Related Experiment Videos
[Comparison of 2 methods for calculating uncertainty in laboratory analysis]
1Cátedra de Medicina Preventiva y Salud Pública, Facultad de Medicina, Universidad de Cantabria. llorcaj@medi.unican.es
Gaceta Sanitaria
|March 29, 2001
Summary
The bootstrap method provides more accurate uncertainty estimation in laboratory quality control compared to the delta method. This simulation-based study highlights bootstrap
Area of Science:
- Laboratory Science
- Statistical Analysis
- Quality Control
Background:
- Accurate estimation of uncertainty is crucial for reliable laboratory quality control.
- Existing methods like the delta method have limitations in diverse analytical conditions.
Purpose of the Study:
- To compare the performance of the delta method and a bootstrap-based method for estimating uncertainty in laboratory quality control.
- To evaluate method robustness under various environmental conditions and variable relationships.
Main Methods:
- Computerized simulation was employed to compare the delta method and the bootstrap method.
- Simulations incorporated varied environmental conditions and inter-variable relationships.
Main Results:
- Bootstrap method yielded higher and more nominal coverage percentages for confidence intervals compared to the delta method.
- Bootstrap demonstrated consistent performance across different conditions, including correlated and unmeasured variables.
- Delta method showed dispersed coverage percentages, with instances of both overly conservative and inadequate coverage.
Conclusions:
- The bootstrap method is a more accurate and robust approach for estimating uncertainty in laboratory quality control.
- Bootstrap-based uncertainty estimation is recommended for improved reliability in laboratory settings.