State v. Streich

658 A.2d 38 (1995)

STATE of Vermont
v.
Todd STREICH.

No. 91-335.

Supreme Court of Vermont.

February 17, 1995.

Scot Kline, Chittenden County State's Atty., and Pamela Hall Johnson, Deputy State's Atty., Burlington, and Jeffrey L. Amestoy, Atty. Gen., and David Tartter, Asst. Atty. Gen., Montpelier, for plaintiff-appellee.

Robert Andres and Michael Johnson, Burlington, and Charles S. Martin of Martin & Paolini, Barre, for defendant-appellant.

Before ALLEN, C.J., and GIBSON, DOOLEY, MORSE and JOHNSON, JJ.

DOOLEY, Justice.

Defendant Todd Streich was convicted of sexual assault in violation of 13 V.S.A. § 3252(a)(1)(A). On appeal, he raises eleven allegations of error which can be consolidated into five main issues: (1) whether the court improperly admitted DNA evidence which linked defendant to the crime scene; (2) whether the court improperly admitted blood-type evidence of another individual which excluded that individual as a possible perpetrator; (3) whether comments made by the judge during jury selection warranted a mistrial; (4) whether the court's instructions to the jury were wrong; and (5) whether the court exhibited open bias and prejudice against defendant in its rulings. Although we disagree with the trial court's rationale regarding the admissibility of DNA statistical evidence, we affirm.

I.

On June 22, 1989, a young woman was sexually assaulted in her home. Immediately after the assault, the victim was taken to the hospital where she underwent an extensive internal examination. A variety of evidence was collected including vaginal swabs, and blood and pubic hair samples. The victim's underwear, which was stained with the assailant's semen, was also taken.

During the investigation, a detective spoke with a man named Mark Rouelle, who gave the detective specific details of the crime, and told the detective that defendant had related these details to him. The investigation then focused on defendant, and the Vermont State Police forwarded to the FBI criminal lab the evidence collected at the hospital and blood samples from the victim and defendant for DNA and blood-typing analysis. The FBI compared the DNA from defendant's blood sample to the semen found on the victim's underwear and reported that defendant's genetic profile matched the genetic profile of the semen at three genetic locations. The FBI concluded that the probability of another person chosen at random having the same DNA profile was 1 in 50,000.

Prior to the July 1991 trial, defendant filed a motion to exclude the DNA evidence. He argued that DNA profiling is not sufficiently reliable to be admitted in Vermont criminal trials, and in the alternative, the FBI used faulty procedures that undermined the probative value of the evidence. The trial court held a four day hearing, and heard extensive expert testimony on behalf of both the State and defendant. Following the hearing, the court issued a bifurcated order rejecting defendant's arguments.

Shortly after the court issued its DNA order, defendant began pursuing the theory that Mark Rouelle, the individual who had initially led investigators to defendant, actually committed the crime. When the State became aware of this theory, it requested blood samples from Rouelle to compare to the evidence taken from the victim and from the crime scene. Rouelle was uncooperative, and the State's efforts to obtain these samples was further delayed by his frequent change of counsel and separate hearings to determine his competency.

One week before trial, the trial court issued a nontestimonial identification order (NTO) requiring Rouelle to provide the requested samples. The State never exercised this NTO because Rouelle voluntarily offered the blood samples. The trial court's issuance of the NTO is relevant, however, because defendant sought to introduce the statutory standards under V.R.Cr.P. 41.1(c) for obtaining NTOs to support his theory that Rouelle had committed the crime. The trial court held that the statutory standards were not evidence, and therefore, were not relevant.

Rouelle's blood samples were analyzed the week before trial. Because DNA profiling can take up to six weeks, the laboratory was able to determine only Rouelle's blood-type and secretor status. The laboratory report indicated that Rouelle's blood did not match the evidence found at the crime scene, and these results were forwarded to the State on July 8, 1991. The next day, which was the day before trial, the State provided defendant with a copy of the report.

Defendant immediately complained that the notice was untimely, and moved for a continuance so that his expert could review the report. Although the trial court agreed that the Rouelle report was untimely, it denied the motion. Instead, the court ruled that the State was prohibited from mentioning any information contained in the report during its opening statement or case-in-chief, and that the report might be barred altogether depending on what evidence defendant presented at trial. The court reasoned that this postponement would provide defendant's expert with an opportunity to review the evidence. During cross-examination of one of the State's key witnesses, defendant advanced his theory that Rouelle had committed the crime. The trial court ruled that because defendant initiated the theory, the State was entitled to rebut it. Consequently, the court permitted the State to admit into evidence the Rouelle report, which indicated that Rouelle's blood-type was inconsistent with evidence from the crime scene and the victim.

A major network filmed the trial for a television documentary. On July 8, 1991, prior to jury selection, the court discussed with counsel how the presence of cameras and bright lighting should be explained to the jury. Both parties agreed that the potential jurors should be told that the court was conducting a media experiment, and they were so informed. Before resuming the jury draw on July 9th, the State expressed its discomfort with the fabrication, especially since some of the jurors already knew that the filming was for television. The court agreed to examine the remaining jurors on whether they had read or heard anything in the media regarding the trial. Only one juror responded affirmatively, and he was excused from the panel. The judge then informed the jurors of the true reason for the cameras and lighting, and inquired whether the filming would interfere with their responsibilities as jurors. No one indicated that it would.

