NIST Traceability Numbers - Confused Yet?

[Marc]

Date: Fri, 14 May 1999 11:06:07 -0500
From: TSmith
To: Greg Gogates
Subject: Re: NIST Numbers RE2

Greetings all,

I've been on the road (as usual) and I open my e-mail and see NIST trace numbers as a topic. Hummmm. This subject has reared is QS9000 misinterpreted head again. Well lets ponder a moment about Traceability. NIST has published a position paper about Traceability, The NCSL has published a position paper about Traceability and speaking loosely, A2LA has a position about Traceability. Now, I have copies of all 3 and have read them and spoke with the responsible authors and have identified a interesting parallel. (sit down everyone in the 9000 community and read this very slowly and carefully) NIST Trace numbers are a nist internal tracking systems for jobs (measurement jobs) (order tracking so to say). The number does not mean the artifact passed or FAILED its measurement tolerance. IT could have failed miserably (out of tolerance) but, It has a TRACE NUMBER. The person who cut the PO owns the information about that NIST traceable measurement that failed its tolerance, but you, the person who has put faith in a trace number will never know that it failed because NIST will only provide information to the client that owns the PO that paid for the job. Now in a community that is telling you the belief system you have in a trace number has no foundation, merit or support (even from NIST) ,how many QS 9000 auditors will continue to ask for NIST trace numbers on Certs.? My guess is sign language isn't even enough to convince the deaf ears of the metrology challenged accregistration QS9000, Guide 25 come lately. If they dont get this basic principal of metrology, how the heck are they going to assess competence in a measurement technique, evaluate a procedure or (god forbid) look at an uncertainty budget to determine the contributing elements have been addressed. Does having a NIST number mean The Lab sent it in to nist or someone else the lab used sent it in to nist. (sorry , I got carried away there) The final point (sharp as it may be) " are you looking to see that it was sent to NIST by someone down the chain, or are you concerned that it met the specifications. Hummmmm. The foundation your belief system is built needs to be solid and on pillars with deep rooted support . NIST trace numbers are the far off oasis in the desert of 9000 that will not quench the thurst metrologists.

Tom Smith

-----snippo-----

Date: Wed, 19 May 1999 15:22:07 -0500
From: TSmith
To: Greg Gogates
Subject: Re: NIST Numbers RE4

Greetings Roger,

Welcome to the cybercave of metrology, where those that fear the dark are best to carry a flashlight with extra batteries. I am also a ISO9000 auditor but have converted from the dark side. I say that with great respect for 9000 groupies as for many years I was in conflict with 9000 Vs metrology guide 25. I felt as my background was deeply rooted in metrology, 9000 didn't make good sense to me, but so much for that. I'll save you the learning curve and send you to NIST web site which explains Traceability. Now , after reading that, if you still need more information , I will send you a copy of the NCSL position paper. If still you need to see Moses part the great sea after already seeing the burning bush and hearing the crys from the mothers carrying their first born beliefs of trace numbers to their final resting places, I will send you to the mountain and let you speak with the almighty him self. ( we here on earth just call him Ted). So , go to ***DEAD LINK REMOVED*** and read the position of trace numbers from nist . I hope all 9000 auditors will read this and never again ask for something that is not even supported by nist or any other national lab on this planet. Dont get me wrong about 9000 auditors. I mean no disrespect. I just wish the registratration clan would pop into the 90's real soon and not put their clients through a dance routine that only amuses the registration clan.

Keeping the faith (but somedays its easier than others)

Tom Smith

-----snippo-----

Date: Wed, 12 May 1999 16:33:24 -0400
From: Jack Gale
To: 'Greg Gogates'
Subject: RE: NIST Numbers

Frustrated by NIST Numbers,

This topic has come up more times than I care to remember. I can share these facts with you:

1) NCSL position statement 96-1 on NIST numbers does not carry much weight when "the all-mighty auditor" says you must have them. Bad habits are even harder to break than good ones.

2) I personally have never seen a generally accepted standard that makes this a requirement. Customers that send me this requirement are kindly corrected.

3) We have collected some letters from some of the same OEMs that helped to generate NCSL 96-1; they back up the document. For all that is worth.

4) Sharrill Dittmann of NIST (the person who assigns the NIST numbers) agreed to this position, but NIST is forbidden to arbitrate issues of traceability, so they will not put it down on paper. She also addressed the issue of standards versus those interpretting the standards. The various interpretations are a much bigger problem than the standards themselves. (ASQ-MQD conference, Boston, Sept 1998). She said contact her and she'll repeat what she said. 301/975-2005

5) Hand a certificate of calibration to your customer with NIST numbers on it and ask if the standard used was calibrated by NIST or traceable to NIST. Either answer will likely be wrong. Demonstrate why the passing down of numbers is smoke and mirrors and auditing is the only true way to verify traceability.

Jack

Jack Gale
ASQ-CQE
Essco Cal Lab


-----snippo-----

From: Greg Gogates
Sent: Wednesday, May 12, 1999 12:01 PM
To: iso25
Subject: NIST Numbers

Moderator Note,
This issue just came up at the A2LA conclave. The number is meaningless to all outside of NIST and has no validity towards traceability. The ONLY recognized traceability is through an accredited lab. Would someone else like to elaborate?

Greg

Date: Tue, 11 May 1999 23:09:49 -0700
From: "Kevin J. Haddock"
To: Greg Gogates
Subject: Re: News Article/ISO 10012 Vs ISO 17025

Hello. I have been quite busy lately, and I might have missed this one, but I would like to throw-out the general question to calibration lab people, and to Q.A. people. I have Customers asking me for the NIST trace number on our calibration certs. I do know of NCSL's position on this, but what do some of you do when this issue comes-up? A question for the ISO in the know folks: Where does it state in any ISO requirement that NIST trace numbers should be on the calibration certificates following each piece of equipment used to perform the calibration? What do some of you folks use to prove traceability to the ever-questioning ISO people?

Kevin J. Haddock-Metrologist
Testwave Laboratories


Date: Wed, 12 May 1999 13:22:26 -0700
From: "Nielsen, Larry E"
To: 'Greg Gogates'
Subject: RE: NIST Numbers

This is a relatively simple concept. If your organization's standards or M&TE are calibrated by NIST you should have a report of calibration from NIST citing measured values, a statement of uncertainty and a NIST Test No. Citing the NIST Test No. in your test reports would constitute a demonstration of traceability. If your organization's standards or M&TE are calibrated someplace other than NIST, citing a NIST Test No. in your test reports by itself would be meaningless because it leaves out the intervening steps. You can fill in the blanks here anyway you want but that is the essence of it.

************************************
Larry E. Nielsen
So. Cal. Edison - Metrology
7300 Fenwick Lane
Westminster, CA 92683
(714) 895-0489; fax (714) 895-0686
************************************

[Roger Eastin]

Really interesting (and, at times, confusing) stuff!!! So I guess the bottom line is have the NIST test # and certificate and you're OK when visiting Hi-I-am-your-ISO-auditor-and-I-want-to-see-your-NIST-traceability-evidence land?

[Marc]

It looks to me, if the last fellow (Neilson) is correct (and it makes sense to me) the test number is not relevant unless your cert is from a NIST calibration verification. The whole process is a cascade where you have to go back through each 'supplier' all the way to the NIST contract if you really want to verify the exact traceability thread.