Should Conclusions Based Upon Field Sobriety Tests Be Admitted
February 14, 2008
As anyone who has ever tried an alcohol related operating offense can attest, in most instances the field sobriety test is an inescapable item of proof. In the average case, there is little chance of escaping the conclusions of the arresting officer that because the motorist was unable to complete several seemingly innocuous tasks that he or she was intoxicated.
Despite there frequent use, one can engage in a rather futile search to locate authority addressing the admissibility of these customary exercises. This is not to say that no authority exists. Last issue we discussed People v. Quinn, 153 Misc2d 139, 580 NYS2d 818, in our opinion Quinn, goes further than most in validating the use of such tests. Even so, the manner in which it does so is of little help. Quinn did not deal preciously with admission of the more commonly administered examinations. Instead, Quinn, sorted out detailed scientific questions as it held that conclusions based upon the DRE protocol were properly admitted. Even so, the Quinn court cited with approval People v. Jacquin, (1988) 71 NY2d 825, 527 NYS2d 728, to support the proposition that protocols covering psychophysical testing are admissible. We, however, do not read Jacquin as broadly. In Jacquin, the issue which was found to be properly before the Court was whether admission of the results of physical performance tests need be proceeded by the five-fold warning necessitated by the holding of the United States Supreme Court in Miranda v. Arizona. Holding they do not, the Court similarly found the contention that specific responses to various questions should be supressed were unpreserved.
Closer to the point is People v. Torrey, (1988, 3rd Dept) 144 AD2d 865, 534 NYS2d 807. In Torrey, the defendant claimed that the weight of the evidence was insufficient to warrant a conviction. While admissibility of the various field sobriety tests given the defendant was never placed directly at issue, the Third Department nonetheless took note that the defendant was observed swaying and staggering when attempting to touch the tip of his nose while standing with his feet together and his eyes closed. When asked to bend down and touch his toes and stand back up, he was again seen swaying back and forth by the Troopers. This evidence, found the court, when combined with his refusal to submit to a chemical test constituted sufficient credible evidence to support the verdict (supra., at 866).
The real question of course is just how reliable are the standard field sobriety tests? Are they accorded sufficient acclaim within the scientific community to warrant admission of the conclusions drawn therefrom?
Not many now practicing remember the days when finger tapping and maze drawing were given to determine whether a motorist was under the influence of alcohol. Just the same, such tests were, at one time and in some areas, frequently given. In 1977, however, Burns and Moskowitz attempted to catalogue the varying tests then given and determine whether or not establishment of a particularized battery of tests would lead to greater accuracy and ease of training. Carried out on behalf of the National Highway Traffic Safety Administration, (Burns, M., and Moskowitz H., Psychophysical Tests for DWI Arrest, National Highway Traffic Safety Administration Contract No. DOT-HS-5-01242) the authors identified sixteen tests which were currently utilized in drinking driver detection and elected a battery of six as the most reliable and easily administered. From this list, four, the walk and turn, the one leg stand, the finger to nose, and the horizontal gaze nystagmus, were ultimately selected.
Was the Burns/Moskowitz test battery the panacea which NHTSA sought? Let's look at their data. True, the test subjects reportedly exhibited an ability to identify those with a excess of .10% of alcohol in the bloodstream with remarkable exactness (84%). Even so, the report also notes a dramatic increase in the potential for error as the BAC approached .10%. According to the study, use of the standard battery of tests would have resulted in a false positive rate of 47%. It is perhaps this figure which is the most telling. Take the following example. Assume 100 test subjects, 99 of whom are sober (.00% BAC) and 1 who has a BAC of .12%. Further assume that in this study the police identify all 100 subjects as having a BAC of .10% or greater. The participants in this example have an accuracy rate in identifying legally intoxicated persons of 100%; that is, all of the legally intoxicated persons were identified. The failure rate, however, would be an unacceptable 99%. The Burns/Moskowitz figures, of course, are worse. Not only did the participants incorrectly identify 47% of the participants with a BAC of .10% or less as having a BAC of .10% or greater, they also failed to identify 16% of those who were legally intoxicated.
