Bootstrapping with Constraints: Analysis of Scattering Asymmetry for Polarized Beam Studies,
Abstract
In polarized beam studies, an asymmetry statistic of physical interest is an estimate of the ratio of the difference and the sum of the Poisson rate parameters for two scattering processes. Typically, an additive background signal contributes to measurements of each scattering process. Background is measured in a third experiment. Data is corrected by subtracting measured background. When the measured background is larger than one of the other measurements, the asymmetry computed from the background corrected data is nonsensical. For such cases, true asymmetry and an associated conservative interval are estimated using a bootstrap procedure. Bootstrap replications of the observed data satisfy a constraint that insures physically meaningful results.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1992
- Accession Number
- ADP007156
Entities
People
- Kevin J. Coakley
Organizations
- National Institute of Standards and Technology