On the fourth day of trial, defendant moved for a mistrial on the ground that the court's misrepresentation concerning media coverage had poisoned the jury selection process. In support of his motion, defendant noted that members of the July 8th jury array who were not selected were upset that the court had not been candid with them about the television coverage. Defendant argued that because of the court's misrepresentation, jurors lacked confidence in the overall integrity of the trial. The court denied the mistrial motion, stating that any potential problem with the jury was effectively addressed on July 9th. The rest of the trial proceeded without incident, and the jury found defendant guilty of sexual assault; this appeal followed.

II.

A.

The most significant issue in this case concerns the admissibility of DNA profiling in a criminal case to prove the perpetrator's identity.^[1] Before we can adequately address this issue, it is necessary to outline the procedures involved in DNA profiling.

Deoxyribonucleic acid (DNA) is the codified genetic blueprint of humans, and with the exception of identical twins, no two people share the same pattern of DNA. DNA is found in almost every cell of the body, including hair follicles, blood, and semen. The DNA molecule is composed of 3 billion "base pairs" of four different chemicals, and the particular order or pattern of these base pairs dictates genetic characteristics. Because 99% of the DNA molecule is the same for all humans, DNA profiling focuses on those areas of the DNA molecule where there is significant differentiation of the base pair pattern. These areas of significant differentiation are called "polymorphic," and base pair patterns in polymorphic areas are called "alleles."

The basic profiling procedure is to compare DNA from the defendant with DNA from the assailant. The matching process involves the use of two distinct disciplines: molecular biology and population genetics.^[2] The biological component of the test utilizes a process called Restriction Fragment Length Polymorphism (RFLP), where the specific alleles in polymorphic areas of the molecule are isolated, photographed, and measured. If the RFLP process concludes that a match exists, scientists then use population genetics to determine the probability that the match occurred merely by coincidence.

The first step of RFLP entails extracting DNA from both the evidence obtained from the victim and found at the crime scene and from samples provided by the defendant. It is then "cut" with chemical scissors at all places along the molecule where polymorphic chemical base pair sequences occur. The cut fragments are placed in a gel to which an electrical current is applied. The process, known as gel electrophoresis, causes the larger pieces to remain at one end of the gel and the smaller fragments to move to the other. Because gel consistency may vary and thus cause a difference in the speed with which the fragments move through the gel, the defendant's sample and the comparison sample are run on the same gel but in different tracks or lanes.

After the DNA fragments are sorted by size, they are transferred to a nylon membrane, which is easier to handle than the gel. During this stage, the DNA fragments are split lengthwise along each base pair so that the base pairs are separated into two strands. Next, radioactive probe markers designed to match or complement the single-standed alleles are applied to the membrane. The use of several probes is necessary because it is common for two or more individuals to have the same alleles even in a polymorphic area. It is less common, however, for two individuals to share several alleles as identified by four or five different probes.

The nylon membrane is then placed against a piece of x-ray film so that the radioactive probes can expose the film where all the tagged alleles are located. This picture is called an autoradiograph or autorad, and resembles the bar code on grocery store packages. It is also widely known as the "DNA fingerprint." The last step of the RFLP test involves measuring the dark bands or bar codes on the autorad to determine whether a match exists between the sample from the defendant and that of the assailant. The length and width of each band on the autorad is measured, and a match is declared if the defendant's and the assailant's samples are within 5% of each other. The 5% margin of error serves to account for mobility and quality differences between the samples. A visual assessment is also made which compares the relative position and color intensity of the autorad bands.

The existence of a match does not necessarily prove that the DNA obtained from the victim or from the crime scene came from the defendant; it shows only that the DNA profile of the defendant is consistent with the DNA profile of the crime scene evidence. Using population genetics, the significance of the match is determined by the probability that an individual randomly selected from the general population has the same DNA profile as the defendant.

The science of population genetics relies heavily on a database comprised of RFLP results of a sample population. The database serves two functions. First, it guarantees that the targeted alleles are truly polymorphic. If an allele appearing on the autorad is shared by everyone, it imparts no information about defendant. Second, the database is thought to provide information about the frequency with which specific alleles appear in the population. Whether the database actually serves this function, however, is the subject of controversy.

One of the most significant criticisms involves the population experts' derivation of probabilities from multiple allele matches. Since the DNA profile is based on numerous individual alleles, the overall probability determination involves a combination of numerous individual probabilities. The probability of a match for each allele is derived from the database. For example, if a given allele appears in 10% of the database samples, it is assumed that it appears in 10% of the general population. Because these probabilities vary from one allele to another, the second allele on the autorad might have a probability of 20%, and the third allele a probability of 5%.

The overall probability that an individual randomly selected from the general population has the same DNA profile as the defendant involves a combination of the individual allele probabilities. Population experts have usually relied upon the "product rule" to calculate the overall probability, simply by multiplying the individual probabilities for each allele. Thus, in the hypothetical in the foregoing paragraph, the odds that another individual, randomly selected from the population, would share defendant's DNA profile is 1 in 1,000 (1/10 x 1/5 x 1/20).

The validity of the product rule depends on whether the matches at each allele are statistically independent. If the three alleles are related, or linked, the product rule can not be applied because there is a significant probability that an individual who has one allele has two or more of the alleles. For example, suppose the probabilities described above relate to whether an individual has blond hair, blue eyes, and fair skin. If the probability of having all three traits is interrelated, such that someone who is blond is more likely to have blue eyes, then the characteristics are interrelated. The product rule is not accurate because instead of a one in 1,000 match, the probability of all three events occurring together is almost certainly much higher.

