A recent paper by Ignacio Lira in Metrologia 43 (2006) S231-S234 addresses a well-known problem in combining information from interlaboratory evaluations. Lira presents an expression, which he claims to be the kernel of a Bayesian posterior probability density function (pdf) concerning the value of the common measurand. Lira claims that his expression encodes the collective state of knowledge irrespective of whether or not the data are consistent. We note that the expression claimed by Lira as a pdf for the value of the measurand is baseless.
Pub Type: Journals
Bayesian statistics, interlaboratory evaluations, key comparisons