Stephen BissShort Calibration Interval and Reliability
Updated: Aug 9, 2022
Why is disclosure of information about the "calibration" of the approved instrument essential to any assessment of the reliability of a particular subject breath test result at time of use?
Why does reliability depend on a recent calibration or re-calibration?
Purpose:
To introduce the concept of "calibration interval"
To obtain admissions as to the importance of calibration interval in the international scientific literature
To obtain an admission that neither the ATC/CFS nor the manufacturer mandate a calibration interval for approved instruments, though the manufacturer so specifies for ASDs
To introduce some international standards on reliability including ISO 17025
To introduce the concept of "uncertainty growth"
To connect "reliability", "calibration interval", and "uncertainty growth" "over time"
Sample cross-examination of CFS expert on the international literature related to calibration interval:
Q. Mr. P, some questions about the concept of "calibration interval". And, again, calibration is something that’s done by the factory or the factory authorized service centre. Right?
A. Yes.
Q. And I want to ask you some questions about you had mentioned literature on calibration interval. And I just wanna first of all show you a paper...
A. Oh, sorry. Here it is.
Q. ...by an author Allen Bare doing a paper at the 2006 NCSL International Workshop and Symposium on Simplified Calibration Interval Analysis. And I just want to suggest to you that the calibration interval for a measurement instrument is often based on the manufacturer’s recommendations?
A. Yes, such as the approved screening devices.
Q. All right. But, unfortunately, with respect to CMI, there is no manufacturer’s recommendation with respect to calibration intervals for that instrument?
A. Not that I'm aware of, no.
Q. So, we are left in Canada with a situation where all we have is a recommendation by the Alcohol Test Committee for a periodic inspection, generally, annually, to check the instrument against the manufacturer’s specifications?
A. Yes.
Q. There is no recommended practice in Canada with respect to a specific calibration interval for approved instruments?
A. Correct.
Q. But out there in the scientific world, there is an expectation that measurement instruments do need to receive calibration at regular intervals?
A. I can't speak to what recommendations are done in the United States or in Europe. I can only speak to what’s done in Canada and what you’ve described is accurate.
Q. On page 2 of that document, there’s a list of a number of international standards which speak in terms of calibration interval. The first one that’s listed is ISO17025. That’s an instrument we talked about on the – oh, I'm sorry, a standard that we talked about on the last day.
A. Yes.
Q. That talks about there being a calibration certificate. There’s another standard referred to there called Ansi, A-N-S-I/NCSLZ540. Do you know what that organization is?
A. No. idea.
Q. Okay, all right. But in any event, that standard seems to imply that calibration shall be calibrated or verified at periodic intervals established and maintained to assure acceptable reliability. And there’s also another document called ISO9001 which talks about the verification at specified intervals?
A. Yes.
Q. And there's also a military standard, 45662A. I think that’s a US military standard?
A. I couldn’t confirm that.
Q. All right. But my question is that do you agree with the comment made on the third-last paragraph on that page where it says that “As calibration intervals are reduced, reliability should increase due to a higher percentage of intolerance calibration results”?
A. Sorry, I'm looking at the wrong paragraph. Which one were you...
Q. The paragraph that begins with “In addition.”
A. Oh, okay. Yes.
Q. There’s a line there that says – the last line in that paragraph says “Therefore, as calibration intervals are reduced, reliability should increase due to a higher percentage of intolerance calibration results.”
A. That is what that says.
Q. All right. Do you agree with that concept, generally?
A. Yes. As a principle, yes.
Q. So, generally speaking, reliability increases if you have frequent re-calibrations?
A. Yes.
Q. If the calibration interval, as a length of time, is reduced?
A. Yes. And that would be dependent on the technique that’s being used and the circumstances under which it is being used, which could be different for breath testing than for analytical testing in the laboratory.
Q. All right.
A. Again, the calibration of the Intoxilyzer 8000C is verified, as per the language here, with every breath test.
Q. The calibration isn't verified across the measuring interval?
A. Correct.
Q. The calibration – there is a control test at 100 milligrams per 100 mls?
A. Correct. As it’s been since before I was a forensic scientist and with the instruments before that including the Breathalyzer 900 and 900A.
Q. Linearity is not checked?
A. Correct.
Q. I'm showing you another document from National Aeronautics and Space Administration of a Measurement Uncertainty Analysis and a chapter on – sorry, at page 121 of that book, talking about the concept of "uncertainty growth". And the statement there is that “An error or bias in an MPTE...” I think that’s a measuring and testing equipment, “...attribute or parameter may increase or remain constant or decrease. The uncertainty in this error, however, always increase with time since measurement or calibration. This is the fundamental postulate of uncertainty growth.” Do you agree with that statement?
A. That’s what it reads, yes.
Q. But do you agree with that statement?
A. Yes, in principle.
Q. All right. Immediately under figure 9.1, measurement of uncertain measurement of uncertainty growth, there’s a statement that says “Uncertainty growth over time corresponds to an increase in out of tolerance probability over time or equivalently to a decrease in intolerance probability or measurement reliability over time.” Do you agree with that statement?
A. Yes. Again, as a principle.
Q. All right.
MR. BISS: So, Your Honour, the simplified calibration interval analysis, could that be Exhibit Number...
CLERK REGISTRAR: Fifty-one.
MR. BISS: Fifty-one?
THE COURT: Yes.
EXHIBIT NUMBER 51: Simplified Calibration Interval Analysis – produced and marked.
MR. BISS: And the NASA uncertainty growth document be Exhibit 52?
EXHIBIT NUMBER 52: Measurement Uncertainty Analysis Principles and Methods, 2010 – produced and marked.
[See also "Establishment and Adjustment of Calibration Intervals Revising NCSLI Recommended Practice RP-1" Mark Kuster, Pantex Metrology, 2007 NCSL International Workshop and Symposium, excerpt below]
#crossex #calibrationinterval #uncertaintygrowth #overtime
#uncertaintygrowth #calibrationinterval #reliability #overtime