Whether the targeted alleles in the RFLP process are statistically independent is said to depend on the absence of population substructure in the database. Scientists have accounted for some population substructuring by segregating databases into racial groups. The conclusion, however, that population substructuring does not occur even within racial classifications assumes that the general population freely migrates and mates in a totally random fashion. The validity of this assumption, and its impact on the statistical accuracy of the database, has divided the scientific community. See, e.g., People v. Barney, 8 Cal.App.4th 798, 10 Cal.Rptr.2d 731, 740-41 (1992) (contrasting Lewontin & Hartl, Population Genetics in Forensic DNA Typing, 254 Science 1745 (1991), and Chakraborty & Kidd, The Utility of DNA Typing in Forensic Work, 254 Science 1735 (1991)).

The FBI has developed a statistical process known as "fixed binning" which it contends resolves the substructure problem in the defendant's favor. As a result of population studies, scientists have observed that for a particular probe, polymorphic allele lengths will vary among individuals. These allele lengths are sorted into categories called bins whereby each bin represents an observed range of allele lengths. By counting how frequently alleles in the population sample fall within a specific bin, scientists are able to calculate the likelihood that a match on a given probe will be coincidental. For example, in a study of 200 individuals, if twenty of those individuals fit into a particular bin for a particular probe, that bin is said to have a 10% occurrence within the population.

This process purportedly accounts for population substructure in two ways. First, if a given bin contains fewer than five individuals, this bin is "collapsed" into an adjoining bin. As a result, the percentage of the population with an allele within this redetermined bin will be higher than it would be otherwise, and the ultimate statistical probability calculation will be more conservative. Second, if a suspect's allele length for a particular probe falls in between two bins, it will be assigned to the bin with the greater frequency figure, ostensibly generating a conservative statistical probability. This conservatism is said to make up for any inaccuracies due to the presence of population substructure.

The "ceiling principal," recommended in a 1992 report by the National Research Council (NRC), is another method of accounting for population substructure through conservative calculation of the probability statistic. Lander & Budowle, DNA Fingerprinting Dispute Laid to Rest, 371 NATURE 735, 736 (Oct. 27, 1994). It has gained favor among scientists and the courts. Id.; see also Commonwealth v. Lanigan, 419 Mass. 15, 641 N.E.2d 1342, 1347 (1994). This method assumes that population substructures exist and requires the probability statistics to be increased to reflect a worst-case scenario. Thus, this approach attempts to resolve any inaccuracy in the defendant's favor. Lander & Budowle, supra, at 736. The practical effect of this approach is to pare back the statistical significance of a DNA match. Id. at 737. For example, although a four-probe match utilizing the product rule might reveal "frequencies of about 10-8-10-10, the ceiling principle pares them back to about 10-6-10-7." Id.

In this case, FBI experts used the fixed-bin method, and determined that defendant's DNA matched that of the assailant's DNA at four targeted alleles, but one allele match was excluded by the court. The experts concluded that the frequency probabilities of the first and second probes were one in seventy, and one in ten, respectively. For the two remaining probes, the autorad on defendant's recent sample could not be interpreted, so the FBI used another autorad of defendant which they had on file in connection with an earlier analysis. This comparison of two different autorads is known as cross-gel analysis, and it revealed that defendant's DNA also matched with respect to the third probe. This probe had a frequency probability of one in seventy-eight. On the fourth probe, the experts also concluded that a match existed, but admitted that their analysis of the autorad from the earlier sample had declared the measurements of the fourth probe to be unreliable. The trial court excluded any evidence regarding the fourth probe. Using the product method, the FBI witness testified that the probability that a person selected from the general population would have the same DNA characteristics was approximately 1 in 50,000.

Defendant raises three arguments contesting the use of DNA evidence at his trial: (1) DNA profiling is not generally accepted by the scientific community, and therefore, the introduction of this evidence at his trial was error; (2) it was error to admit evidence regarding the statistical probability of the DNA profile match because these statistics unduly prejudice the jury; and (3) the FBI testing laboratory failed to follow established procedures. We combine the first two arguments and start with them.

B.

Defendant's arguments are premised on his claim that the admissibility of novel scientific evidence is governed by the test set forth in Frye v. United States, 293 F. 1013 (D.C.Cir.1923). Frye held that such evidence could be admitted only if it is generally accepted within the scientific community. Id. at 1014. Although the Frye standard has been widely used, this Court has never adopted it. The issue is governed by V.R.E. 702, which allows admission of scientific evidence if it "will assist the trier of fact to understand the evidence or to determine a fact in issue." See State v. Percy, 156 Vt. 468, 475, 595 A.2d 248, 252 (1990). Our rule is identical to the federal rule, and the United States Supreme Court recently held in Daubert v. Merrell Dow Pharmaceuticals, Inc., ___ U.S. ___, ___, 113 S.Ct. 2786, 2794, 125 L.Ed.2d 469 (1993), that Fed. R.Evid. 702 superseded the Frye test. Noting the identity of the evidence rules, we recently decided to follow Daubert's principles in scientific evidence cases. State v. Brooks, 162 Vt. ___, ___, 643 A.2d 226, 229 (1993). Daubert, not Frye, controls our inquiry.

The Daubert Court concluded that Rule 702 relaxed the traditional barriers to admission of expert testimony, Daubert, ___ U.S. at ___, 113 S.Ct. at 2794, and replaced the Frye "general acceptance" requirement with a flexible standard governed by two principles: reliability and relevance.^[3]Id. at ___, 113 S.Ct. at 2797. The principles are derived from two components of Rule 702, the requirement of "scientific knowledge" and the requirement that the evidence assist the trier of fact. Id. at ___, 113 S.Ct. at 2796.

