Ballard v. State

955 P.2d 931, 1998 Alas. App. LEXIS 17, 1998 WL 150774

• Court: Court of Appeals of Alaska
• Decided: April 3, 1998
• Docket: A-5679
• Status: Published

Opinion

MANNHEIMER, Judge.

This appeal requires us to decide whether the horizontal gaze nystagmus test, or “HGN” test, satisfies the standard for admissibility of scientific evidence established in Frye v. United States, 293 F. 1013 (D.C.Cir.1923), and reaffirmed as the governing standard in Alaska in Contreras v. State, 718 P.2d 129, 134-36 (Alaska 1986).

In the HGN test, the subject is asked to cover one eye and then use the remaining eye to track the lateral progress of an object (usually a pen) as the officer moves the object at eye-level across the subject’s field of vision. As the moving object travels toward the outside of the subject’s vision, the officer watches the subject’s eye for “nystagmus”— an involuntary jerking movement of the eyeball. If the person’s eyeball exhibits nystag-mus, and especially if the nystagmus occurs before the moving object has traveled 46 degrees from the center of the person’s vision, this is taken as an indication that the person is intoxicated. See the description of the HGN test contained in State v. Grier, 791 P.2d 627, 629 (Alaska App.1990).

As we explain in more detail below, we conclude that the horizontal gaze nystagmus test is premised on generally accepted scientific theory. Although there is genuine dispute as to whether a person’s precise level of intoxication or precise blood-alcohol level can be ascertained by his or her performance on the HGN test, there is no dispute that alcohol consumption causes observable nystag-mus. We conclude that, if a person demonstrates nystagmus during the HGN test, this is a reasonable indication that the person may have consumed alcohol and is potentially intoxicated. Therefore, with certain qualifications explained below, we hold that evidence of a person’s performance on the HGN test is admissible under the Fry e-Contreras rule.

How This Case Arose

In the early morning of March 1, 1994, Alaska State Trooper Lee Robert Oly was patrolling the Glenn Highway, accompanied by his supervisor, Trooper Oscar Siegfried. 1 Oly observed an oncoming Toyota pull off to the other side of the highway. Intending to check on the welfare of the Toyota’s driver, Oly turned his patrol car around and pulled up behind the stopped Toyota. Ballard, who was driving the Toyota, got out of his vehicle. Ballard had to steady himself against the Toyota, and he staggered as he walked toward Oly. When Ballard stopped walking, he continued to have difficulty maintaining his balance; he swayed from side to side. Oly smelled the odor of alcoholic beverages on Ballard’s breath, and he saw that Ballard appeared to have difficulty focusing his eyes (which were bloodshot). Based on these observations, Oly decided to administer field sobriety tests to Ballard.

Ballard performed several field sobriety tests; one of these was the horizontal gaze nystagmus test. After Ballard failed all of the field sobriety tests, Oly arrested him and charged him with driving while intoxicated.

Ballard’s Pre-Trial Objection to the HGN Evidence and the Testimony Presented at the Frye Hearing

Before trial, Ballard asked the district court to exclude all evidence of his performance on the HGN test, contending that the HGN test did not meet the Frye standard for admissibility. District Court Judge Natalie K. Finn conducted an extensive hearing into the scientific foundations of the HGN test, as well as its reliability and validity. 2

At this hearing, the State presented an instructional videotape produced by the National Highway Transportation Safety Administration (NHTSA), demonstrating its recommended procedure for administering a three-part HGN test. The State also called Trooper Tim Schoenberg, who demonstrated the three parts of the test in the courtroom: asking the subject to track a moving object, asking him to focus on the object at the limits of his peripheral vision, and asking him to focus on the object at an angle of 45 degrees from the center of his field of vision. Schoenberg not only described how the HGN test is to be administered, but he also described how the HGN test is taught to officers.

Trooper Oly, the officer who arrested Ballard, then testified concerning his own training in the three-part HGN test. Oly told the court that his training included administering the HGN test to nineteen test subjects. Like Schoenberg, Oly demonstrated the three portions of the HGN test on a volunteer in the courtroom.

(One of Ballard’s expert witnesses, Dr. Ronald Nowaczyk, later critiqued Oly’s performance. Nowaczyk noted that Oly had performed the three portions of the HGN test on one of the subject’s eyes, then had performed the three portions of the test on the subject’s other eye; the NHTSA recommends performing each portion of the test on each eye before moving to the next portion of the test. However, Dr. Nowaczyk did not suggest that this variation would affect the test results.)

