Attacking the Citadel of Bad Science


Courts are notorious for allowing bad science into evidence at trials even when the outcome can be so severe to the defendant.  It is even more difficult to persuade a Court that what has been accepted as “scientific” is wrong, and has been wrong for a long time, and the Courts must change.

In Minnesota, after the widely publicized explosion of the St. Paul Forensic labs for bad science, the legislature enacted a law requiring all labs to be certified.  The lawyers exposing all this were voted “Lawyers of the Year”.
Incredibly, our own Bureau of Criminal Apprehension (BCA) was not certified.

Now they are.

The parent agency that certifies the BCA requires that in any measurement instrument, the lab must calculate the real, scientific, uncertainty of measurement.

In the past, for breath test machines, the BCA assigned an arbitrary “Margin of Error”.  It was not a scientific calculation of actual uncertainty of measurement.  It was a fudge factor they used when their calibration solution (known sample solution) did not report the correct, expected result.  They allowed a margin of error of plus or minus .01.

So when we used to expect the control solution to read .10 to show an accurately functioning machine, it could read as high as .11 and as low as .09 and be considered working “accurately”.

Even more bizarre, the Courts did not allow the drivers whose results came from the same machine the same fudge factor.  There was no duty on the State to prove the result was at least .10 (now .08) within the plus or minus .01 margin of error the State enjoyed when calibrating the machine.  It was OK to use it to save their machine’s inability to calculate a known standard, but ignored when the driver sought the same benefit of the doubt.

Now, true Uncertainty of Measurement can be, and must be, and is being calculated.  It is not “Margin of Error”, yet many courts have already ruled it is the same thing.  Again, one was an invented, arbitrary fudge factor, and only used to help the State guess its machine was working accurately.  The other, Uncertainty of Measurement, is an actual scientific calculation of the limits of the machine’s accuracy.

With the uncertainty of measurement calculated, we can now at best say the result that we really want, the true BAC, is somewhere in this range, and we cannot say more!

This is hugely important now that people can go to prison based on the machine’s reported results.  There will be true ”reasonable doubt” for anyone testing at .08 all the way up to .09.  For those who receive greater penalties for testing .16, the current calculation of uncertainty of measurement would show reasonable doubt all the way up to .18 results.

This is not guesswork.   It is the true range - the true limitation - on the machine’s ability to report accurate BACs.  Trying to say that the BAC is any one of the many possible results within the uncertainty range is the guesswork.

Moreover, every machine has its own bias, just as every scale measures a little low or a little high.  Without ascertaining the bias on a particular machine - on the one used in this case - we never even get to use the uncertainty of measurement.  We don't yet know the bias to correct for, before applying the scientifically calculated uncertainty of measurement.

The lawyer protecting a driver must be familiar with this state of the art science, which science is required to be applied by the accrediting agency, and by the overwhelming agreement of the forensic community, despite the fact that Courts are stuck in the past, and do not know all this, do not understand all this, and conflate this with decades of non-scientific “Margin of Error” case law.

The lawyer must familiarize himself or herself with the requirements of the agency that accredits the lab, and the learned treatises and peer-reviewed work, and the accepted forensic science in breath testing, from many sources.

The Breath Testing Forensic Expert will testify regarding the foundational reliability of the breath test result. This testimony will pertain to the uncertainty of measurement values that apply to these breath test results, the metrological traceability of these test results, and the ultimate accuracy of the results. Hypothetical situations that may arise in the testimony include analogies to the effect of bias on the final reported values, the effect of random error of the final reported values, and the BCA’s practices of calculating or ignoring both bias and random error in reporting final values. 

The Breath Testing Forensic Expert’s testimony will rely, in part, on numerous expert treatises, which are listed here.