To explain the seeming fifty - fifty accuracy of their battery when used to eliminate those who were not actually intoxicated, Burns and Moskowitz turned to what they perceived to be the effect that alcohol can generally be expected to have upon the human body. Concluding that signs of intoxication tend to naturally appear at .08% the authors hypothesized that it was the artificially selected .10% threshold and not the accuracy of the battery that was at fault. This conclusion was supported, they felt, by test data that indicated that when lower limits were chosen as a threshold the error in selecting intoxicated drivers dramatically decreased, this time to 30%. Again, the authors have shrewdly manipulated the numbers. Assuming that .001% were chosen as a the cutoff, one can easily see how the reported ability to detect would dramatically increased tenfold.
In what may have been an effort to minimize the incidence of error noted by Burns and Moskowitz, in 1983, Anderson, Schweitz, and Snyder, on behalf of NHTSA, returned to the standardized battery. This time, in a report entitled Field Evaluation of a Behavioral Test Battery For DWI (National Highway Traffic Safety Administration, Washington, DC. Office of Driver and Pedestrian Research, [Publication DOT‑HS‑806‑475]), the researchers concluded that the battery is effective in determining whether the motorist's BAC is above or below .10%.
In 1990, Sussman, Needalman and Mengert, attempted to evaluate the reliability of the standard battery of tests when employed in a marine environment. In Experimental Evaluation of a Field Sobriety Test Battery in the Marine Environment (Transportation Systems Center, Cambridge Mass.) the Coast Guard Office of Research and Development found that in a study of 97 volunteers dosed with alcohol in a recreational boating setting, experienced enforcement agents correctly identified the subjects with a BAC of .10% or greater in only 82% of all cases. Further, estimates of the operators BAC showed an error rate of 30%.
Since the original Burns/Moskowitz study there has, of course, been a major improvement in Field Sobriety testing brought on by the introduction of highly portable electronic preliminary breath testing devices. Even so, unrelenting reliance upon such devices may prove a trap for the unwary. By and large the potential for error lurks in the fact that the most commonly employed detector, the Taguchi cell, is non-specific for ethanol. Although somewhat dated, this nonspecificity, particularly the ALERT model J3A was highlighted in a study on the effect of various substances other than ethyl alcohol upon it. In the study it was found that the ALERT J3A falsely indicated the presence of significant concentrations of ethanol when exposed to acetaldehyde, acetone, isopropanol, methanol and paraldehyde, all non‑ethanolic combustibles (Papple, The Effect of Non‑Ethanolic Volatiles on the Measurement of Blood Ethanol Concentrations with an ALERT Roadside Screening Device, 15 Can Soc Forensic Science Journal No. 3/4 ). Indeed, the usefulness of a PBT device seems highly dependant upon the ability of the user. In Measurement of Blood Alcohol Concentration in Michigan Drivers: A Review of the Literature, (Michigan University, Transportation Research Institute, Michigan Office of Highway Safety Planning) Olson concluded that although the introduction of preliminary breath testing devices can be very effective at reducing false positives, just how well it improves the identification of alcohol influenced offenders, particularly those at BAC's close to 0.10%, is not at all clear. The investigation found reported results ranging from no improvement to very great improvements.
Are there any motorists to whom the standard battery should simply not be administered?
Most definitely. In their 1981 study, Development and Field Test of Psychophysical Tests for DWI Arrest, (National Highway Traffic Safety Administration Contract No. DOT-HS-8-01970 ), Burns, Tharp and Moskowitz, speaking of the so-called walk and turn informed that certain individuals have difficulty with this test under sober conditions, including people over 65 years of age; people with back, leg or middle ear problems; people who are overweight by 50 or more pounds.
The foregoing is by no means intended to imply that field sobriety tests have no place in DWI arrests or should somehow not be used. Such would be ludicrous; field sobriety tests serve a valuable role in assisting the officer in forming his or her conclusions. The overarching difficulty encountered in the area of field sobriety testing is the dispositive evidentiary significance that courts routinely ascribe to what is actually a rough art indeed.