The State’s third witness was Mareelline Burns, a research psychologist who was one of the scientists who conducted two seminal studies of the HGN test under the auspices of the National Highway Traffic Safety Administration. These two studies, conducted between 1975 and 1981, were admitted as exhibits at the Frye hearing.

Burns testified that, within the scientific community, it was well-accepted that alcohol intoxication causes nystagmus. Nystagmus is not, however, an unerring indicator of alcohol intoxication; it can also be caused by other drugs, by brain trauma, or by certain diseases. Moreover, approximately half of the population will exhibit an eyeball tremor when fixing their eyes on an object at the limit of their peripheral vision. For this reason, of the three parts of the HGN test conducted on Ballard, checking for nystag-mus at maximum deviation of the eyes was the least informative. Burns conceded that only “distinct nystagmus” during this portion of the test would be a meaningful indicator of intoxication.

Although Burns declared that the HGN test is a “highly sensitive” measure of whether a person is intoxicated, she also stated that the HGN test was never designed to be used alone, but rather in conjunction with other field sobriety tests such as the walk- and-turn test and the one-leg-stand test. Burns added, however, that among such a battery of tests, the HGN test has distinctive value because nystagmus is an involuntary response. An alcoholic can learn to control or compensate for the effects of intoxication on balance, physical agility, and verbal acuity, but it is all but impossible to disguise nystagmus.

Burns told the district court that, for purposes of assessing the reliability and validity of the HGN test, the relevant community of researchers and practitioners consisted of scientists concerned with alcohol and drug research, scientists concerned with traffic safety, and law enforcement professionals. Among this community, Burns testified, the battery of three field sobriety tests that includes walk-and-turn, one-leg-stand, and HGN is generally accepted as an accurate indicator of a person’s intoxication.

Ballard presented two expert witnesses at the hearing. The first was Dr. Craig Smith, a neuro-ophthalmologist. Smith conceded that the scientific community generally accepts the fact that alcohol intoxication affects people’s ability to move their eyes, and that intoxication will cause nystagmus. However, Smith pointed out that nystagmus is also caused by diseases of the brain, by damage to or disease of the eye muscles, and by various medications and drugs. According to Smith, nystagmus can also appear naturally in about two to five percent of the population.

Smith conceded that alcohol-produced nys-tagmus is observable by police officers in the field, but he strenuously questioned whether precise observations of nystagmus could be made outside of a controlled laboratory setting. In particular, Smith asserted that it is “impossible” for a police officer conducting a traffic stop to detect whether a driver’s eyes are smoothly pursuing a moving object. Smith pointed out that the lighting at the scene of a traffic stop can be inadequate to observe the subject’s eyes. He also pointed out that, if the driver is balancing unsteadily, the driver’s eyes may not move smoothly when tracking an object because the driver’s head is itself moving. Smith also pointed out the possibilities that the officer may not move the target object smoothly, or that the driver is myopic and therefore not able to see the object clearly. Smith told the court that he had ridden along with police officers some ten to twelve times, and he had observed that the HGN test was very difficult to perform during a roadside traffic stop.

Smith also testified that, even when nys-tagmus is visible, it is quite difficult to determine whether the angle of onset is less than 45 degrees unless one is armed with a protractor. Smith, like Burns, testified that the occurrence of nystagmus at maximum deviation is not a good indicator of intoxication because some 50 to 60 percent of the normal population displays such nystagmus. However, on cross-examination by the court, Smith admitted that the nystagmus manifested by half the population at the edges of their vision was a “tiny” nystagmus, readily distinguishable from the “distinct” nystag-mus that occurs in people who have blood-alcohol levels exceeding .10 percent.

Ballard’s second expert witness was Dr. Ronald Nowaczyk, a psychology professor who specializes in behavioral statistics and research. Nowaczyk explained the concepts of “reliability” and “validity” as they apply to scientific testing. (See footnote 2, supra.) Nowaczyk told the court that, having analyzed the results of the NHTSA studies on horizontal gaze nystagmus, he had concluded that the reliability of the HGN test (that is, the uniformity of HGN testing results) appeared to be between 60 and 66 percent — too low a percentage for the HGN test to be acceptable as a method of “clinical diagnostic testing” for determining “whether someone meets a certain level [of intoxication]”.

