Data analysis & uncertainty

For simple measurements, mathematical models are usually straightforward, and it is clear how to analyse the data and calculate the uncertainty.

The BIPM Joint Committee on Guides for Metrology Working Group 1 (JCGM-WG1) has provided a large suite of guidance documents, including the well-known GUM, the Guide to the Expression of Uncertainty in Measurement.

More Guides on Uncertainty Estimation

Based on these documents, other bodies have written guides for more specific application domains, such as:

the EURACHEM guide for analytical chemistry

Good practice in evaluating measurement uncertainty, produced in the EMPIR project EMUE, and the project consortium works closely with the Mathmet community. The document gives examples dealing with more complex, practical problems where the application of the principles outlined in the GUM may be not straightforward.

Click the italicised links above to see the full guides.

However, there are still many practical cases left in which it is not clear how the data can best be modeled and how to calculate the uncertainty. This can be due to a complex mix of systematic and random effects, unknown model errors, a very small or a very large amount of measurement data, hidden correlations or other complex structures in the measurement procedure or data model. Also, for regression and inverse problems optimal guidance is sometimes missing.

The Mathmet research topic ‘Data analysis and uncertainty’ will address these cases. Mathmet is actively engaging with stakeholders and collecting such research topics. These topics will be grouped and ranked before being added to the Mathmet Strategic Research Agenda.

Do you have a measurement science related need in this area? Or would you like to collaborate with Mathmet on a research topic in this area? Please don’t hesitate and contact us.

Related metrology research projects

Novel mathematical and statistical approaches to uncertainty evaluation

Short Name: Uncertainty, Project Number: NEW04

Software for evaluating PET cardiac perfusion imaging uncertainties for more accurate diagnosis

Short Name: TracPETperf, Project Number: 19SIP04

Improved metrology for quantitative MRI

Short Name: iMET-MRI, Project Number: 20NRM05

Metrology for low-frequency sound and vibration

Short Name: Infra-AUV, Project Number: 19ENV03

Metrology of automated data analysis for cardiac arrhythmia management

Short Name: MedalCare, Project Number: 18HLT07

Advancing measurement uncertainty - comprehensive examples for key international standards

Short Name: EMUE, Project Number: 17NRM05

Metrology for the Factory of the Future

Short Name: Met4FoF, Project Number: 17IND12