Ensuring the potential for an intended quality of a measurement result must be done before and during the measurement, not after it (when it is too late)

Abstract

‘‘Providing quality’’, be it in a material thing or in a ser-vice, is a very fashionable slogan nowadays. Yet it is stilldifficult to realize it unambiguously, due to an insuffi-ciently simple and clearly understandable definition of‘quality’.What is ‘quality’?Meeting the end user’s requirement?What if this is not precisely formulated, as is many timesthe case?And, if it is, is it based on a real need?Has the end user arrived at a justified, and justifiable,formulation?Was the formulation specified so as to be understoodunambiguously?Fortunately for the quality of a measurement result, thenew VIM [1] is very helpful in answering those questions.A requirement on a measurement result can be formu-lated as a target measurement uncertainty (TMU):‘measurement uncertainty specified as an upper limit anddecided on the basis of the intended use of the measure-ment result’ [VIM-2.34] [1]. Thus, the TMU can be seen asa measure of the ‘fitness-for-intended-use’. That reflectsthe importance of a formulation of the use of that mea-surement result, which is understood by the analyst interms of the intention of the end user.The concept of TMU offers a tool which both serves• the end user: it forces him/her to think about his/herintention of the use of a measurement result, and,• the analyst: it enables him/her to examine how (s)hethinks to obtain a measurement result fit for theintended use, prior to the measurement i.e. at the timeof its planning.Whether the analyst will have any chance of achievingthe right ‘quality’ expected by the end user can then beevaluated against a jointly agreed TMU, which is anumerical measure of the fitness-for-intended-use of thatresult.The importance of the potential of the combination of aTMU and a fitness-for-intended-use cannot be overesti-mated. It couples better thinking by the end user (in termsof his/her real need) to the ability of the analyst (in terms ofmeeting that need).The feeling of ‘quality’ is born in the perception of theend user that his/her need is met. The feeling of ‘quality’ isborn in the perception of the analyst that (s)he might beable to meet that need.There are various ways to build-in quality of the mea-surement result before the measurement (a priori): thechoice of a ‘quality calibrator’, or the choice of a shorter‘calibration hierarchy’ entailing a shorter metrologicaltraceability chain of the measurement result, or the choiceof a good measurement procedure including a good mea-suring instrument and entrusting the measurement to agood analyst.A convincing feeling of ‘quality’ in both the minds ofanalyst and end user is good. But, let us note that a feelingof exceeding the claimed quality, makes the feeling of‘quality’ really unavoidable, and, at the same time, highlysatisfactory. Slightly exceeding the expectations of both theproducer (the analyst) and the end user (the customer) ofthe measurement result, seems to be the simple secret ofremoving any doubt on the ‘quality’ of the measurement