Turning to the validity of the HGN test (that is, the likelihood that the test accurately measures intoxication or non-intoxication), Nowaczyk told the court that he could not come to a statistical conclusion based on the data available to him. However, he believed that the HGN test “tend[s] to overestimate” people’s levels of blood alcohol. Nowaczyk referred the court to a field study of the HGN test which showed that significant numbers of drivers failed the HGN test even though their blood-alcohol levels were below .10 percent. According to Nowaczyk, this field study showed that 15 percent of drivers with blood-alcohol levels of .00 to .04 percent failed the HGN test. Among drivers with blood-alcohol levels of .05 to .09, 64 percent failed the HGN test. However, among drivers with blood-alcohol levels of .10 to .15, the percentage failing the HGN'test rose to 95. Under cross-examination, Nowaczyk admitted that the HGN test appeared to be quite accurate in confirming the intoxication of people whose blood-alcohol level is .10 percent or higher.

After considering this testimony and reviewing the exhibits, Judge Finn ruled that the horizontal gaze nystagmus test met the Frye standard for admission of scientific evidence. The judge concluded that the scientific community accepted the connection between the appearance of nystagmus and the consumption of alcoholic beverages. While Judge Finn acknowledged that the HGN test might not be reliable enough to satisfy the standards for tests employed to make medical diagnoses, she concluded that the HGN test was a sufficiently reliable indicator of intoxication to be introduced in conjunction with the results of other field sobriety tests to establish whether a driver was under the influence. Responding to Ballard’s further objections that Trooper Oly might not have administered the HGN tests in the approved manner, Judge Finn ruled that such questions were properly left to the jury.

Although Judge Finn decided Ballard’s Frye motion, District Judge William H. Fuld was the judge who presided over Ballard’s trial. Shortly before the trial began, Judge Fuld warned the State that its evidence concerning Ballard’s performance on the HGN test could not include any testimony estimating Ballard’s level of blood alcohol based on the HGN results. Judge Fuld further warned the State not even to try to quantify Ballard’s performance by referring to the point score used by the troopers to assess a driver’s performance on the test. The judge told the prosecutor, “Keep the numbers out of it.... We won’t hear numbers. You can ask [the trooper] whether [Ballard] passed [the test] or failed it.”

Trial Testimony Concerning Ballard’s Performance on the HGN Test and the Various Other Field Sobriety Tests

At trial, Trooper Oly described how he administered the three parts of the HGN test to Ballard. Holding a pen about six inches in front of Ballard’s face, Oly moved the pen to the left and then to the right, watching Ballard’s eyes for “smooth pursuit” of the moving pen. Oly next held the pen all the way to the left, then all the way to the right, to see whether Ballard’s eyes exhibited involuntary jerking when he looked at the pen at the extremes of his vision. Finally, Oly brought the pen to approximately a 45-degree angle from the center of Ballard’s face, to see if Ballard’s eyes jerked while looking at the pen at that angle.

Oly testified that Ballard’s performance on the HGN test was among the worst of all the people Oly had tested. According to Oly, Ballard’s eyes jerked during all three stages of the test. Oly told the jury that Ballard’s performance on the test was “dramatically different” from what one would expect from a sober person — that Ballard’s performance indicated that he was “a very impaired individual”.

Oly also described the balance and agility tests he had administered to Ballard. In one of these tests, Oly had Ballard stand on one leg. According to Oly, Ballard did “very poorly” on this test: to maintain his balance, Ballard was forced to spread his arms and either put the other foot down or hop onto the other foot. Again, Oly characterized Ballard’s performance on this test as characteristic of a “very impaired person”. Ballard also failed the walk-and-turn test: he wobbled and lost his balance while standing, walking, and turning.

Following these tests, Oly administered verbal tests to Ballard. Ballard was unable to recite the alphabet completely, and he was unable to count backwards from 100 by fives without error. Again, Oly testified that Ballard’s performance was consistent with someone who had been drinking and was impaired.

Based on Ballard’s performance on all these field sobriety tests, as well as Oly’s observations of Ballard’s eyes, breath, and movements when he first got out of the Toyota, Oly concluded that there was “absolutely no doubt” that Ballard was intoxicated (as opposed to simply being tired and having consumed one or two drinks). Oly then arrested Ballard for driving while intoxicated.

Trooper Oscar Siegfried, who was riding as Oly’s supervisor and who observed this entire encounter, confirmed that Ballard was swaying while standing and walking, that Ballard failed the walk-and-turn and one-leg-stand tests, that Ballard appeared highly intoxicated, and that there was a strong odor of alcoholic beverages on Ballard’s breath, an odor that was obvious once Ballard was placed in the patrol car. Siegfried testified that it was “not a close call” as to whether Ballard was intoxicated. Siegfried did not mention the HGN test during his testimony. 3

Ballard took the stand and testified that he had not been intoxicated. He said that he had consumed three drinks the evening before, then had gone to sleep for approximately seven hours. Ballard said that he was still drowsy from this lengthy sleep when the troopers stopped him. Ballard also testified that his balance was normally not that good and that his eyes sometimes had trouble tracking moving objects.

