Lies, Damn lies, and Statistics a 1995 Colorado study commissioned and funded by the NHTSA and Colorado DOT acknowledges that 12 people out of 234 were erroneously arrested for drunk driving in Colorado when their breath tests showed levels not criminally actionable. The study also was forced to admit that they released 21 subjects later proven over what was then the legal limit in Colorado, .05 grams ethanol per 210 L breath. This apparently doesn't take into account 22 refusals, for which data is impossible to calculate. 



Does Colorado NHTSA study actually invalidate rather than validate SFSTS? 

by DAC-LAW PLC


​In a blog here DAC-LAW PLC questioned a twenty year old San Diego study commissioned by the federal transportation agency to validate Standardized Field Sobriety Tests, or, SFSTs. An even earlier study is also often cited by SFST proponents who conclude that the SFSTs are a reliable predictor of BrACs higher than .05 grams ethanol per 210 L breath, which will become important as special interests motivated by grief politics lobby to lower national BrAC to that figure.

The study, authored again by Marcelline Burns, of Southern California Research Institute, and Ellen Anderson, then a Deputy with the Pitkin County Sheriff’s Office in Aspen, Colorado, and funded by the National Highway Traffic Safety Administration, examined 234 individual suspected drunk driving cases, and concluded that SFSTs are a reliable predictor for BrACs over .05 grams per 210 L breath.

The study acknowledges that in 12 out of 163 arrests, law enforcement predicted BrACs over .05 when in fact they were lower. Moreover, the study admits that in 21 cases, officers predicted that BrAC values would be lower than .05 when in fact breath tests showed they were higher. Thus, the study’s authors concluded that SFSTs cannot reliably predict BrACs for 14 out of 100 test subjects.

The salient fact here is that the study’s authors designed the study and chose when the officers would predict BrACs. Since the study was intended to determine the validity of SFSTs, the study’s authors designed it to predict BrACs after the administration of those tests.

However, the study acknowledges that from the time officers began administering the SFSTs, another 38 “correctly released,” or, incorrectly suspected of drunk driving to begin. That would make OWI/DUI investigations faulty until the third phase of detection.

In other words, 38 out of 234 test subjects were asked to step out of the vehicle and erroneously identified as intoxicated prior to SFST test administration. Their breath tests proved otherwise.

However, 22 test participants refused the breath test and were still arrested and those numbers do not appear to be included in the calculations of false positives, so the data is skewed. Of course, only an expert witness retained to examine SFST test results and trained in these areas can testify to them in court.

If you or a loved one has been arrested for OWI (DUI) in the State of Iowa, contact DAC-LAW PLC at 319-389-4276 for a free initial consultation or visit https://www.cedarrapidsduilawyer.lawyer.