Uncertainty at very low analytical levels – The probabilistic approach |
|---|
| Lampi, E., Boussias, S. |
- Abstract:
- Chemical metrology is a specific field where the process in order to achieve a traceable and comparable results is difficult to be standardized. Moreover, a more holistic approach is required for the evaluation of their reliability. On the other hand, a continuous increasing number of contaminants are intended to be determined in foods and for a high percentage of them e.g. the more toxic, at very low quantification levels. At these low levels, close to detection, the classical way of GUM using statistical techniques for the determination of uncertainty, as the range of values that could reasonably attributed to the measurand, becomes weak.
Probabilities have been used for the determination of type B uncertainties. The probabilistic approach is regarded appropriate in order to determine the concentration level of the analyte and to estimate its uncertainty at the detection and quantification level of the method. More specifically Monte Carlo Simulation and Bayesians can be considered as useful approaches. Bayesians are very useful in case of detection methods where the measurement is not a deterministic procedure but it is based on the capability of the method to classify or identify the analyte from an indication which could lead to false positive or false negative observation.
In the present work, Bayesians are applied in order to estimate the value of the analyte and its uncertainty in the determination of food contaminants at very low levels as well as in identification techniques. - Keywords:
- uncertainty, probabilistic, Bayesians, low concentrations
- Download:
- IMEKO-TC23-2017-119.pdf
- DOI:
- -
- Event details
- IMEKO TC:
- TC23
- Event name:
- 3rd IMEKOFOODS Conference
- Title:
Metrology Promoting Standardization and Harmonization in Food and Nutrition
- Place:
- Thessaloniki, GREECE
- Time:
- 01 October 2017 - 04 October 2017