The State introduced no Intoximeter evidence; Ballard was prosecuted under AS 28.35.030(a)(1), which prohibits driving “while under the influence of intoxicating liquor”.

In his summation to the jury at the close of the trial, the prosecutor mentioned the HGN test once: “You have Trooper Oly saying that he administered the HGN test and, based upon his training and experience, he saw that the result was that the defendant was impaired.” However, the prosecutor immediately moved on to the details of the other field sobriety tests that Ballard had failed. The prosecutor emphasized that the troopers had decided to arrest Ballard based on the results of the entire battery of tests, and he urged the jury to determine Ballard’s guilt using the same totality-of-circumstances analysis: “You’ve just got to look at everything.”

Ballard’s attorney mentioned the HGN test once during his summation as well, arguing that it was inconsistent for Trooper Oly to claim that Ballard had been unsteady on his feet and yet able to hold his head steady during the HGN test. The defense attorney also made a general remark that could have been interpreted to include the HGN test along with the other field sobriety tests: “[WJhat should ... be clear to you from hearing how these different tests work is [that], although they may in certain circumstances identify those people who are under the influence of alcohol or impaired by alcohol, ... these tests [also] have the potential for identifying those people who don’t perform well under pressure.” Other than these quoted excerpts, the two attorneys did not emphasize or even mention the HGN test to the jury during summation.

The jury convicted Ballard of driving while intoxicated. He now appeals, contending that the State should not have been permitted to introduce evidence of his performance on the HGN test.

HGN Evidence Satisfies the Frye Test

Under the Frye test (which sets the minimum standard for admissibility of scien- tifie evidence in Alaska), the proponent of scientific evidence is required to demonstrate that the evidence rests on generally accepted scientific principles or discoveries. Contreras, 718 P.2d at 134 (citing Frye, 293 F. at 1014). Ballard argues that the scientific theory underlying the HGN test is not generally accepted in the scientific community, and thus HGN evidence does not satisfy the Frye test.

As can be seen from our description of the testimony presented in the district court, all of the expert witnesses agreed that alcohol consumption causes nystagmus. While there are other substances and disorders that also cause nystagmus, nystagmus is generally recognized as one effect of alcohol consumption.

Further, there appears to be scientific agreement that the more alcohol in a person’s blood, the more likely the person is to exhibit observable nystagmus. For example, in one study cited by Ballard’s expert, Dr. Nowaczyk, the HGN test was administered to drivers with a wide range of blood-alcohol levels: 15 percent of drivers with blood-alcohol levels of .00 to .04 percent failed the HGN test; 64 percent of drivers with blood-alcohol levels of .05 to .09 percent failed the test; and 95 percent of drivers with blood-alcohol levels of .10 to .15 percent failed the HGN test.

In the last decade, numerous state courts have addressed the question of whether HGN evidence meets the Frye standard. Of the courts that have resolved this issue, most have held that HGN evidence does not satisfy Frye if the HGN test results are introduced to establish that the person tested had a particular level of blood alcohol. However, these same courts hold that HGN evidence satisfies Frye if the HGN test results are admitted for the more modest purpose of indicating that the person has consumed alcohol and is potentially under the influence. See Whitson v. State, 314 Ark. 458, 863 5.W.2d 794, 797-98 (1993); State ex rel. Hamilton v. City Court of City of Mesa, 165 Ariz. 514, 799 P.2d 855, 857-58 (1990); State v. Superior Court, 149 Ariz. 269, 718 P.2d 171, 181-82 (1986); People v. Joehnk, 35 Cal.App.4th 1488, 42 Cal.Rptr.2d 6, 15-17 (1995); Hawkins v. State, 223 Ga.App. 34, 476 S.E.2d 803, 808-09 (1996); State v. Garrett, 119 Idaho 878, 811 P.2d 488, 491 (1991); People v. Buening, 229 Ill.App.3d 538, 170 Ill.Dec. 542, 547-48, 592 N.E.2d 1222, 1227-28 (1992) 4 ; People v. Wiebler, 266 Ill.App.3d 336, 203 Ill.Dec. 597, 600, 640 N.E.2d 24, 27 (1994); State v. Armstrong, 561 So.2d 883, 887 (La.App.1990); State v. Taylor, 694 A.2d 907, 911-12 (Me.1997); Schultz v. State, 106 Md.App. 145, 664 A.2d 60, 69-70, 74, 75-77 (1995); People v. Berger, 217 Mich.App. 213, 551 N.W.2d 421, 423-24 (1996) (per curiam); City of Fargo v. McLaughlin, 512 N.W.2d 700, 707-08 (N.D.1994); State v. O’Key, 321 Or. 285, 899 P.2d 663, 689-690 (1995) 5 ; State v. Sullivan, 310 S.C. 311, 426 S.E.2d 766, 769 (1993); Emerson v. State, 880 S.W.2d 759, 769 (Tex.Crim.App.1994); Richardson v. State, 766 S.W.2d 538, 540 (Tex.App.1989).

