Confidence Intervals on Measures of Variability in R&R Studies
Richard K. Burdick, Greg A. Larsen · Journal of Quality Technology · 1997
Several methods have been proposed for constructing confidence intervals on measures of variability in a classical repeatability and reproducibility study. These methods are compared using computer simulation to determine if the intervals maintain the stated level of confidence and to compare the average interval lengths. Recommendations are provided for selecting an appropriate method and for selecting the number of operators and samples to use in a study.