We discuss methods for analysing calibration data to estimate the effective
area of a pressure balance. These include the so-called P- and Delta P-met
hods, and a method in which the estimate is made by solving a generalized r
egression problem that reflects more correctly the error structure in the m
easurement data. We compare these methods, both from a theoretical point of
view and on simulated data, and show that the differences in the parameter
estimates can be significant compared with the claimed measurement uncerta
inty.