Precluding Use of Standardized Field Sobriety Tests in non- per se
Prosecutions and When Not Performed According to NHTSA Standards.
Ted Vosk*, Attorney at Law/Consultant

Evidence concerning performance of Standardized Field Sobriety Tests (SFSTs) posses some unique challenges for those engaged in the practice of DUI defense. Unlike breath tests or even Drug Recognition Expert (DRE) evidence, SFST’s are not yet viewed by courts in this state as being ground in systematic research or recognized standards.1 Ironically, while this would demonstrate a weakness and perhaps even a basis for suppression of other similar evidence, it actually makes SFST’s more difficult for a defense attorney to deal with. This results for two reasons: First, Washington courts have determined that performance on at least the balance and agility components of the SFSTs can be understood and independently assessed by a lay jury based on common experience so that it is not necessary to lay any type of particularized foundation for the introduction of such evidence; and second, an arresting officer may render a “quasi-expert” opinion of impairment in reliance upon an individual’s performance of the SFSTs based on nothing more than his personal experience utilizing them.2

Two of the most common approaches to dealing with SFST evidence are to attack the adequacy of the NHTSA sponsored validation studies underlying the development of the SFSTs3 and/or demonstrate the arresting officer’s failure to adhere to the NHTSA standards for administering the SFSTs.4 While ultimately a successful attack on the SFSTs will likely end up making these arguments, alone neither addresses the difficulties alluded to above. The first fails because if performance on SFSTs is something a lay individual can understand and independently assess on his own, whether or not the validation studies are adequate is irrelevant. The second misses the mark in part for the same reason and because if Washington does not recognize NHTSA standards but instead relies upon an officer’s experience for the interpretation he places on SFST performance, then NHTSA’a standards are also irrelevant. To succeed, ultimately two things should be established preliminarily: (1) performance of SFSTs does not lie in the realm of lay understanding; and (2) for the performance of SFSTs to have any demonstrable meaning the administration of SFSTs must not deviate from uniform standards.

Both are established, paradoxically enough, by educating the court on the strengths of the validation studies. While counter intuitive, establishing that the SFSTs were developed using scientific methodology should ultimately limit their application. As lamented by the International Association for the Chiefs of Police (IACP), however, unfortunately many courts do not realize that the SFSTs constitute a technology with a scientific basis.5 Nonetheless, given their history and development it is clear that the SFSTs are scientific in nature.6 This fact is explicitly recognized by Washington’s Criminal Justice Training Commission (CJTC).7

While the SFSTs have been considered as evidence for years, they have never been deemed scientific. By establishing that they are scientific in nature, their consideration involves new scientific principles before the court for the first time. Accordingly, their use as evidence of guilt must pass the Frye test of general acceptability before being admitted by the court.8 While it is tempting to attack the complete validity of the SFSTs on Frye grounds, winning such an attack would simply put us back to square one with the difficulties discussed above still present. Instead, our goal is to develop the criteria of general acceptability but to show that such acceptability is strictly limited.9 In doing this, one thing we will depend on is that whoever comprises the relevant scientific community for purposes of determining the interpretation of the SFSTs, it certainly includes the researchers who developed the SFSTs for NHTSA.10

The prosecution ordinarily seeks to introduce the SFSTs for one of two purposes: (1) as direct evidence of impairment; and/or (2) as indirect of evidence of impairment through the estimate of an individual’s breath alcohol content (BAC).11 This is natural because “[m]any individuals, including some judges, believe that the purpose of a field sobriety test is to measure driving impairment.”12 Nonetheless, it “is based on the incorrect assumption that field sobriety tests are designed to measure driving impairment.”13 Although measuring such impairment was one of the initial goals in developing the SFSTs, an early Department of Transportation study found that the 3-test battery was a poor tool for doing so.14 In fact, the developers of NHTSA’s SFST battery have frankly admitted that it is “unlikely that complex human performance, such as required to safely drive an automobile, can be measured at roadside.”15 Accordingly, despite attempts to the contrary, no link between the performance of SFSTs and driving impairment could be established by the research on SFSTs.16

This conclusion accomplishes two things. First, in a prosecution for DUI on the “under the influence/affected by” prong,17 it makes any use of the SFSTs as a direct indication of guilt irrelevant because they are unrelated to any element of the crime charged. This is a result of the fact that the SFSTs are incapable of determining whether an individual is under the influence of or affected by alcohol. While standing alone this result is significant, its reach is broader. In fact, it does away with one of the primary difficulties mentioned above when confronted by evidence of SFSTs: it removes them from the realm of lay observation.

