6 Errors, corrections and uncertainty

6.1 Summary

Errors can be divided broadly into two types - systematic and random. Systematic errors are repeatable and can usually be predicted and hence corrected. Errors caused by processes that are either truly random or too difficult or too costly to predict are usually classified as random errors and treated statistically. An uncertainty value associated with a measurement or a calibration is conventionally interpreted as the scatter or spread of a random process. Conventional practice is to correct a measurement for all known systematic errors, and estimate an uncertainty associated with only the random components of the error. Corrections are seldom perfect and uncertainties associated with corrections are included in the overall uncertainty estimate. Uncorrected systematic errors should not be included in an uncertainty estimate.

6.2 Introduction

No measurement is ever correct. Every measurement has an associated error. The true value of a measurand is never known. Measurement error is the difference between the (unknown) true value and the measured value. The value of the error is never known. If we know the error in magnitude and sign we can correct the measurement to obtain the true value of the measurand. In practice the best we can do is correct known systematic errors to the extent that is possible or practical and estimate a range of values within which the random component of the error is likely to lie. Such an estimate is termed the uncertainty of the measurement.

This topic examines the classification of measurement errors into systematic and random, and how the two types of errors are treated in the estimation of uncertainty.

This topic is incomplete.

©2002 Martin Turner B.Sc. (Eng) Ph.D.
Engineering and Measurement Consultant
12 Goodman Place, Cherrybrook, NSW 2126, Australia
Tel: 0403-007 305 (International: +61-403-007 305)
Email: mjturner at biccard.com
Disclaimer The views expressed and information provided in these documents are the opinions of the authors and do not represent specific advice on any topic.

First published: 27 Nov 02 Last modified: 29 June 03