Introduction to Measurement Uncertainty 
This course covers the fundamentals of calculating uncertainty in measurement. Based on our new "Introduction to Statistics in Metrology" textbook (Springer, 2020), the course begins with a review of basic concepts and the terminology of metrology. Next, details of Type A and Type B uncertainty evaluations are presented using a direct measurement model. Uncertainty propagation using an indirect measurement model is then presented using both the JCGM 101: GUM and Monte Carlo approaches. The topics of design of experiments and curve fitting in metrology are presented next, followed by a section on special topics in metrology. These topics include statistical process control (SPC) in metrology, a Bayesian alternative to uncertainty analysis, evaluation of binary measurement systems, and sample size determination in metrology studies. Much of the material is illustrated using case studies from our work, and the analyses are performed using our open source uncertainty calculator.

Quantifying Measurement Decision Risk and Defining Decision Rules
This class provides a statistical framework for incorporating measurement uncertainty into the assessment of risk. Topics include Test Uncertainty Ratios (TURs), false accept and false reject probabilities, guardbanding, and the effect of measurement bias on decision risk. These topics are presented both in the context of manufacturing pass/fail decisions and calibration lab pass/fail decisions. Details regarding the Test Uncertainty Ratio (TUR), including assumptions and limitations are covered, along with different strategies for guard-banding to mitigate risk. Finally, an overview of some common decision rules is discussed. The risk calculations introduced are supported by our uncertainty calculator.

Quantifying Uncertainty in Curve Fitting Problems
In this course, methods for fitting linear and nonlinear curves to experimental data are presented, with an emphasis on uncertainty in the resulting curves. The linear calibration problem, in which a "calibration curve" is used to relate an accurate reference measurement to a less accurate measurement, is explored in detail. Additional topics include assessing uncertainty due to drift in calibration assets, and determining optimal calibration intervals for a set of measurement devices.

Statistics and Metrology in Uncertainty Quantification (UQ)
This course teaches the student how to properly account for input parameter uncertainties in a UQ framework, in which the “measurement” is the output of a computer code. Sources of uncertainty that are covered include physical boundary conditions, temporal boundary conditions, and material property uncertainties. A case study approach is taken using examples from real world electrical, thermal, and nuclear modelling problems.

Statistical Process Control (SPC)
Our basic class covers an introduction to the various aspects of quality, representations of variability, an overview of key SPC tools, and a detailed look at commonly used control charts. More advanced classes are available upon request.

Design of Experiments (DOEx)
Our basic class discusses the importance of efficient experimentation, followed by an overview of DOEx tools. The specific designs presented include full factorial designs, fractional factorial designs, response surface designs, and ANOVA (variance components) designs. The importance of each of these designs in measurement studies is emphasized, along with numerous case studies to illustrate the material.

Statistical Computing in R
This course provides an in-depth introduction to the R programming language with an emphasis on using R for data analysis and statistical programming. The first part of the course covers the foundations of R such as syntax, data types and control structures. The second part focuses on data analysis workflows in R including data cleaning/manipulation, visualization (using the popular ggplot2 package), and basic statistical analysis. All concepts are presented with hands-on exercises using real data. Additional topics can include regression and ANOVA analyses, Monte Carlo methods, and advanced data visualization in R.

Additional classes available upon request. All classes can be offered virtually as needed.

Classes