Statistical Methods Metrology

3 Credits · Level: Graduate, Undergraduate

The course teaches concepts and methodologies of statistical methods and interplay between statistics and measurements in metrology through in-depth discussion of the best statistical methods for the assessment and expression of measurement uncertainty. This is a topic that is essential to metrology but not commonly known to even experienced statisticians. Topics to be covered, but are not limited to, include (1) probability modeling of measurement error, (2) elementary one-/two-sample statistical methods and measurement and inter-lab comparison, (3) intermediate-level statistical methods to quantify the importance of sources of variability in measurements, (4) statistical analysis of repeatability and reproducibility (R&R), (5) linear regression and calibration, (6) quantization/digitalization effects, (7) quantifying how measurement precision affects one’s ability to detect differences between measurands, (8) models and methods of spatial statistics applied to environmental problems. The course also introduces R language and its application for environmental statistics and discusses statistical modeling methods used by EPA. Overall, the course is designed to help students understand the quality and reliability of the measurements, acquire the ability to parse a measurement problem, identify the critical sources of variability and bias, and develop approaches to address these using the core concepts of uncertainty, traceability, validation, and quality systems.