Validation in Chemical Measurement

Validationofmeasurementmethodshasbeenusedforavery ciated measurement uncertainty? The answer must be: no. longtimeinchemistry. Itismostlybasedontheexamination Therecanneverbeamechanismorrecipeforproducing - of a measurement procedure for its characteristics such as tomatically 'valid' results becaus [...]
Visa längre beskrivning

Validationofmeasurementmethodshasbeenusedforavery ciated measurement uncertainty? The answer must be: no. longtimeinchemistry. Itismostlybasedontheexamination Therecanneverbeamechanismorrecipeforproducing - of a measurement procedure for its characteristics such as tomatically 'valid' results because one can never eliminate precision, accuracy, selectivity, sensitivity, repeatability, re- theskills,theroleandtheresponsibilityoftheanalyst. producibility,detectionlimit,quanti?cationlimitandmore. ISO 9000:2000, item 3. 8. 5 de?nes validation as 'con?r- When focussing on quality comparability and reliability mation by examination and provision of objective evidence in chemical measurement, the ?elds of interest to this Jour- that the requirements for an intended use are ful?lled'. The nal, one stumbles into various interpretations of the term revised edition of the VIM ('VIM3'), is likely to ?ne-tune validation. It is one more example of a term which is used thisde?nitionoftheconcept'validation'tobe'con?rmation sometimes very consistently, sometimes very loosely or in- through examination of a given item and provision of - deed ambiguously.Since the term is very common in the jective evidence that it ful?lls the requirements for a stated chemical community, it is important that its meaning be intendeduse'. nd clear. Turning to the 2 edition of the International Vo- Lookingatsimplepractice, manypeoplearelookingfor cabulary of Basic and General terms in Metrology (VIM) aformaldecisionthatagivenmeasurementmethod automat- (1993), surprisingly we do not ?nd a de?nition. Webster's ically gives them 'valid' i. e. reliable results. One wonders Dictionary of the English language (1992) tells us that val- what this has to do with 'stated intended use'. Reliab- idation is 'making or being made valid'. Obviously valida- ity clearly is a property of a measurement result.