Opportunity at National Institute of Standards and Technology (NIST)
Bayesian Methods for Uncertainty Quantification in High Precision Experiments
Information Technology Laboratory, Statistical Engineering Division
Please note: This Agency only participates in the February and August reviews.
NIST scientists perform high precision experiments in many science disciplines with the requirement that measurement uncertainty estimates be as comprehensive as possible. The results of the experiments are often outputs of complex systems of measurement equations. The uncertainty estimates must therefore include propagation of variability of many inputs of these equations as well as quatification of variability in the measurements due to external factors, both designed and incidental. Bayesian statistical modeling has proven to be especially well suited to the analysis of such experiments.
NIST statisticians have developed analysis methods based on Bayesian hierarchical models for many particular measurement problems; some examples are evaluation of chemical purity using quantitative 1H-nuclear magnetic resonance, estimation of dose response curve due to the toxic effects of NH2-polystyrene nanoparticles on living human cancer cells, and estimation of the Planck constant. The goal of this opportunity is to apply recent advances in Bayesian statistical modeling and computation to complex measurement systems.
 Toman B, Roesslein M, Elliott J, Petersen E (2016) Metrologia 53, S40 -S45.
 Toman B, Nelson M, Lippa K (2016) Metrologia 53: 1193-1203.
 Toman B, Nelson M, Bednar M (2017) Metrologia 54, S16 - S28.
Markov Chain Monte Carlo methods; Complex measurement systems; Hierarchical models; Non-parametric models
Open to U.S. citizens
Open to Postdoctoral applicants