Discrepant measurements and experimental knowledge in the early modern era (Q2458889)

Before the 19th century the average was hardly used to deal with observational discrepancies. In considerable detail the author studies how observers dealt with discrepant measurements before the averaging became probabilistically justified and generally accepted. This yields a complex story. Essentially early observers like Brahe, Kepler, Boyle, Hooke and others were looking for the unique correct value. Great emphasis was put on the experimenter's skill. Unwanted measurements were simply ignored. Means were sometimes calculated, but never as a methodological principle. Newton is in this respect the exception. Near the end of his life he applied averaging methods in his redating of ancient chronology.