Reliability is assured if the expert testimony is supported by "scientific knowledge," defined as information that is more than a subjective belief or unsupported speculation, and that is grounded in the methods and procedures of science. Id. at ___, 113 S.Ct. at 2795. Science represents a "`process for proposing and refining theoretical explanations about the world that are subject to further testing and refinement.'" Id. (quoting Brief for American Association for the Advancement of Science and the National Academy of Sciences as Amici Curiae 7-8) (emphasis in original). Of course, there are no certainties in science.

Based on this overview and using the self-descriptor of "general observations," the Court outlined four nonexclusive factors designed to assist a trial judge how to determine whether expert testimony was sufficiently rooted in "scientific knowledge" to be admissible: (1) whether the theory or technique involved is capable of being tested, id. at ___, 113 S.Ct. at 2796; (2) whether the theory or technique has been subjected to peer review and publication, id. at ___, 113 S.Ct. at 2797; (3) the known or potential rate of error associated with the scientific technique, id.; and (4) whether the theory or technique has been generally accepted in the scientific community. Id. The fourth factor is, of course, the Frye standard, but it is now only one of many factors without controlling significance. If the evidence is sufficiently rooted in scientific knowledge, it must also be relevant such that it will "assist the trier of fact to understand the evidence or to determine a fact in issue." Id. at ___, 113 S.Ct. at 2796.

In applying the reliability and relevancy framework outlined in Daubert to the present case, it is clear that the relevancy component of the analysis is not at issue. Because the identity of the perpetrator is a key factual issue, the fact that DNA found on the victim matches that from defendant is probative and helpful to the trier of fact. See State v. Anderson, 118 N.M. 284, 881 P.2d 29, 47 (1994). Consequently, our analysis need examine only the reliability of the two components of DNA profiling.

To ensure consistency in our approach to the admissibility of DNA evidence and the statistics derived from DNA matches, we must look at the foundation evidence received in this case, the scientific literature, and the analysis in other courts. As has been typical of criminal cases in which the prosecution seeks to admit DNA evidence, the foundation evidence shows sharp differences in the scientific community and supports virtually any outcome. The trial court is, of course, entitled to discretion in evaluating foundation evidence. See V.R.E. 104(a); State v. Ogden, 161 Vt. 336, 343, 640 A.2d 6, 11 (1993).

We also admit openly that our conclusion is greatly influenced by the timing of this decision. As is typical in DNA cases, the legal and scientific climate is vastly different as we decide the issue from that which was present when the trial court rendered its decision. Daubert has emerged, clearly stating a more liberal approach to the admission of such evidence. Part of the scientific debate has essentially ended in favor of DNA admissibility. See, e.g., State v. Vandebogart, 616 A.2d at 492 (authorizing trial courts to take judicial notice of RFLP). That part which involves the proper probability statistics to employ has not abated, and strong support has emerged for the ceiling principle. See, e.g., United States v. Porter, 618 A.2d 629, 637 (D.C.1992); see also Lander & Budowle, supra, at 737 (noting disagreement over best solution to population probability theories). Although admissibility cannot turn on the content of the most recent publication in a scientific journal, we must be cognizant of emerging trends.

The first aspect of the DNA profiling process is the RFLP test, which determines whether a DNA match exists. The RFLP process is firmly rooted in "scientific knowledge," and an application of the four Daubert factors confirms this conclusion. RFLP has been widely tested and has been the subject of a variety of learned articles. Lander & Budowle, supra, at 735. The process is not error-free, but adherence to accepted procedures and controls minimizes this error. See People v. Castro, 144 Misc.2d 956, 545 N.Y.S.2d 985, 993-95 (Sup.Ct.1989) (describing techniques to minimize error). The technique is generally accepted in the scientific community. State v. Cauthron, 846 P.2d at 511 (citing fifteen cases supporting general acceptance of the test). Indeed, we cannot find any recent decision under any standard of admissibility which refuses to admit the DNA match result based on the invalidity or risk of error of the underlying technology. See, e.g., Porter, 618 A.2d at 636 (case law overwhelmingly supports validity of match technology).

We cannot reach the same conclusion about probability statistics if they are generated by the product method. Even the conservatism introduced by fixed-bin analysis does not attempt to account for the dependence of allele frequencies in particular populations. The decisions from other jurisdictions on the use of the product method are mixed. Compare State v. Bible, 175 Ariz. 549, 858 P.2d 1152, 1190 (1993) (product method statistics inadmissible); Lanigan, 641 N.E.2d at 1349-50 (DNA frequency statistics admissible only if the ceiling principle is used); Vandebogart, 616 A.2d at 494-95 (same as Bible); and Cauthron, 846 P.2d at 517, with United States v. Bonds, 12 F.3d 540, 565 (6th Cir.1993) (dispute over statistics goes to weight not admissibility); State v. Pierce, 64 Ohio St.3d 490, 597 N.E.2d 107, 115 (1992) (same as Bonds); and People v. Soto, 30 Cal.App.4th 340, 35 Cal.Rptr.2d 846, 858 (1994) (product rule statistics admissible). We note that the courts that refuse to accept statistics based on the unmodified product method continue to rely on the more narrow Frye standard.

We believe, however, that even under Daubert it is inappropriate to allow evidence based on the unmodified product method. In the lexicon of Daubert, we are concerned that the accuracy of the results cannot be ensured by testing, there is an unknown potential for error, and these calculations are not generally accepted within the scientific community. The endorsement of the ceiling principle by the NRC and more recently by leading advocates in the dispute, including a representative of the FBI, leads us to this conclusion. Lander & Budowle, supra, at 735. Further research may show that the unmodified product rule is reliable, but the current state of scientific development speaks forcefully that we must insist on ceiling principle statistics to fully protect rights of criminal defendants.

