An Exercise in the Real World of Design and Analysis
Lorraine Denby, Jim Landwehr, C. L. Mallows · The American Statistician · 2001
In order to win a contract to manufacture certain devices, it was necessary for our company to prove that we could reliably measure the devices at our factory to within a stated tolerance of what corresponding measurements would be if made by the customer. Our measuring equipment was different from that of the customer, so it was necessary to calibrate and certify our measuring process relative to the customer's. This raised challenging and interesting problems, both for the design of the calibration experiment and for the data analysis and modeling of the experimental results.