A calibration is Statistical transformation of a variable to correct for scale or bias, or otherwise to impose the properties of another variable. Example: let two labs use different protocols of chemical analysis. Designate one protocol as the standard and generate a transformation function to convert the data from the second protocol into representation of the standard. calibration will be required as a routine statistical procedure in many aspects of EMAP. New protocols will replace old ones. Surrogate attributes will be extensively measured, to be calibrated to represent the target attribute. Use of multiple laboratories invariably involves laboratory bias, which can be reduced by calibration. analytic equipment requires periodic calibration, and the data generated by such a process can also benefit from statistical calibration. Double-sample methods can utilize calibration in many ways to enhance the precision of estimates based on small subsamples.
List of books: Calibration