To understand why, one must realize that the only thing common experience could inform a lay person of with respect to performance on SFSTs is impairment itself.18 After all, a lay person has no training or experience in administering and scoring the performance of SFSTs for purposes of establishing the probability that an individual is over a particular BAC. Since the SFSTs cannot detect impairment then, they do not involve observations easily understood by a jury.19 Although there are sure to be individuals who will claim that they could interpret these actions based on their personal experiences and determine whether an individual was impaired as a result of them, direct scientific research indicates that they are wrong. Any belief to the contrary is simply the unsupported bias of the individual possessing it.20

Based on these considerations, NHTSA determined that “the only appropriate criterion measure to assess the accuracy of SFSTs is BAC.”21 “As a consequence, they pursued the development of tests that would provide statistically valid and reliable indications of a driver’s BAC, rather than indications of driving impairment.”22 While a modicum of success was achieved, the uncertainty involved in estimates of BAC was found to be at least a quarter of the .08 limit itself. Quite simply, there are no behavioral cues which differentiate infallibly in a ± .02% BAC margin.23 Worse still, the results of the 1981 study showed the “[O]fficer estimates of the BACs of people they tested differed by .03 on the average from the actual BAC.”24 This has not changed in the intervening years.25

Given these uncertainties, NHTSA explicitly requires officer’s be trained that the SFSTs are “not to be used to estimate specific BAC level.”26 What SFST performance has been correlated with is the likelihood that an individual will be above or below the per se limit, not the ability to predict particular BAC values. Thus, to allow SFSTs as evidence of a particular BAC in and of itself misrepresents what they have been validated for. Based on this then, it can be strongly argued that use of SFSTs as indirect evidence of impairment through the estimate of an individual’s breath alcohol content (BAC) should be precluded. Further, if the argument for strict adherence to NHTSA guidelines is accepted (see below) NHTSA’s requirement that the SFSTs are “not to be used to estimate specific BAC level[‘s]” should be binding.

On the other hand, the prosecution is likely to respond that it is not trying to demonstrate a specific and particular BAC but simply the likelihood that a defendant was in excess of a .08. Remember, though, that the .08 is being advanced as a starting point for an expert opinion that by itself this BAC is evidence of impairment. It will not be disputed here that stating an opinion as to the likely impairment of an individual based upon his BAC as determined by an accurate method, such as a breath or blood test, is reasonable. The problem is that the SFSTs don’t give us an accurate measure of a particular BAC to begin with.27 In fact, the average error associated with estimates of BAC based on the SFSTs (± .024)28 is a full order of magnitude greater than that associated with breath and blood alcohol testing.29

Taking just the average error into account then, it is easily seen that even where an individual has “failed” the SFSTs any estimate of a BAC greater than a .056 is difficult to support. Even this figure is questionable when one considers that in the 1998 study officers overestimated the BAC of individuals who were between a .0 and .08 approximately 65% of the time30 with the largest overestimate being a .077,31 nearly equal to the per se limit itself. Given these facts, it now becomes quite significant that “[I]mpairment varies widely among individuals with the same BAC level.”32 This is because the opinion of impairment which is already speculative, even with a precise measure of BAC, is now being based upon a BAC which is nothing more than sheer conjecture. Pyramiding inference upon inference in this way by attempting to apply the SFSTs in a manner they were neither designed nor meant to be used for is clearly misleading, confusing and far more prejudicial than probative. Thus, again we find that one can argue that use of SFSTs as indirect of evidence of impairment through the estimate of an individual’s breath alcohol content (BAC) should be precluded.

Now that we’ve assembled a stable of arguments removing the SFSTs from the arena of lay understanding and prohibiting their use for purposes of a prosecution on the “under the influence/affected by” prong, we can turn our attention to the issue of mandatory standards for the administration and “scoring” of SFSTs.