Ballard’s brief cites several cases in which the admission of HGN evidence was held to be error because the trial court failed to recognize the Frye issue or hold a hearing to decide it. However, Ballard cites no eases in which the Frye issue was actually litigated and an appellate court decided that HGN evidence did not meet the Frye standard. 6

From our examination of the record in this case, as well as our examination of the lengthy discussions of the scientific research and literature concerning HGN that is contained in such decisions as State v. O’Key, Schultz v. State, and State v. Superior Court, supra, we conclude that the science underlying the HGN test — and, in particular, the principle that alcohol consumption causes nystagmus — is generally accepted within the relevant scientific community. 7 Moreover, there is general agreement that a properly trained police officer (or other trained lay person) can observe nystagmus under field conditions. While there is considerable debate about the power of the HGN test to disclose or predict a person’s precise level of blood alcohol, there is widespread agreement that horizontal nystagmus is a reliable and valid indicator of the presence of alcohol in a person’s blood and, consequently, of the person’s potential intoxication.

Ballard contends that HGN testing is not a “reliable” indicator of intoxication — that is, the test does not yield sufficiently uniform results. He points to the testimony of Dr. Nowaczyk, who quoted a study indicating that HGN testing is reliable only 59 to 66 percent of the time. Other studies have shown much higher reliability for the HGN test; these studies suggest that the three-test battery recommended by the NHTSA (HGN, walk-and-turn, one-leg-stand) is the most accurate field technique for gauging whether a person is under the influence. See Schultz, 664 A.2d at 73 n. 12. In our view, the difference among the studies is not significant.

Dr. Nowaczyk told the court that a reliability rate of 59 to 66 percent would be acceptable for some purposes (specifically, medical research), but would not be considered acceptable for making medical diagnoses or for “making decisions as to whether someone meets a certain level” of blood-alcohol. As we have explained above, we do not find that HGN testing is a reliable indicator of a person’s precise level of blood alcohol. Our conclusion is more modest: we find HGN testing to be a reliable indicator of a person’s alcohol consumption and, to that extent, HGN results are relevant circumstantial evidence to prove a person’s impairment. Even assuming that the HGN test is reliable only 60 percent of the time, this is a sufficient level of reliability for the HGN test to be admitted as an indicator of potential intoxication.

Ballard also argues that HGN testing is not a “valid” indicator of intoxication — that is, the test does not yield sufficiently accurate evidence of intoxication. He relies on studies showing that many people fail the HGN test even though their blood-alcohol levels are lower than .10 percent. Ballard asserts that the validity of the HGN test is undermined by these “false positives”. But Ballard’s argument (and most of the studies he cites to support it) are premised on the assumption that, if the HGN test is a valid indicator of intoxication, then a person whose blood-alcohol level is lower than .10 percent should pass the test. This premise is mistaken.

Alaska law forbids a person from driving while he or she is under the influence of alcohol — when the person’s ability to control a motor vehicle is impaired. AS 28.35.030(a)(1). While the law also forbids people from driving when their blood-alcohol level is .10 percent or higher, see AS 28.35.030(a)(2), people can be “under the influence” for purposes of 030(a)(1) even when their blood-alcohol level is less than .10 percent. Thus, when a person with a blood-alcohol level of less than .10 percent fails the HGN test, this does not necessarily mean that the HGN test has falsely identified that person as being under the influence of alcohol.

Moreover, as all of the expert witnesses pointed out, nystagmus is caused by substances other than alcohol, as well as by various disorders and diseases. The fact that a sober person fails the HGN test does not mean that the tester mistakenly identified nystagmus when there was none. It could also mean that the person exhibited nystagmus, but the nystagmus occurred for reasons unrelated to alcohol consumption. In this respect, the HGN test is no different from the other field sobriety tests that measure coordination, balance, or verbal responses: people can fail these tests for reasons other than intoxication.