We are also influenced by a practical consideration. The debate is not about whether to allow introduction of probability statistics. We agree with the courts that have held that evidence about the fact of one or more allele matches is not helpful without some evidence about the probability of a match in the population as a whole. See, e.g., Vandebogart, 616 A.2d at 494; Cauthron, 846 P.2d at 516. The debate instead is about how conservative the probability statistic will be. Although use of the ceiling principle will increase the likelihood of a random match, the difference is unlikely to have any real effect on jury deliberations. We strongly doubt whether it will make much difference to a jury whether the probability of random selection of a person with the same alleles is 1 in 100,000 or 1 in 1,000,000.

Thus, we find it wholly impractical to say the dispute over statistical measurement should go before the jury for a clash of experts on the proper probability figures. Such evidence will not assist the jury in understanding the evidence or determining a fact in issue as required by Rule 702. It would threaten to become an evidentiary sideshow raising serious Rule 403 concerns. On the one hand, it may inappropriately undercut the legitimate force of DNA match evidence. On the other hand, it may focus too much attention on the significance of the probability statistics. These are exactly the considerations that prompt concerned experts to try to find a truce in the scientific debate. See Lander & Budowle, supra, at 736.

The failure of the unadjusted product method to pass muster under the reliability standards of Daubert does not sound the death knell for the use of DNA profiling in Vermont courts. The use of the ceiling principle will eliminate the objections we have itemized and allow introduction of DNA match statistics consistent with legitimate accuracy concerns of defendants. There is general acceptance within the scientific community that the ceiling principle over-community pensates for any population substructure or allele linkage. See Barney, 10 Cal.Rptr.2d at 745 (ceiling principle points way to common ground). Because the ceiling principle offers an unabashedly conservative estimate, the risk of error is minimal.

C.

The ceiling principle, however, was not used in this case. The trial court admitted the FBI's calculations under the fixed-bin method, which calculated that the odds of a coincidental match were 1 in 50,000. Because admitting this statistic was error, we must evaluate the effect of the evidence and determine whether the error was harmless, or adversely affected defendant's substantial rights. V.R.Cr.P. 52(a); see State v. Lynds, 158 Vt. 37, 42, 605 A.2d 501, 503 (1991).

If the error here were of constitutional dimension, we could find the error harmless only if it is clear beyond a reasonable doubt that the jury would have returned a guilty verdict regardless of the error. Lynds, 158 Vt. at 42, 605 A.2d at 503. We have not created a definitive standard for nonconstitutional errors. See State v. Curavoo, 156 Vt. 72, 76-77, 587 A.2d 963, 965-66 (1991) ("Vermont harmless-error rule appears in need of clarification"). We conclude that we do not have to adopt a definitive standard to decide this case.

Whatever the standard, the factors that guide our inquiry are the same. Lynds, 158 Vt. at 42, 605 A.2d at 503. These include the importance of the evidence in the prosecution's case, whether the evidence was cumulative, the extent of cross-examination, and the overall strength of the prosecution's case. Id. We start by emphasizing the nature of the error in this case. The error is not that the State's witness declared a DNA match, nor even that he offered probability statistics on whether the DNA sample found on the victim came from someone other than defendant. Instead, the error is that the probability of a coincidental match, placed at 1 in 50,000 by the witness, is too low under the ceiling principle. See Vandebogart, 616 A.2d at 495.^[4] Although we do not know the exact number that would be produced by the ceiling principle, we understand that it would be in the range of 1 in 1000 to 1 in 5000.

The disputed DNA statistics were certainly not the most incriminating aspects of the State's case. Other scientific tests were performed which linked defendant to the crime scene. Samples of his pubic hair matched those found on the victim's bedsheet. Secretor status tests revealed that defendant's secretor status was consistent with the blood type and secretor status found in the semen on the victim's underwear. The victim conclusively identified defendant's voice as that of her assailant from a voice line-up. She identified her attacker as a six-foot-tall man with brownish hair parted in the middle and feathered down the sides; defendant matched this description. She also identified rings that defendant gave his girlfriend as those that were stolen from her when she was attacked. Finally, the victim testified that her assailant had threatened her with a.45 caliber semiautomatic pistol, and investigators retrieved this type of gun from someone who had purchased the gun from defendant.

In the wake of all of this evidence, the admission of the DNA profile statistic was cumulative in nature and not very significant. See Delaware v. Van Arsdall, 475 U.S. 673, 684, 106 S.Ct. 1431, 1438, 89 L.Ed.2d 674 (1986). We recognize that DNA statistics can have a powerful impact on a jury. Where other identification evidence is overwhelming, however, other courts have found beyond a reasonable doubt that the erroneous admission of probability statistics was harmless. See Bible, 858 P.2d at 1191; Barney, 10 Cal.Rptr.2d at 747-48; Nelson v. State, 628 A.2d 69, 77 (Del.1993). On this basis, we find the error in the admission of the product method probability statistics to be harmless error.

D.

The last DNA issue we must address concerns whether the FBI used proper procedures in performing the RFLP test. We recognize the sensitivity of the RFLP test, and note that even minor errors in the laboratory can have a dramatic impact on the results. See Vandebogart, 616 A.2d at 491 (referring to courts excluding DNA evidence because laboratory failed to follow established procedures). Indeed, additional problems exist when the test is used for forensic purposes because the quality of forensic samples may be adversely affected by weather, time, or insufficient quantity. See Commonwealth v. Curnin, 565 N.E.2d at 441. Nonetheless, defendant's arguments on this issue are not persuasive.

