John, On Page 11 of NIST Handbook 133, Internet 2005, is the requirement that "the larger of the two declarations must be verified" (after comparison of SI and non-SI declarations to at least six digits) as you note.
The "Maximum Allowable Variation" (MAV) addresses deviations from the larger *declared value*, not from the "mean value" of the samples, as I asserted below. I don't know why there is not a later posted version of NIST HB 133 than 2005 January. Gene. ---- Original message ---- >Date: Sun, 16 May 2010 13:07:03 -0700 (PDT) >From: "John M. Steele" <[email protected]> >Subject: [USMA:47392] Re: A silly milliliter--Coca Cola commemorative >To: "U.S. Metric Association" <[email protected]> > > I'm not an inspector and never worked in this > industry. Accordingly to published data, they must > compare (by conversion accurate to at least six sig. > figures) the two declared contents and determine > which claim is larger. They then test to that > claim, using a sample average and statistics to > determine whether there is a problem. > .... > > ------------------------------------------------ > > From: "[email protected]" <[email protected]> > To: [email protected]; U.S. Metric > Association <[email protected]> > Sent: Sun, May 16, 2010 3:56:36 PM > Subject: Re: [USMA:47387] Re: A silly > milliliter--Coca Cola commemorative > ____________________________________ > What do inspectors of "net contents" actually do? > > NCWM prescribes a tedious statistical process of > maximum allowed deviations from mean values from > selected samples of a product. > > Are the measured values the SI values or the non-SI > values? > > Is there first a comparison of declared values (SI > vs non SI) on the labels of the product to determine > the larger? > > What is taught in NIST certification courses for > inspectors? > Are inspectors allowed (legally) to do the analysis > only on samples bearing the larger declared value? > > What do inspectors actually do before and after > certification? > > ---- Original message ---- > >Date: Sun, 16 May 2010 02:46:00 -0700 (PDT) > >From: "John M. Steele" > <[email protected]> > >Subject: [USMA:47387] Re: A silly milliliter--Coca > Cola commemorative > >To: "U.S. Metric Association" <[email protected]> > > > > Then I must play the Devil's Nitpicker. > > > > Both the Customary and metric claim must be true > in > > the sense that average net contents must equal or > > exceed the claimed amount...
