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

Blaza Toman Presentations

  Bayesian Analysis of Interlaboratory Data

Blaza Toman November 8, 2001
NIST

Abstract This talk will introduce the Bayesian approach to the researchers involved in the SRM certification. It will aim to show the advantages of the Bayesian paradigm, illustrating the methods on the data from the Lake Superior Fish Tissue SRM.
  Statistical Analysis of the Results of the CCPR Key Comparison on Near-Infrared Spectral Responsivity

Blaza Toman
October 21, 2001
Washington, DC

Abstract This is a presentation of the statistical analysis of a recently completed international comparison of near infrared spectral responsivity executed under the direction of the Consultative Committee on Photometry and Radiometry (CCPR). The comparison was structured in a star pattern, with the National Institute of Standards and Technology (U.S.) the host laboratory. A total of fifteen national laboratories participated in one of 4 rounds. Three indium gallium arsenide (InGaAs) detectors were sent to each laboratory, one of a particular type and two randomized detectors from 5 different vendors. NIST measured the photodetectors between rounds to establish their radiometric stability. The goals of the statistical analysis were to assess the agreement between the laboratories and further to arrive at a common reference value for each wavelength. Cluster analysis of the responsivity data together with data analysis of the combined uncertainties was used to identify groups of laboratories that were in particular agreement. A Bayesian hierarchical model was used to calculate the common reference values and to assess the agreement between laboratories and the reference value.
Bayesian Sample Size Calculations

Blaza Toman
Joint Statistical Meetings (JSM)
August 2001

Abstract This was an invited poster presentation giving a review of methods for Bayesian optimal sample size selection. The emphasis was on the computational aspects of the methods. The session, which was devoted to sample size and power calculations, included presentations from academics as well as software developers such as SAS and Insightful. The session was very well attended.

Date created: 8/28/2001
Last updated: 8/28/2001
Please email comments on this WWW page to sedwww@nist.gov.

SED Home |  Bayesian Home |  Previous |  Next ]