Defendant contests the fact that the FBI did not repeat the RFLP process to verify its conclusions. We find nothing in the record or the relevant literature to suggest that RFLP is not reliable unless the same results are obtained twice. In addition, defendant challenges the FBI's use of cross-gel analysis whereby his known sample on one autorad was compared to the evidence sample on another autorad. Cross-gel analysis was employed because the results from defendant's known sample on the initial autorad were inconclusive. It is important to note that the inconclusive results were on defendant's known sample, not the sample obtained from the victim and the crime scene, and that adequate controls were in place which allowed for this type of cross-gel comparison. The trial judge carefully considered the evidence on the reliability of results based on cross-gel analysis, and ruled the results admissible. We think the ruling was within his discretion.

III.

Defendant contests the trial court's admission of the blood-type evidence of Mark Rouelle, the person defendant theorized had committed the crime. He maintains that his due process rights were violated because the trial court admitted the evidence contrary to its July 10th pretrial ruling, adversely affecting his trial strategy.

Defendant misinterprets the trial court's pretrial ruling as a fixed prohibition against the introduction of the blood-type evidence. The court instructed the prosecution that the ruling could be revisited depending upon what transpired at trial. It stated that the prosecution could not mention the blood-type evidence in its opening statement or its case-in-chief, but it specifically left open the question of whether the prosecution could use the evidence on rebuttal. The prosecution honored these instructions, and with the court's permission, introduced the evidence on rebuttal. Although defendant may have misinterpreted the court's instructions, his misunderstanding does not create reversible error.

Even assuming that the pretrial ruling specifically precluded the admission of the blood-type evidence, we have held that pretrial rulings are provisional, and are subject to later modification. See State v. Bruno, 157 Vt. 6, 8, 595 A.2d 272, 273 (1991). Thus, a ruling on a pretrial motion to exclude evidence is tentative and subject to revision at trial. See id. (citing State v. Blondin, 128 Vt. 613, 617, 270 A.2d 165, 167 (1970)). In light of these decisions, defendant's reliance on the pretrial ruling to determine his trial strategy was imprudent. Because he could not legitimately expect to rely on the ruling, no due process violation exists.

Defendant also asserts that the trial court's decision to admit the blood-type evidence of Mark Rouelle was error. The blood-type evidence was offered to rebut defendant's theory that Rouelle committed the crime. It is well within the trial court's discretion to admit evidence that rebutted a theory aggressively advanced by defendant. See State v. Jones, 160 Vt. 440, 446, 631 A.2d 840, 846 (1993) (trial court has wide discretion over evidentiary rulings, including the admission of rebuttal testimony).

IV.

Defendant next contends that the State violated V.R.Cr.P. 16(a)(1) and V.R.Cr.P. 16.2(b) by failing to give the defense timely notice of the Rouelle blood- and hair-analysis report. The State provided the defense with this report on July 9th, the day before trial.

V.R.Cr.P. 16(a)(1) imposes a duty on the State to disclose to the defendant "as soon as possible the names and addresses of all witnesses then known to [it]." V.R.Cr.P. 16.2(b) further provides that the State has a continuing duty to disclose relevant material discovered after initial compliance with a discovery request. The purpose of these rules is to prevent the State from assuming an unfair advantage over the defense. See State v. Percy, 149 Vt. 623, 636, 548 A.2d 408, 416 (1988). To establish reversible error, however, defendant must show both a violation of the rule and resulting prejudice. Jones, 160 Vt. at 446, 631 A.2d at 845 (citing State v. Sird, 148 Vt. 35, 39, 528 A.2d 1114, 1116 (1987)); State v. Heath, ___ Vt. ___, ___, 649 A.2d 243, 244 (1994). Defendant has shown neither.

Nothing suggests that the State gained an unfair advantage by its untimely delivery of the Rouelle report. Indeed, at least two years prior to trial, the State provided defendant with test results showing the blood-type and secretor status of the victim and defendant. Defendant should have anticipated that similar blood-type evidence would be taken from Rouelle once he proposed that Rouelle was the real perpetrator. Furthermore, the State's explanation for the delay in producing the Rouelle report was reasonable. The State did not learn of the defense plan to name Rouelle as the perpetrator until May 20, 1991. Testing of Rouelle was also delayed by his competency determination and frequent change of counsel. There is no evidence that the State deliberately withheld the reports or acted in bad faith.

Moreover, the only objection defendant's counsel raised below concerned his inability to confer with his expert about the test results given the short notice. The trial court addressed this problem when it precluded the State from using any of the contested evidence in its opening statement or case-in-chief. The record shows that the blood-type evidence of Mark Rouelle was not introduced until late in the trial, and defense counsel conceded that his expert would be able to review the reports by that time.

Even if the expert could not have reviewed the report, defendant has failed to show prejudice. He makes no claim how the expert review would have affected his defense strategy or made it more viable. His only claim is that the defense would have been stronger without the evidence, a claim that may be true but is not responsive to the asserted error.

Defendant also asserts that the State's failure to provide timely notice of the Rouelle report violated Article 10 of the Vermont Constitution and the Sixth Amendment to the United States Constitution. For the reasons explained above, no constitutional violations exist.

Defendant argues further that the State's failure to disclose the Rouelle report in a timely fashion required the trial court to grant his motion for a continuance. The granting of a continuance by the trial court is a matter of discretion. State v. Bailey, 144 Vt. 86, 93, 475 A.2d 1045, 1049 (1984). This Court will not interfere with a discretionary action of the trial court as long as a reasonable basis exists for the court's decision. Brooks v. Brooks, 131 Vt. 86, 93, 300 A.2d 531, 535 (1973). To support a claim of error, defendant must show that the trial court's decision was untenable or clearly unreasonable. State v. Ahern, 137 Vt. 253, 267, 403 A.2d 696, 705 (1979). He fails to make this showing.

