Raul R Cordero and Pedro Roth 2005 Metrologia 42 L15 doi:10.1088/0026-1394/42/2/L01
Raul R Cordero1,2 and Pedro Roth1
Show affiliationsBy international accord, the standard uncertainty of a directly measured quantity is taken as the square root of the variance of its probability density function (pdf). The pdf can be inferred from the dispersion of unbiased observations obtained by repeatedly measuring the quantity involved. If the quantity is not repeatedly measured, the apparent scarcity of experimental data may require consideration of information other than experimental data to assign the pdf and to evaluate the uncertainty; in this context, an evaluator tends intuitively to consider information linked with his experience. In this work, we report on a method, based on the theorem of Bayes, to assign the pdf and to evaluate the uncertainty of a quantity measured only once. This method allows us to include in the uncertainty evaluation the information concerning the evaluator's experience. It gives theoretical support to the intuition since it satisfies the formalities of statistical inference. Two examples are presented.
06.20.Dk Measurement and error theory
02.50.-r Probability theory, stochastic processes, and statistics
Issue 2 (April 2005)
Received 16 December 2004
Published 21 March 2005
Raul R Cordero and Pedro Roth 2005 Metrologia 42 L15
John S Cullen et al 1999 Meas. Sci. Technol. 10 812
Shili Xiao et al 2009 J. Phys.: Conf. Ser. 188 012015
I V Fialkovsky and D V Vassilevich 2009 J. Phys. A: Math. Theor. 42 442001
Brahim Lounis and Michel Orrit 2005 Rep. Prog. Phys. 68 1129
J E Allnutt and J A Staniforth 1971 J. Phys. E: Sci. Instrum. 4 730
W Budde 1980 Metrologia 16 89
Nathan Berkovits and Chris Hull JHEP02(1998)012
H S Allen 1930 Proc. Phys. Soc. 42 372
E K Sandeman 1923 Proc. Phys. Soc. London 36 132