ILAC-P14:12/2010

[WestCoastCal]

Anyone receive an audit finding, observation or recommendation pertaining to expanded uncertainty significant figures? IAW ILAC P14:12/2010, section 6, ILAC Policy on Statement of Uncertainty of Measurement on Calibration Certificates step 6.3, "The numerical value of the expanded uncertainty shall be given to, at most, two significant figures. Further the following applies:
a) The numerical value of the measurement result shall in the final statement be rounded to the least significant figure in the value of the expanded uncertainty assigned to the measurement result.
b) For the process of rounding, the usual rules for rounding of numbers shall be used, subject to the guidance on rounding provided i.e in Section 7 of the GUM."
I was wondering if an auditor will fully enforce 2 signifcant figures reporting requirement. (I know, dumb question, but...)
If your lab was recently assessed any feedback would be appreciated.

[RFCAL]

Your guess is right. The auditors are not letting anything slide as they cannot have issues be resolved while they are at your facility inspecting. In other words, if the inspector sees a cracked cal void sticker on a standard, they are obligated to report that as a deficiency. Before, they would not report anything you could correct while they were there.

[WestCoastCal]

ILAC P14 has been revised (2013), I haven't had a chance to download/read it yet.  There has been much discussion on this topic on the elsmar dot com forum. Back to my original question, does anyone have experience receiving an observation, finding or recommendation due to significant figure reporting during an accreditation audit?  If so, was it resolved?

[Duckbutta]

This post is more evidence of what a joke accreditation has become, or more accurately, always was. Not an attack on you WestCoastCal, just an observation of the insanity.

[RFCAL]

That is true--ILAC P14 does state to report UNC's to 2 significant digits after the decimal point, starting with any lab being inspected this year. And yes, they will report you--no leeway. This entire uncertainty thing is getting out of hand!!

[Duckbutta]

The problem with your typical uncertainty analysis is that it fails to take into account the greatest source of uncertainty in any measurement - the technician. This is especially true in dimensional but also true in the other disciplines.

[RFCAL]

That should be taken care of in the operator Bias input for the total uncertainty. You should be letting 2 different technicians take your "A" samples and then average them. The more technicians you average, the lower that number will be. In electronics, you can see errors real quick. When reading a DMM, if you are looking for + numbers and you see - numbers, it's pretty logical the technician messed up. Time to re-train him and take another set of readings. For all of our Generate and Measure disciplines, we use an automated program. Still, we have to connect and dis-connect correctly. I don't know whatever happened to the old 4:1 & 10:1 method, Seemed to work pretty well to me.

[Lasario]

IMHO uncertainties are the most important factor in a calibration. Without uncertainties you have no traceability. Without uncertainties maintaining Z540.3's 2% rule is nearly impossible. Without uncertainties how can you make a compliance statement about a TI being within tolerance?  As metrologists  and calibration technicians we should be encouraging the use of uncertainties in every calibration, not just the ones that someone says to do. Uncertainties are just not for the statistician anymore.

[beadwork]

Uncertainty of a measurement is suposed to show how well the measurement was made.

[OlDave]

I look at it as more of an estimation, based on a statistical probability, of how well a measurement could have been made. It is never an exact value for an individual data point.

Personally I think that there is way too much BS and wasted time, money and effort put into needless uncertainty analysis for the majority of equipment. For top level instrumentation it is invaluable. For working items such as a 5% rotometer, a digital timer used to measure a process to within a minute or so, the DVM that rides around in our toolbag, it makes no sense what-so-ever. But the quality manually states that they will be calibrated by "accredited sources". What a waste!

[RFCAL]

I agree, this is a big waste of time & money for most people. The majority of us are not NASA type labs.

[metrologygeek]

It's the biggest reason why Z540.3 has not been widely accepted by the metrology world. To require a full up uncertainty analysis on every single measurement is just adding huge costs with no technical benefit. That's why 17025 has become the defacto world-wide standard and Z540 will become more and more marginalized.

[Duckbutta]

Without uncertainties you have no traceability.
Without uncertainties how can you make a compliance statement about a TI being within tolerance? 

These statements are just factually incorrect. We've been doing both for years.

[RFCAL]

We have had Zero requests for Z540.3 although we are accredited to that standard. I am hoping that the big Defense Companies will continue to ignore it and NOT include this in their contracts.
I agree with Duckbutta, we have been traceable 40 years ago without uncertainties and have not had that questioned at all!!

[Lasario]

I.B    Questions about Establishing Metrological Traceability
I.B.1 What is involved in establishing metrological traceability?

    Per the VIM definition, metrological traceability is established through "an unbroken chain of calibrations, each contributing to the measurement uncertainty." In practical terms, this means having in place, for each link in the chain:

        a clearly defined particular quantity that has been measured

        a complete description of the measurement system or working standard used to perform the measurement

        a stated measurement result or value, with a documented uncertainty

        a complete specification of the specified reference at the time the measurement system or working standard was calibrated against it

        an 'internal measurement assurance' program for establishing the status of the measurement system or working standard at all times pertinent to the claim of traceability

        an 'internal measurement assurance' program for establishing the status of the specified reference at the time that the measurement system or working standard was calibrated against it.

Copied from the Nist Website

By exception, and where it has been established during contract review that only a
statement of compliance with a specification is required, then the measured quantity
value and the measurement uncertainty may be omitted on the calibration certificate.
The following shall however apply:

- The calibration certificate is not intended to be used in support of the further
dissemination of metrological traceability (i.e. to calibrate another device);

- As specified in ISO/IEC 17025:2005 clause 5.10.4.2, the laboratory shall
determine the uncertainty and take that uncertainty into account when issuing the
statement of compliance; and

- The laboratory shall retain documentary evidence of the measured quantity value
and the uncertainty of measurement, as specified in ISO/IEC 17025 clauses
5.10.4.2 and 4.13, and shall provide such evidence upon request.

From the OP Referenced Document