The trial court fully considered defendant's need to have his expert review the Rouelle report, and specifically postponed the State's introduction of this evidence to provide defendant with an opportunity for his expert to review the report. This approach is consistent with our previous cases. See State v. Burnham, 145 Vt. 161, 168-69, 484 A.2d 918, 923 (1984) (no abuse of discretion where trial court declared recess to permit defendant opportunity to interview witness not disclosed by State); State v. Connarn, 138 Vt. 270, 271, 413 A.2d 812, 814 (1980) (no abuse of discretion where trial court gave defendant opportunity to depose witness not disclosed by State). Thus, there was no prejudice to defendant from the denial of the continuance. Compare Heath, ___ Vt. at ___, 649 A.2d at 244. Accordingly, we can not hold that the trial court's refusal to grant a continuance was an abuse of discretion.

V.

Defendant next challenges the trial court's refusal to allow into evidence the statutory standards under V.R.Cr.P. 41.1(c) for obtaining a nontestimonial identification order (NTO).^[5] He argues that because an NTO was issued against Mark Rouelle, the requirements for such orders are relevant to support his theory that Rouelle committed the crime. Thus, he asserts the trial court violated V.R.E. 402 (relevant evidence is admissible unless otherwise excluded) by excluding this evidence.

V.R.E. 401 defines evidence as relevant if it has "any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence." Despite this broad rule of relevancy, the decision to admit evidence lies within the sound discretion of the trial court, and will not be overturned on appeal absent a showing of an abuse of discretion. See State v. Larose, 150 Vt. 363, 371, 554 A.2d 227, 233 (1988). In this case, the fact that the trial judge found that the NTO standards were met with respect to Rouelle adds nothing to the jury's task of determining the identity of the victim's assailant. Moreover, introduction of the NTO standards might mislead the jury into thinking that the court's action somehow bound their determination. See V.R.E. 403. The trial court's decision to exclude this information was a proper exercise of its discretion.

VI.

Defendant argues that the trial judge's misrepresentation to potential jurors of the reason for the presence of television cameras in the courtroom warranted a mistrial because this deception destroyed the jury's confidence and trust in the proceedings. He further contends that the trial judge erred by failing to either examine the jury regarding this matter or issue a curative instruction. Our review of the record does not support these arguments.

We agree that the trial judge's explanation to the jury pool about the presence of television cameras was inaccurate. As defendant points out, prospective jurors complained to the Judicial Conduct Board about the deception, demonstrating their negative reactions. There is no evidence, however, that the empaneled jurors were negatively affected by the deception. Indeed, the empaneled jury was questioned about whether it had read or heard anything in the media regarding the trial; only one juror responded affirmatively, and he was excused from the panel. In addition, the judge instructed the empaneled jury that the trial was going to be televised. He inquired whether this filming would interfere with any juror's ability to work in an unbiased, impartial manner; none indicated that the filming would interfere with their work.

We agree with defendant that he is entitled to an impartial jury free from any suspicious taint by outside influences, and that he need show only that an irregularity had the capacity to influence the result of the trial, not that it actually did so. See State v. Griffin, 152 Vt. 41, 45, 563 A.2d 642, 645 (1989). He failed, however, to show any irregularities or outside influences on the empaneled jurors. Accordingly, the trial court's denial of his motion for a mistrial was proper.

VII.

Defendant claims that the trial court improperly instructed the jury on the State's burden of proof. There is no error if the jury charge as a whole conveys the "`true spirit and doctrine of the law, and there is no fair ground to say the jury has been misled by it.'" State v. Benoit, 158 Vt. 359, 362, 609 A.2d 230, 232 (1992) (quoting State v. Williams, 154 Vt. 76, 81, 574 A.2d 1264, 1267 (1990)). Moreover, because defendant failed to make a timely objection to the language he now challenges, the verdict may be overturned only for plain error. V.R.Cr.P. 52(b); State v. Welch, 136 Vt. 442, 444, 394 A.2d 1115, 1116 (1978). Plain error will be found only in rare and extraordinary cases where the error is obvious and strikes at the heart of defendant's constitutional rights or results in a miscarriage of justice. See State v. Ross, 152 Vt. 462, 469, 568 A.2d 335, 339 (1989). Defendant fails to meet this standard.

Defendant bases his claim on two isolated phrases from the trial court's lengthy charge to the jury. These phrases note that the jury's role is to "seek the truth." Relying on State v. Giroux, 151 Vt. 361, 365, 561 A.2d 403, 406 (1989), he maintains that the inclusion of these phrases caused the jury to lose sight of the prosecution's burden to prove defendant's guilt beyond a reasonable doubt. When we look at the entirety of the charge, however, we find that the court repeatedly instructed that the State's burden was proof beyond a reasonable doubt. Taken as a whole, the court's charge correctly conveyed to the jury the State's burden of proof. There is no plain error.

VIII.

Finally, defendant argues that the trial court exhibited bias and prejudice toward him and defense counsel during pretrial and trial proceedings. He lists twelve incidents that he claims individually and collectively demonstrate the court's bias and prejudice. We have already addressed several of these incidents in our discussion above, and determined that the court's actions were proper; thus, they require no further discussion here.^[6] Defendant's remaining examples of judicial bias are as follows: (1) imposition of a thirty-minute limit on defense counsel's cross-examination on rebuttal of a State witness during a hearing on a pretrial motion; (2) attempts to "coerce" defense counsel to waive his client's Fifth Amendment privilege; (3) reference to one of defense counsel's statements as "fantasy"; and (4) the grant of the State's request for an NTO of Mark Rouelle.