The SFSTs were developed pursuant to research commissioned by NHTSA beginning in the 1970s.33 Researchers noticed early on that “[T]here are wide differences between officers in using tests to assess a driver’s state of intoxication, and they may exist within a department as well as between agencies and locales.”34 The problem with this is that such variation negates any reliability the SFSTs may have.35 To be reliable, standardization of the tests and observation procedures “is highly important.”36 In recognition of this, subsequent studies sought to establish and validate such standardized tests and observation procedures. This process resulted in the 3-test battery we have today consisting of the walk and turn, one leg stand and horizontal gaze nystagmus (HGN) tests. SFSTs consist of this battery of tests performed in the prescribed standardized manner. Emphasizing the importance of such standardization to the usefulness of the SFSTs, the researchers responsible for the 1995 report indicated that:

The validity of the SFSTs hinges on standardized administration and scoring. To the extent that officers’ instructions and demonstrations, or their interpretations of observations, differ from those established by research, it diminishes the meaning which can be attached to drivers’ test performance.37

As a result, “[I]f any one of the standardized field sobriety test elements is changed, the validity is compromised.”38 The SFSTs are valid “only when the tests are administered in the prescribed, standardized manner…[and]…the standardized criteria are employed to interpret that performance.”39 When this is not done, they are deprived of all predictive power. Accordingly, when the administration of SFSTs does not conform to the established standards, an individual’s performance on them is irrelevant because little, if any, reliable information can be derived from them. Standing alone this offers strong support for the proposition that, in order for evidence concerning an individual’s performance of SFSTs to be admissible, they must have been administered strictly in accordance with the prescribed standards. There is more, however.

In 1989, the International Association of Chiefs of Police (IACP) passed an official resolution adopting NHTSA SFST standards.40 Subsequently, in 1992, the IACP teamed with NHTSA in formulating and adopting uniform minimum training standards for use in training all law enforcement officers the administration and scoring of the SFSTs.41 As a result, NHTSA’s SFSTs are presently employed in all 50 states.42 They “have become the standard pre-arrest procedures for evaluating DWI in most law enforcement agencies.”43 In order to be certified under these standards, an officer is required to pass a course of NHTSA-approved SFST instruction.44

The CJTC was created by the Washington State Legislature “to provide programs and standards for the training of criminal justice personnel.”45 All law enforcement officers in the State of Washington must complete basic law enforcement training which complies with standards adopted pursuant to these powers.46 If an officer fails to comply with this requirement, he must be discharged as a law enforcement officer.47

The State of Washington has determined that all of its officers must be certified in the NHTSA SFSTs.48 As a result, one of the requirements of the CJTC curriculum is that all officers must attend a 3 day class, and pass a written examination, on SFSTs.49 This is the same program established through the cooperation of the IACP and NHTSA and recognized throughout the United States.50 As a result, if the test is not passed, an officer will be suspended from the academy.51 Thus, before an individual can become a law enforcement officer in Washington, he must be trained in the SFSTs. This illustrates the fact that, in addition to their development and promulgation by NHTSA and their recognition in the law enforcement community in general, the CJTC, under the authority granted to it by RCW 43.101.080, has established NHTSA’s SFSTs as the standard for roadside sobriety tests in the State of Washington as well.52 This rebuts the court’s conclusion in Staudenmaier that “no Washington case law, statute, or administrative code adopts [the NHTSA SFST] standards.”53 Accordingly, the proposition that in order for evidence concerning SFSTs to be admissible they must have been administered strictly in accordance with NHTSA standards is supported not only by the nature of the SFSTs themselves but by the fact that the State of Washington, through the CJTC, has explicitly recognized them as the standards to be utilized by law enforcement officers.

In this brief overview we have seen that by establishing the scientific nature and development of, as well as the standards governing, the SFSTs, certain difficulties surrounding their use as evidence can be ameliorated. This results because the science and standards governing their use demonstrate that: (1) the performance of SFSTs does not lie in the realm of lay understanding; (2) SFSTs cannot be used do determine impairment; and (3) for the performance of SFSTs to have any demonstrable meaning they must be administered according to uniform standards.54 This article has laid the ground work for these arguments. It is hoped that through their application, the courts of Washington will be able to develop a more complete understanding of the SFSTs and their use as evidence.

1 See e.g., City of College Place v. Staudenmaier, 110 Wn.App. 841, 848, 43 P.3d 43 (2002)( “[N]o Washington case law, statute, or administrative code adopts [recognized SFST] standards.”). But see also, State v. Baity, 140 Wn.2d 1, 17, 991 P.2d 1151 (2000) (“HGN testing is scientific in nature [and] the forensic application of HGN to drug intoxication in the DRE context satisfies Frye.” [Emphasis added]).

2 City of Seattle v. Heatley, 70 Wn.App. 573, 579-82, 854 P.2d 658 (1993), review denied, 123 Wn.2d 1011, 869 P.2d 1085 (1994).