The real question is whether there is a sound scientific basis for concluding that a person’s performance on the HGN test is a relevant factor to be considered when determining whether that person is under the influence of alcohol. As we have already discussed, even the studies cited by Ballard show that performance on the HGN test is related to blood-alcohol level: people with low blood-alcohol levels (.00 to .04 percent) are quite likely to pass the HGN test, some 50 to 65 percent of people with blood-alcohol levels of .05 to .09 percent fail the test, and people with high blood-alcohol levels (.10 percent and above) are quite likely to fail the test. Despite the fact that a person’s level of performance on the HGN test can not necessarily be tied to a particular blood-alcohol level, the studies show that the HGN test is a significant indicator of alcohol consumption. Thus, it is circumstantial evidence of impairment caused by alcohol consumption.

For these reasons, we conclude that HGN evidence meets the Frye standard for admission of scientific evidence if the test results are admitted for the limited purpose of establishing that a person has consumed alcohol and is therefore potentially impaired. While HGN testing may not, of itself, be sufficient to establish intoxication, HGN test results are admissible as a factor to be considered by the fact-finder when determining intoxication. Testimony concerning a defendant’s performance on a properly administered HGN test is admissible on the issue of impairment, provided that the prosecution claims no greater reliability or weight for the HGN evidence than it does for evidence of the defendant’s performance on any of the other standard field sobriety tests, and provided further that the prosecution makes no attempt to correlate the HGN test result with any particular blood-alcohol level, range of blood-alcohol levels, or level of impairment. 8

The Foundational Requirements for Introducing HGN Evidence

Although we have ruled that HGN evidence (offered for the limited purposes described above) satisfies the Frye standard, another question of evidence law remains. Ballard questions whether Trooper Oly was qualified to testify concerning Ballard’s performance on the HGN test.

The fact that particular scientific evidence meets the Frye standard means only that this evidence is a proper subject for testimony. Schultz, 664 A.2d at 70-71, 75-76. The party offering such evidence must still present a witness who is qualified to testify on this subject. In particular, under Alaska Evidence Rule 702, whenever a party intends to offer evidence concerning any “scientific, technical, or other specialized knowledge”, the party must establish that the witness who will offer this evidence possesses the “knowledge, skill, experience, training, or education” to offer an informed opinion on the subject.

Among the courts who agree that HGN evidence satisfies the Frye test, the virtually unanimous opinion is that a police officer may testify to the results of HGN testing (that is, a scientist is not required), but the government must establish as a foundational matter that the officer has been adequately trained in the administration and assessment of the test. See Hawkins, 476 S.E.2d at 808-09; Buening, 170 Ill.Dec. at 547-48, 592 N.E.2d at 1227-28; Taylor, 694 A.2d at 911-12; Schultz, 664 A.2d at 75-76; Berger, 551 N.W.2d at 423-24; McLaughlin, 512 N.W.2d at 707-08; O’Key, 899 P.2d at 689-690; Emerson, 880 S.W.2d at 769.

We adopt the same rule. Further, we find that the State met this foundational requirement in Ballard’s case. At the pre-trial hearing, the State presented extensive evidence concerning the proper administration and assessment of the HGN test: the State played a videotape produced by the NHTSA that demonstrated the proper administration of the test, and the State offered the testimony of Trooper Schoenberg, an HGN test instructor, who demonstrated the test in the courtroom and explained how the test results are scored. Trooper Oly, the officer who arrested Ballard, then testified regarding how he had been trained to administer and assess the HGN test, and he too demonstrated the three portions of the test on a volunteer in the courtroom.

As noted above, Ballard presented only one criticism of Oly’s in-court demonstration of the test: Oly had performed all three portions of the test on one eye before moving to the other eye, while the NHTSA recommends alternating eyes. However, Ballard gave the court no reason to believe that this variation could be expected to make any difference in the test results. On this record, Judge Finn could properly conclude that Trooper Oly had received adequate training in the administration of the HGN test and that Oly was therefore qualified to testify about the results of Ballard’s HGN test.

We recognize that Ballard potentially had another foundational attack on the State’s HGN evidence. Ballard could argue that Trooper Oly failed to substantially comply with proper testing procedures when he administered the test to Ballard. Compare Williams v. State, 884 P.2d 167, 174 (Alaska App.1994) (holding that “substantial compliance” with Intoximeter procedures is required to establish the foundation for admitting breath test evidence).

Most of the courts cited above treat this issue as affecting the weight of HGN evidence rather than its admissibility. They recognize that a poorly administered HGN test may yield results that are not particularly probative, but they note that the same is true of the other field sobriety tests.

