How Should Breath Alcohol Measurements Be Reported? An Interview with A.W. Jones
A case for reporting the average result with a deduction for uncertainty.
In Minnesota, breath alcohol measurements are reported by taking the lower of the two measurements and truncating the third digit. So a 0.083 becomes a 0.08 g/210L. At first glance, this might seem to benefit the person taking a breath test.
The problem with this practice is that there is no scientific rationale for it, and it arbitrarily benefits some people more than others based on chance.
For example, a person who blows a 0.079 and a 0.099 gets a final value of 0.07 g/210L. That person benefits greatly by taking the lower sample and truncating the third digit. They will not have their license automatically rescinded based on the results. They get a deduction of 0.019 from their average breath result of 0.089 g/210L.
In another example, a person blows a 0.080 and a 0.080. Taking the lower of the two samples gives 0.08 g/210L as the final result. In this case, the person will automatically have their license suspended despite having a lower average result (0.080 g/210L) than the first example!