Usually, it is said that systematic errors can not be handled in a well-defined way, unlike statistical errors.
A) How can systematical errors be calculated for any experimental device or experiment? Maybe does it involve a "calibration" of experimental devices?
B) I have heard that bayesian methods provide THE tool to estimate and calculate systematical errors. How is it done?