There are six reports that NHTSA has referred to as “extensive scientific research” regarding roadside field sobriety testing. These reports were submitted between 1977 and 1998. Since 1998,the NHTSA student manuals have not referred to any additional scientific reports or studies regarding the administration of SFSTs.
These six reports, the only scientific reports that NHTSA has cited, were never peer-reviewed by the scientific community. Peer-review is “a process by which a scholarly work is checked by a group of experts in the same field to make sure it meets the necessary standards before it is published or accepted. (Merriam-Webster Dictionary) The lack of peer-review would explain why each validation study contains a number of flaws.
The 1977 SCRI Study
In 1977, the Southern California Research Institute test subjects were instructed not to eat or use any stimulants for four (4) hours prior to testing and to arrive alcohol free. However, these instructions were violated by a number of subjects. Some even arrived with positive BACs. Even worse, these violations were ignored by the researchers. Additionally, the comparison for a relationship to driving impairment was not made with actual driving. The researchers used visual tracking and other non-driving based tests for comparison.
The 1981 SCRI Study
In this study, the researchers examined drinking volunteers, not persons who were actually stopped for suspected DUI. Participants in the study were asked not to eat, drink or use drugs prior to their arrival at the testing facility. Again, not all of the participants complied with those instructions. Three subjects arrived with BACs of 0.05g/dL or more and no chemical testing was done to determine whether participants were under the influence of drugs.
Additionally, one (1) of the subjects had a zero (0.00) BAC, but was classified as being over 0.10g/dL by the police. Four (4) other subjects who had a zero (0.00) BAC were classified as being over 0.10g/dL by the research staff. Lastly, the researchers simply relied on the truthfulness of the subjects to self-report any illegal drug use.
The 1983 SFST Field Evaluation
In the 1983 SFST Field Evaluation Study, “it was not possible for researchers to routinely accompany the patrols and supervise the actual data collection. Therefore, no statements can be made as to how close the requested data collection procedures were followed.” Also, the data relating to the frequency of the monitors that were riding with the police were not mentioned. This leads one to believe that the frequencies of the monitors were extremely low.
Even though the focus of the study was to determine the accuracy and reliability of the three test battery, the presence and use of portable breath testing devices (PBTs) tainted the reliability of the study. With the exception of the North Carolina State Police, all of the other police agencies (Virginia, Maryland and Washington DC) equipped their officers with PBTs. The presence and use of the PBTs would appear to be a crutch for the officers when they had to make the decision whether or not to arrest a suspect.
Furthermore, if an officer stopped and tested someone with SFSTs and determined that the driver was not impaired, the officer was still required to use the PBT. If that PBT showed a .150 BAC, then that officer would have to arrest the driver. The converse is also true. Assuming the driver failed all three tests, yet blew a .002 BAC, the individual would still have to be arrested for the sake of obtaining research results. The study even acknowledges this problem: “There are a number of problems in using these data. We do not know how those given a PBT differ from or are representative of the rest. Perhaps most significant of all, except for North Carolina, all agencies had PBT’s available, and in the great majority of the cases, PBT data were available to the officer for a driver before he was arrested. Thus, most of the arrest decisions were based on PBT data, rather than just the test battery data.”
Of the missed classifications, 16% of the drivers were incorrectly classified as having a BAC of 0.10g/dL or more. (197 drivers were falsely arrested.)
The Colorado Study
The 1995 Colorado study recorded but did not account for many variables such as: weather conditions, temperature, type of roadway, roadway conditions, conditions of the roadway surface, reasons for the stop, open containers in the vehicle, whether the suspect’s clothing was adequate for weather conditions, suspect’s footwear, whether or not there was an observer present, whether or not the police officer had a PBT, and whether the study was done around recreational events, holidays, taverns, bars or restaurants in the area. (Note that the 1998 San Diego study collected the above variable data, but did not release the raw data for analysis.)
Additionally, the Colorado study made a correct or incorrect arrest decision based on a statutory limit of .05g/dL. That is well below the illegal standard today, of .08g/dL.
The San Diego Study
In the 1998 report regarding SFST validation, the researchers used only seven officers from one police agency in San Diego. The officers were advised not to use any other tests besides the SFST battery for their initial determination of estimating a BAC. However, all officers were actually equipped with a NHTSA-approved PBT. There was also very little supervision. Project staff monitors “periodically” rode with participating officers. In this study, there were 24 false positive cases. (24 people were falsely arrested.)
Additionally, the San Diego study wanted “only trained and experienced test administrators” to participate in the study. In essence, the study’s findings only apply to officers with a great deal of experience in DUI detection, apprehension, and standardized field sobriety testing. The study does not apply to inexperienced or partially-experienced officers.
The Florida Study
In this 1998 study, 379 drivers were stopped and only 256 were evaluated for the report. Monitors were present only 63% of the time. There is no breakdown as to the percentage of the 256 investigations, which went without observation.
Read more in this series:
Say No No to the Po Po! (Part 1 of 6)
NHTSA Manuals (Part 2 of 6)
The “Reports” or “Studies” (Part 3 of 6)
What Does It All Mean? (Part 4 of 6)
The Ohio Legislature’s Response (Part 5 of 6)
Conclusion (Part 6 of 6)