On appeal, Ballard argues that proper testing procedure is a foundational matter— something that the State must prove before HGN evidence can be admitted. Quoting People v. Leahy, 8 Cal.4th 587, 34 Cal.Rptr.2d 663, 667, 882 P.2d 321, 325 (1994), he argues that the proponent of scientific evidence “must demonstrate that correct scientific procedures were used in the particular ease”. Ballard then faults Judge Finn for failing to decide whether Oly properly administered the HGN test to him.

Even assuming that a substantial variance from accepted testing norms could render HGN evidence inadmissible, Ballard presented no evidence of such a substantial variance at the pre-trial hearing. It must be remem bered that the primary purpose of that pretrial hearing was to determine whether HGN evidence met the Frye standard for admissibility. Even if Judge Finn ruled that HGN satisfied the Frye test, Ballard would still have another opportunity (at his trial) to attack the particular procedures that the trooper employed when administering the test to him.

This is seemingly the reason why the evidence presented at the pre-trial hearing centered on the scientific basis for the HGN test and the approved methods for administering the test, and not on the particular procedures Oly employed when administering the test to Ballard. Ballard’s attorney did cross-examine Oly to a limited extent about the trooper’s administration of the HGN test to Ballard (the procedures he used and the observations he made), but this cross-examination occupies only 7 pages of transcript out of a total of 420 pages.

As Judge Finn later noted in her written decision, “[tjhere was some discussion during the hearings ... about whether the HGN test was administered accurately to [Ballard].” However, given the evidence before her, Judge Finn could justifiably conclude that Trooper Oly was generally familiar with the proper procedures for administering and scoring the HGN test, and that Oly had generally followed these procedures when he tested Ballard. Accordingly, Judge Finn correctly ruled that any remaining questions concerning the administration of the HGN test to Ballard, and how Oly’s testing procedures might have affected the probative force of the test results, were matters to be brought out during cross-examination and then argued to the jury, just as questions about the proper administration of the other field sobriety tests are argued to the jury.

Ballard does not contend on appeal that Trooper Oly’s testimony reveals any significant non-compliance with established testing procedures. We recognize that a trial court retains the discretion to exclude HGN test results if the test procedures were so substandard as to preclude any reasonable juror from concluding that the results were meaningful. However, our review of the record in Ballard’s case reveals that any such foundational objection would have been meritless: there was clearly sufficient evidence for the jury to find that the HGN test had been administered in a manner likely to yield relevant results.

The Trooper’s Assessment that Ballard’s HGN Test Results Showed that He Was ‘Very Impaired"

Thus far, we have concluded that a properly qualified law enforcement officer can testify concerning the administration .of the HGN test and then give an assessment of a defendant’s test results — including an explanation that the presence of nystagmus is an indication that the defendant had been consuming alcohol, thus providing circumstantial evidence that the defendant was under the influence. We have further concluded that, based on the record before her, Judge Finn could properly find that Trooper Oly was properly qualified to administer the HGN test to Ballard and to assess the test results. Because of Judge Finn’s ruling, Oly could testify that Ballard had failed the HGN test and that this failure indicated that Ballard had alcohol in his blood and might currently be under the influence.

However, we have also concluded that HGN evidence is not admissible to prove that a defendant has a particular level of blood alcohol. Because of this ruling, we are concerned by one aspect of Trooper Oly’s testimony — his elaboration that Ballard’s HGN test results showed that Ballard was “a very impaired individual”. Oly’s assertion concerning Ballard’s degree of intoxication intrinsically rests on the premise that the worse a person’s performance on the HGN test, the more impaired that person must be. In our view, this testimony is perilously close to the type of HGN testimony we have rejected: the assertion that HGN test results can be correlated to a particular blood-alcohol level. (We further note that, despite Trooper Oly’s training in the administration and scoring of the HGN test, the record fails to show that he has the kind of scientific expertise that would qualify him to draw any connection between the degree of onset or obviousness of a person’s nystagmus and the degree of that person’s impairment.)

However, even though this aspect of Trooper Oly’s testimony may have amounted to an impermissible attempt to correlate Ballard’s HGN test result with a particular level of impairment, we conclude that the error, if any, was harmless under the facts of this case. Both Oly and his supervisor, Trooper Siegfried, gave detailed testimony describing the numerous ways in which Ballard’s intoxication was manifest: his breath, his eyes, his inability to stand or balance, his inability to recite the alphabet or to count backwards. Moreover, the prosecutor at Ballard’s trial did not accentuate the HGN test. The prosecutor noted that the HGN test results indicated that Ballard was “impaired”, but the prosecutor gave equal prominence to Ballard’s poor performance on the other field sobriety tests, and he urged the jury to “look at everything” when determining whether Ballard was under the influence. Given this record, we conclude that the jury’s verdict could not have been affected by Oly’s arguably impermissible assessment of the HGN test results.