Certainly there is no evidentiary objection that can be raised when a police witness seeks to testify as to what he or she observed when say, a finger to nose test was administered. A witness is always free to testify as to what he or she smells, sees or hears (see, People v Babala, (1989, 3rd Dept) 154 AD2d 727, 547 NYS2d 68 [witness observations of a subject's gait]; People v Ritgers, (1990, 2nd Dept) 158 AD2d 628, 551 NYS2d 598 [speech and odor]; People v Fiacco, (1989, Albany City) 146 Misc2d 330, 549 NYS2d 901 (the appearance of subjects eyes]). However, it is also a fundamental principle of testimonial evidence that a witness should testify as to facts alone, leaving it to the trier of fact to draw the conclusions therefrom. Phrased differently, a witness should not give his or her opinion or conclusion upon the facts (People v Russell (1991, 2nd Dept) 165 AD2d 327, 332, 567 NYS2d 548, affd (1992) 79 NY2d 1024, 584
NYS2d 428; People v Raco (1979, 3rd Dept) 68 AD2d 258, 262, 416 NYS2d 849; see, Kulak v Nationwide Mut. Ins. Co. (1976) 40 NY2d 140, 148, 386 NYS2d 87).
Conclusions should only be permitted when the facts themselves are not capable of precise rendition, such as a fishy odor (Aronette Mfg. Co. v. Capitol Piece Dye Works (1959) 6 NY2d 465, 190 NYS2d 361; and see, People v. Mertz (1986) 68 NY2d 136, 506 NYS2d 290) or where the conclusion would help to clarify an issue calling for professional or technical knowledge, possessed by the expert and beyond the ken of the typical juror (De Long v County of Erie (1983) 60 NY2d 296, 307, 469 NYS2d 611; see also, People v Bennett (1992) 79 NY2d 464, 473, 583 NYS2d 825; People v Taylor (1990) 75 NY2d 277, 293, 552 NYS2d 883).
When the conclusion is received that a motorist failed the finger to nose examination or one leg stand, such an opinion will fall into neither exception. As to the first, the trier has heard the instructions which were given, it knows that the motorist was asked to touch the tip of his or her finger to the tip of the nose. The facts are capable of precise rendition and surely it would be an insult to the trier, be it the court or a jury, to say that whether a person touched his or her nose calls for professional or technical knowledge, possessed by the expert and beyond the ken of the typical juror (De Long v County of Erie ,supra).
The error, although frequent, is immeasurably compounded when, upon completion of all the tests, the officer is thereafter qualified as an expert witness and is permitted to set forth the opinion that he or she draws from the defendant's performance upon all of the tests, which is clearly different than the simple conclusion that the defendant was or appeared to be intoxicated (see, Burke v. East Tower Restaurant (2nd Dept. 1971) 37 AD2d 836, 326 NYS2d 32; see also, People v. Eastwood (1856) 14 NY 562). Arguably, a determination of intoxication, as opposed to impairment based upon the manner in which the motorist completed the tests may require expert interpretation. Even so, as a scientific experiment or test, the conclusions are inadmissible absent a showing that the reasonable accuracy and general scientific acceptance of the procedure are clearly established (People v. Leone, (1969) 25 NY2d 511, 307 NYS2d 430; see also, People v. Middleton, (1981) 54 NY2d 42, 444 NYS2d 581). As the DC Circuit Court of Appeals acknowledged in the seminal Frye v. United States, 293 F. 1013, 54 App.D.C. 46, the point when such acceptance has been achieved is oftentimes difficult to discern. Later cases interpreting Frye have seemingly called for an examination of the number of scientists willing to adopt the view offered by the proponent. For instance, in United States v Williams, (1978, 2nd Cir) 583 F2d 1194, for instance, the Second Circuit, in reference to a spectrographic voice analysis, declared that determination required a survey and categorization of the subjective views of a number of scientists, assuring thereby a reserve of experts available to testify. By no stretch of the imagination can two or three studies conducted by the proponent of the NHTSA test battery be deemed to fulfill the general acceptance requirement mandated by Frye and Leone.
Without question, the decision not to receive the conclusion will require a bit of courage. Admission of conclusions drawn by the administering officer following a recitation of the test has become routine, despite the fact that no on point authority permitting receipt of such conclusions. Even so, noted examples for reexamining the reliability of established or accepted tests exists (see, Giannelli, The Admissibility of Novel Scientific Evidence: Frye v. United States, a Half‑Century Later, 80 Colum L Rev 1197, 1224‑1225 [discussing belated discovery of inaccuracy of paraffin gunshot test]; Neufeld and Colman, When Science Takes the Witness Stand, 262 [No. 5] Scientific Am 46 [discussing belated discovery of inaccuracy of gunpowder detection test]). Indeed, when compared to the paraffin gunshot detection test, little backtracking at all is required in that a New York court is yet to hold that the conclusions drawn from the three test battery are generally accepted.
« Go Back