Although we address each of these claims, we emphasize that rulings adverse to the defense, even if erroneous, do not by themselves demonstrate that the judge was biased. It is incumbent on defendant to make some showing of improper motivation by the rulings, a showing he has not made here.

Defendant first contends that the trial court impermissibly limited defense counsel's cross-examination of a State witness during a hearing on a pretrial motion. Rule 611(a) of the Vermont Rules of Evidence provides that the trial court has the authority to set reasonable limits on the consumption of time in examining witnesses. These time limits must be reasonable and sufficiently flexible to ensure that important evidence is not excluded due to artificial time constraints. See Varnum v. Varnum, 155 Vt. 376, 390, 586 A.2d 1107, 1115 (1990). Here, there is no indication that the limit on the length of the witness's cross-examination was unreasonable or that important evidence was excluded.

Defendant also asserts that the trial court tried to "coerce" defense counsel to waive his client's Fifth Amendment privilege, and that the court referred to one of defense counsel's statements as "fantasy." These incidents stem from the televised coverage of the trial, and relate to discussions between the parties concerning what type of immunity would be extended to defendant in his interviews with television reporters. The prosecution and defense counsel could not agree, and the trial court's frustration over the issue is apparent from the transcript. Defendant is wrong to characterize the court's impatience as an attempt to coerce defense counsel into waiving any of his client's rights. Indeed, the court specifically stated that the decision whether to participate in the television interviews ultimately rested with defendant. The court's reference to "fantasy" was in response to an example defense counsel offered to further explain his position. The court noted that the example was not comparable to the situation at hand, not that defendant's Fifth Amendment concerns were fantasy. The court's comments, when considered in their proper context, do not show bias or prejudice.

Defendant also maintains that the trial court's bias was exhibited by granting the State's request for an NTO of Mark Rouelle, but he fails to provide any reasons or explanation why this ruling was biased. Given that the State's request for the NTO was prompted by defendant's claim that Rouelle had committed the crime, the court's decision was reasonable. Defendant invites us to find that the court was biased merely because it ruled in favor of the other side; for obvious reasons, we decline the invitation.

Finally, defendant argues that the cumulative effect of all these incidents constitutes judicial bias. Because we have not found any instance of judicial bias, no cumulative effect exists. See State v. Wheel, 155 Vt. 587, 612, 587 A.2d 933, 948 (1990).

Affirmed.

NOTES

[1] This Court has previously considered the use of DNA evidence to show that a defendant was not the perpetrator of a crime. State v. Passino, 161 Vt. 515, 640 A.2d 547 (1994). Because the use of DNA evidence for exculpatory purposes does not involve the same issues as when it is used for inculpatory purposes, Passino is not relevant to this analysis.

[2] For a more extensive discussion of these two disciplines see United States v. Jakobetz, 747 F.Supp. 250, 251-54 (D.Vt.1990), State v. Cauthron, 120 Wash.2d 879, 846 P.2d 502, 508-15 (1993); State v. Vandebogart, 136 N.H. 365, 616 A.2d 483, 486-89 (1992); Commonwealth v. Curnin, 409 Mass. 218, 565 N.E.2d 440, 445-48 (1991); Annotation, Admissibility of DNA Identification Evidence, 84 A.L.R.4th 313 (1991); Note, The Dark Side of DNA Profiling: Unreliable Scientific Evidence Meets the Criminal Defendant, 42 Stan.L.Rev. 465 (1990).

[3] The Court also emphasized that the inquiry regarding the admissibility of scientific evidence cannot be divorced from the other rules of evidence. Daubert, ___ U.S. at ___, 113 S.Ct. at 2979. The Court noted that Rules 703 (otherwise inadmissible hearsay permitted if it normally relied upon by experts in the field), 706 (court may procure assistance of expert of its own choosing), and 403 (relevant evidence excluded if probative value outweighed by unfair prejudice) are also relevant to an admissibility inquiry.

[4] The Supreme Court of New Hampshire remanded the case to the trial court to determine whether use of ceiling principle statistics would have made any difference in the outcome of the criminal trial. Vandebogart, 616 A.2d at 495. Although we believe it is unnecessary to remand in this case to assess the effect of the error, we do not rule out such a procedure in other cases.

[5] V.R.Cr.P. 41.1(c) provides:

An order shall issue only on an affidavit or affidavits sworn to before the judicial officer and establishing the following grounds for the order:

(1) that there is probable cause to believe that an offense has been committed;

(2) that there are reasonable grounds, that need not amount to probable cause to arrest, to suspect that the person named or described in the affidavit committed the offense; and

(3) that the results of specific nontestimonial identification procedures will be of material aid in determining whether the person named in the affidavit committed the offense.

[6] The alleged incidents of bias or prejudice which we have already discussed include: (1) the admission of the Rouelle report at trial despite the court's pretrial ruling, (2) the trial court's denial of defendant's motion for a continuance, (3) the exclusion of the NTO standards, and (4) the denial of the mistrial motion based on the trial judge's misrepresentations to prospective jurors. Defendant attempts to squeeze several different arguments out of the trial court's decision regarding the inclusion of the Rouelle report (i.e., that the court's evidentiary rulings were motivated by the State's needs, refusing to issue clear and concise rulings regarding the Rouelle report evidence, and forcing defense counsel to present Rouelle as a witness after it rested). We are unpersuaded by this effort. Our earlier discussion adequately addresses any concerns defendant has regarding the trial court's treatment of the Rouelle report.