Response Drifts Over Time
Updated: Aug 9, 2022
In an article relied upon by the SCC in St. Onge, Brian Hodgson stated:
Defence lawyers need to educate themselves about the meanings of "drift" and "over time". What is it that actually "drifts"? What happens "over time". Are these concepts important to clarify during any cross-examination of a Crown expert? Can the reality of drift over time be satisfied by single point control tests? Is more required to maintain reliability over time?
The ATC and CFS regularly opine that any drift in instrument response will always be caught by a single point control test, is not supported by any empirical study. Is there any empirical study that supports this opinion? Is this mere hypothesis without empirical study?
This is part of a cross-examination. It obviously needs to go much further. However, it appears that there isn't any published empirical study that supports this CFS/ATC hypothesis.
To establish that the hypothesis by the ATC/CFS, that any drift in instrument response (what Hodgson describes as significant drift in accuracy and precision over time) will always be caught by a single point control test, is not supported by any empirical study.
To establish that there is no study abstracted in Wigmore on Alcohol that establishes the ATC/CFS hypothesis empirically.
To confirm that the new calibration curve created after a re-calibration compared with the old calibration curve prior to recalibration can move up, move down, move sideways, or rotate in either direction around an axis that may or may not be at 100 mg/100mls.
Q. Mr. P, did you find any other scientific literature that we should be looking at? A. I didn’t. I just thought about it a lot, and so what can happen with the calibration curve whether it’s a breath testing instrument, or a instrument that’s used in our lab for measuring concentrations of drugs, the calibration curve can shift either up or down, left or right, or as you said, it could potentially pivot around an axis point or any axis point. But once the calibration is set, it’s fixed and doesn’t changed. So for the Intoxilyzer 8000C, right, you – the manufacturer, as you said – described, would run a series of standards and then run the program to get the calibration curve and then that is fixed. That doesn’t change. That doesn’t drift. But what does drift is the response of the instrument and that’s what you’re measuring when you’re doing the calibration check, the stand-alone calibration check, is you’re checking to see whether or not the instrument response has changed over time because again, there’s never been anything shown that the calibration shift – the calibration shifts in any kind of direction, but that the response of the instrument could potentially shift and so the issue is that you’re checking it with that standard. And again, in addition to that, during the diagnostic checks the instrument runs what’s called an I-T-P, and internal test procedure which is an electronic internal check of the calibration of the instrument. So, you have an external standard which has been certified by the Centre of Forensic Sciences, as being accurate for use for checking the calibration of the instrument, plus you have the instrument’s own internal
electronic check of the calibration. And if it’s outside of the acceptable range, then again, it fails and won’t allow testing to proceed. Does that cover? Q. I think so. A. Okay. Q. Um, I can tell you that – do you know this book Wigmore on Alcohol? A. Yes.
Q. All right. Mr. Wigmore’s a very respected scientist for many years at the Centre of Forensic Sciences? A. Yes, he was there when I... Q. And elsewhere. A. ...first started, yes. Q. All right. A. I’ve even written papers with him. Q. The – I just want to show you the inscription of the book at the front. A. Okay. Q. You can read it if you want. A. “To Steven Biss, everything you wanted to know about alcohol, Wigmore on Alcohol, courtroom toxicology, Jim Wigmore, C-L-A Conference, Toronto October 30th, 2015.” Q. Yes. But as far as you know, there’s nothing in this book of an empirical study talking about what you described. A. Correct. There’s never been any presentation presented nationally or internationally with respect to the calibration of either the Intoxilyzer 5000C or 8000C shifting. Q. Right. Now, you just talked about.... A. With time. It’s a hypothetical but it’s never, ever been....
See also blog entry "Reliability - Calibration is a Curve" for more explanation.
See also "Why Does the ATC Say Single Point Control Tests are Good Enough?" for an earlier portion of this cross-examination.
#calibration #linearity #crossex
#drift #precision #Hodgson #accuracy #calibrationcurve #controltest