Cherry Picking with the BCA

Posted On June 22, 2015

For a driver with a breath alcohol concentration (BrAC) that straddles one of the statutory cutoffs (0.04, 0.08, 0.16), measurement uncertainty is reasonable doubt -- reasonable doubt that the BCA refuses to report.

Consider this scenario adapted from the National Academy of Sciences' Strengthening Forensic Science in the United States: A Path Forward (2009):

A driver gives three breath samples. The device calculates three results: 0.08, 0.09, and 0.10. The different test results arise from a number of factors: the internal components of the test device, calibration, and environmental characteristics, to name a few. Without accounting for these factors--by presenting the average of the three samples, 0.09, as an actual BrAC--the prosecution is able to argue that there is no reasonable doubt.

That is not science.

It is scientifically impossible to achieve 100% accurate results. In this scenario, the variation between the three breath sample results clearly indicates some uncertainty of measurement. Because of this scientific reality, test results must be accompanied by the area around those results, known as a confidence interval, in which there is a high probability that the actual BrAC lies.

How is this confidence interval calculated? Here, the average is 0.09 (the sum of the three values, 0.27, divided by 3) and the standard deviation is 0.01 (the three results deviate from one another by 0.01). This means that a two-standard-deviation confidence interval (0.07, 0.11) has a high probability of containing the person's true BrAC. Here, for the per se crime of driving with a BAC at or above 0.08, science establishes reasonable doubt.

But â?¦ not in Minnesota. Here, the BCA refuses to report measurement uncertainty in breath tests, even though other state courts have held that:

Given the requirements of due process, the discovery rules and [rules of evidence], therefore, the State must provide Defendants with a confidence interval for each Defendant's breath-alcohol measurement. Absent this information, a defendant's breath-alcohol measurement will be suppressed.

In other words, cherry picking is not allowed when it comes to science.