3 MARCELLINE BURNS ET AL., A FLORIDA VALIDATION STUDY OF THE STANDARDIZED FIELD SOBRIETY TEST (SFST) BATTERY (1999); MARCELLINE BURNS ET AL., VALIDATION OF THE STANDARDIZED FIELD SOBRIETY TEST BATTERY AT BACs BELOW 0.10. U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DOT-HS-808-839 (1998); MARCELLINE BURNS ET AL., A COLORADO VALIDATION STUDY OF THE STANDARDIZED FIELD SOBRIETY TEST (SFST) BATTERY (1995); THEODORE ANDERSON ET AL., FIELD EVALUATION OF A BEHAVIORAL TEST BATTERY FOR DWI, U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DOT-HS-806-475 (1983); MARCELLINE BURNS ET AL., DEVELOPMENT AND FIELD TEST OF PSYCHOPHYSICAL TESTS FOR DWI ARREST, U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DOT-HS-8-01970 (1981); MARCELLINE BURNS ET AL., PSYCHOPHYSICAL TESTS FOR DWI ARREST, U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DOT-HS-5-01242 (1977); MARCELLINE BURNS ET AL., PSYCHOPHYSICAL TESTS FOR DWI ARREST, U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DOT-HS-5-01242, p.1 (1977).

4 DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DWI DETECTION AND STANDARDIZED FIELD SOBRIETY TESTING, INSTRCTOR MANUAL, HS 178 R9/04 (2004); DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DWI DETECTION AND STANDARDIZED FIELD SOBRIETY TESTING, PARTICIPANT MANUAL, HS 178 R9/04 (2004).

5 IACP, TRAFFIC SAFETY FOR LAW ENFORCEMENT, p.21 (2003).

6 The nearly identical development of the SFSTs with that of the DRE program renders Baity, supra note 1, as support for this proposition.

7 SFST’s are “scientifically tested methods to detect, screen, and process DUI drivers.” CJTC website < http://www.cjtc.state.wa.us/classes/0156.htm> (last modified August 23, 2005).

8 Cf. Baity supra note 1, at 10-11.

9 For instance, an odometer is based upon generally accepted scientific principles and has a generally accepted use: to record the distance traveled. It is not a generally accepted device for measuring an individual’s speed at a given instant, however. For that there is a different device: a speedometer.

10 Cf. Baity, supra note 1, at 17.

11 SFSTs cannot be used to establish a motorist’s BAC for purposes of a per se violation because such an offense requires the BAC to be “shown by analysis of the person's breath or blood.” RCW 46.61.502.

12 MARCELLINE BURNS ET AL., VALIDATION OF THE STANDARDIZED FIELD SOBRIETY TEST BATTERY AT BACs BELOW 0.10. U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DOT-HS-808-839, p.27 (1998).

13 Id. at 28.

14 MARCELLINE BURNS ET AL., PSYCHOPHYSICAL TESTS FOR DWI ARREST, U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DOT-HS-5-01242, p.1 (1977). K.J. SNAPPER ET AL., DEPARTMENT OF TRANSPORTATION, AN ASSESSMENT OF BEHAVIORAL TESTS TO DETECT IMPAIRED DRIVERS, FINAL REPORT 4-2 (1981).

15 See supra note 12, p.28.

16 See supra note 12, p.28.

17 RCW 46.61.502(b).

18 Cf. Heatley, supra note 2, at 580-2.

19 Instead such observations are evidence consisting of what would otherwise be scientifically determined clues within the SFST battery.

20 In one study, researchers had 21 individuals who had consumed no alcohol or intoxicating drugs perform the SFSTs (not including HGN) and other tasks. Despite the fact that all of these individuals were completely sober, police officers perceived 46% of the subjects performing the tests as drunk and worthy of arrest. SPURGEON COLE ET AL., FIELD SOBRIETY TESTS: ARE THEY DESIGNED FOR FAILURE? 79 PERCEP. & MOTOR SKILLS, 99 (1994).

21 See supra note 12, p.10.

22 See supra note 12, p.28.

23 MARCELLINE BURNS ET AL., PSYCHOPHYSICAL TESTS FOR DWI ARREST, U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DOT-HS-5-01242, p.27 (1977).

24 MARCELLINE BURNS ET AL., DEVELOPMENT AND FIELD TEST OF PSYCHOPHYSICAL TESTS FOR DWI ARREST, U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DOT-HS-8-01970, p.i (1981).

