< back to main site


Statistical error modelling.

Cox, M G; Harris, P M (2004) Statistical error modelling. NPL Report. CMSC 45/04

[img] Text

Download (563kB)


Statistical error modelling is regarded as a process that involves developing relationships between measurement data, the required measurement result(s), and the measurement deviations (or ‘errors’) associated with the measurement data. The problem of evaluating the measurement result(s), and the associated uncertainty, in cases where a statistical model is assigned to the values of the measured quantities is considered. A number of cases are addressed, such as where the measurement deviations arise from random effects only or from a combination of random and systematic effects, and where the statistical model for the effects is completely or only partially specified. The focus is on regression and its application to the calibration of an instrument or measurement system. A number of examples are discussed, concerned with characterising the performance of a clock, developing RF coaxial frequency standards, modelling spectral characteristic data, and assessing the roundness of a nominally circular artefact.

Item Type: Report/Guide (NPL Report)
NPL Report No.: CMSC 45/04
Subjects: Mathematics and Scientific Computing
Mathematics and Scientific Computing > Measurement Uncertainties
Last Modified: 02 Feb 2018 13:16
URI: http://eprintspublications.npl.co.uk/id/eprint/3034

Actions (login required)

View Item View Item