2-Lab Consensus Value)

Introduction The accurate measurement of toxic elements relies heavily on the expertise of members of Steve Wise's Organic Analystical Methods Group in CSTL's Analystical Chemistry Division. A wide variety of analytes are presented to Wise's group for accurate determination by the best available methods.

The data set analyzed below is for the analyte mercury. The calculation of the final SRM certified value is complicated by the fact that two laboratories took part in the measurement process. As is commonly the case, the data from the two labs were not in statistical agreement. The question at hand is thus: what consensus value should appear on the SRM certificate?

The statistical categorization of this problem is the "2-Lab Consensus Mean". A good discussion of this problem is in Mark Levenson's (et al) paper (2000): An ISO GUM Approach to Combining Results from Multiple Methods", Journal of Research of the National Institute of Standards and Technology, 105, 571.

You may also download the Wise data.

Data

Wise Mercury Mass Fraction Interlab Consensus Mean Data

Prior Distribution 1:
Non-Informative
(~ Classical)

Wise Non-informative Prior Distribution

Prior Distribution 2:
Uniform[.30,.36]
 

Wise Uniform Prior Distribution

Prior Distribution 3:
Normal(.32,.02)
 

Wise Normal Prior Distribution

Posterior Distribution 1

Posterior Distribution from Non-informative Prior

Posterior Distribution 2

Posterior Distribution from Uniform Prior

Posterior Distribution 3

Posterior Distribution from Normal Prior

Posterior Statistics 1

Posterior Statistics from Non-informative Prior

Posterior Statistics 2

Posterior Statistics from Uniform Prior

Posterior Statistics 3

Posterior Statistics from Normal Prior

Date created: 9/21/2001
Last updated: 9/21/2001
Please email comments on this WWW page to sedwww@nist.gov.

[ SED Home |  Bayesian Home |  Data Gallery |  Previous |  Next ]