25 In the 1998 study, instead of having officers simply make arrest decisions based on SFST performance as was the practice in the earlier validations studies, NHTSA researchers also had officers estimate drivers BAC. The average error was ± .024 with the largest over estimation giving a BAC level of .077 greater than the true BAC of the driver in question. Officers overestimated driver BAC levels in approximately 45% of these cases. Thus officers assigned an erroneously high BAC level to 130 of the drivers that were stopped during the study. What is more disturbing, is the rate of overestimated BACs in individuals who had BAC’s below the legal limit. For individuals who had BACs between 0 and .04, officers overestimated their BAC 76% of the time while for those with BACs between .04 and .08 they did so 64% of the time.

26 DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DWI DETECTION AND STANDARDIZED FIELD SOBRIETY TESTING, INSTRCTOR MANUAL, HS 178 R9/04, P. VIII-25, (2004).

27 Compare the average error of ± .024 associated with the SFSTs with the that allowed for blood (± .0024 at a .08 as per WAC 448-14-010) and for breath (± .004 at a .08 as per WASHINGTON STATE PATROL BREATH TEST SECTION, BREATH TEST PROGRAM POLICY AND PROCEDURE MANUAL, 27 (10/24/2004)).

28 See supra note 25.

29 See supra note 27.

30 See supra note 12.

31 See supra note 12.

32 U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DWI DETECTION AND STANDARDIZED FIELD SOBRIETY TESTING, STUDENT MANUAL, §VII (2000 ).

33 NHTSA, DEVELOPMENT OF A STANDARDIZED FIELD SOBRIETY TEST (SFST) TRAINING MANAGEMENT SYSTEM, DOT HS 809 400 (Nov. 2001) <http://www.nhtsa.dot.gov/people/injury/alcohol/SFST/>

34 See supra note 23, p.59.

35 Id.

36 Id.

37 MARCELLINE BURNS ET AL., A COLORADO VALIDATION STUDY OF THE STANDARDIZED FIELD SOBRIETY TEST (SFST) BATTERY, §IV-C (1995).

38 DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DWI DETECTION AND STANDARDIZED FIELD SOBRIETY TESTING, PARTICIPANT MANUAL, HS 178 R9/04, VIII-19 (2004).

39 Id.

40 IACP website (visited Sept. 20, 2004)

41 See supra note 33 and U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DWI DETECTION AND STANDARDIZED FIELD SOBRIETY TESTING, INSTRCTOR MANUAL, App. B, p.11 (2002 ).

42 See supra note 12, p.3.

43 See supra note 12, p.3.

44 See supra note 33.

45 RCW 43.101.020.

46 RCW 43.101.200 and WAC 139-05-200.

47 Id.

48 See supra note 7.

49 WASHINGTON STATE CRIMINAL JUSTICE TRAINING COMMISSION & SPOKANE POLICE TRAINING CENTER, BASIC LAW ENFORCEMENT ACADEMY RULES AND REGULATIONS, p.21 (2004). See also supra note 7.

50 CJTC website (last modified September 14, 2004) <http://www.cjtc.state.wa.us/classes/trng_definitions.htm> ; U.S. DEPARTMENT OF TRANSPORTATION, NATIONAL HIGHWAY TRAFFIC SAFETY ADMINISTRATION, DWI DETECTION AND STANDARDIZED FIELD SOBRIETY TESTING, INSTRCTOR MANUAL, App. B, (2002 ); See also supra note 7.

51 See supra note 49, p.22.

52 The CJTC’s reliance on NHTSA standards in the training of Washington law enforcement officers also acts as a party admission of the applicability of their requirements and pronouncements to any individual officer trained and commissioned pursuant to the requirements of the CJTC. United States v. Van Griffin, 874 F.2d 634 (9th Cir. 1989).

53 City of College Place v. Staudenmaier, 110 Wn.App. 841, 848, 43 P.3d 43 (2002).

54 This leaves open the question of how to handle the SFSTs if they have been administered in the appropriate manner and are being introduced in a per se prosecution for purposes of supporting breath or blood test result indicating a BAC in excess of the legal limit. Once again, establishment of the scientific nature of the tests may help to limit their use ultimately restricting their application to questions of probable cause determined prior to trial. Even if this were not the case, however, use of SFST evidence by the State in prosecutions where there is no breath or blood test is far more damaging to a defendant than when there is a valid test so that the focus of the challenges described herein still seems to be the correct one.

* Ted Vosk is a graduate of Harvard Law School with experience as both a prosecutor and criminal defense attorney. His current practice focuses primarily on appeals, consulting and legal writing. He can be contacted at (425) 753-6343.

---------------------------------------------------------------------------------------------------------