Conclusion

The judgement of the district court is AFFIRMED.

COATS, J., not participating.

1 . Oly was a newly-commissioned trooper who was completing his field training under the supervision of Trooper Siegfried.

2 . As explained by Dr. Ronald Nowaczyk, one of the expert witnesses in this case, the "reliability" of a scientific test or procedure refers to the degree of uniformity in the results — the likelihood that the test or procedure will again and again yield the same results under the same testing conditions. The "validity” of a test or procedure refers to the likelihood that the results accurately reflect the presence or absence of the thing the scientists are looking for. See Charles R. Honts & Susan L. AmatoHenderson, "Horizontal Gaze Nystagmus Test: the State of the Science in 1995", 71 North Dakota Law Rev. 671, 677(1995).

3 . Siegfried made one passing reference to Ballard's "high visual problems with motor skills”. From the context of the remark, it appears that Siegfried was speaking of Ballard’s "highQy] vis[ible] problems” with his motor skills, rather than Ballard’s performance on the HGN test.

4 . But see People v. Kirk, 289 Ill.App.3d 326, 224 Ill.Dec. 452, 456-57, 681 N.E.2d 1073, 1077-78 (1997), declining to follow Buening on the issue of whether HGN has been so widely accepted at the appellate level that, even in the absence of a Frye hearing, a court can take judicial notice that HGN meets the Frye test.

5 . Oregon does not follow the Frye test. Nevertheless, the O'Key court found that the scientific principles underlying HGN testing are in fact generally accepted within the relevant scientific community. O'Key, 899 P.2d at 686.

6 . Ballard does cite one case, Young v. City of Brookhaven, 693 So.2d 1355, 1360-61 (Miss.1997), in which a state supreme court ruled that the scientific principles underlying HGN do not meet the Frye standard. However, the Mississippi court's reasoning is simply confused.

There was no Frye hearing in Young, and the defendant appealed on precisely that basis. He claimed that HGN was "scientific evidence” for purposes of the Frye rule, and that therefore the State should not have been able to introduce testimony about the HGN results without first presenting an expert witness to lay a Frye foundation. After reviewing decisions from other states, the Mississippi Supreme Court adopted the majority view that HGN is indeed scientific evidence. From this premise, however, the Mississippi court reached a surprising conclusion. Instead of ruling that the HGN evidence should not have been admitted because the trial court held no Frye hearing, the Mississippi court concluded that the scientific principles underlying HGN are not generally accepted within the scientific community:

We find that the HGN test is a scientific test.... Whereas most other field sobriety tests arise out of a juror's common experi- . enees, ... the HGN test relies upon a scientific[ ] ... set of observations.

Therefore, this Court finds that the HGN test is not generally accepted within the scientific community and cannot be used as evidence to prove intoxication or as a mere showing of impairment.

Young, 693 So.2d at 1360-61.

In this passage, the Mississippi court confuses two issues: (a) does HGN testing rest on scientific principles, so that a Frye hearing must be held before HGN evidence can be introduced? and (b) if a Frye hearing were held, would the scientific principles underlying the HGN test prove to be generally accepted within the scientific community? The Mississippi court decided (as we decide here) that HGN evidence is "scientific evidence” for Frye purposes. It therefore follows that HGN evidence can not be admitted unless the proponent of the evidence presents expert testimony to prove that HGN satisfies the Frye test. There was no Frye hearing in the Mississippi trial court, so the Mississippi Supreme Court could draw no conclusions about the scientific acceptance of HGN from the trial court record. Moreover, none of the case law cited by the Mississippi Supreme Court holds that HGN fails to satisfy Frye. In short, there was no factual or logical basis for the Mississippi court's conclusion that the principles underlying HGN are not generally accepted in the scientific community.

7 . On appeal, Ballard faults the district court for indicating that one of his expert witnesses might not be within the relevant community of scientists whose views are to be considered when deciding the Frye issue. As indicated in our discussion, we have considered the views of both of Ballard’s expert witnesses, and our decision is premised on the assumption that both of Ballard's witnesses are members of the relevant scientific community.

8 . The government may present evidence explaining the scientific underpinnings of the HGN test in response to a defense attack on the meaning of the test results. However, such evidence would have to be offered through the testimony of a person qualified as an expert witness under Alaska Evidence Rule 702.