State v. Hummert

183 Ariz. 484 (1994) 905 P.2d 493

STATE of Arizona, Appellee,
v.
Steven Henry HUMMERT, Appellant.

No. 1 CA-CR 92-098.

Court of Appeals of Arizona, Division 1, Department B.

July 26, 1994. Reconsideration Denied December 12, 1994. Review Granted and Denied November 21, 1995.

Grant Woods, Atty. Gen. by Paul J. McMurdie, Chief Counsel, Crim. Div., and Randall M. Howe, Asst. Atty. Gen., Phoenix, for appellee.

Alex D. Gonzalez, Mesa, for appellant.

Review Granted on issues A and B and Denied on other issues November 21, 1995.

OPINION

TOCI, Judge.

After a jury trial, Steven Henry Hummert ("defendant") appeals from his convictions and sentences for two counts of sexual assault and other related offenses. The primary issue defendant raises on appeal is whether the trial court erred in admitting expert testimony that a declared "match" of DNA samples uniquely identified him as the assailant. We conclude that in the absence of generally accepted population frequency statistics for calculating the probability of a random match of DNA samples, the trial court erred in admitting such testimony. Because we do not find the error harmless, we reverse.

We resolve the remaining issues raised by defendant as follows: (1) evidence of a prior incident in which defendant followed another female was properly admitted by the trial court to prove defendant's identity as the assailant in this case; (2) the trial court erred in precluding defendant from introducing "reverse 404(b) evidence" that excluded defendant as the assailant in a prior sexual assault similar to the attack on the victim; (3) the evidence of defendant's in-court identification was not unduly suggestive under State v. Dessureault, 104 Ariz. 380, 453 P.2d 951 (1969); (4) the trial court did not err in denying defendant's motion to remand to the grand jury for a new determination of probable cause; and (5) the trial court did not err by giving an instruction that excluded "possible doubt" from the definition of reasonable doubt.

I. FACTUAL BACKGROUND

We view the facts at trial in the light most favorable to sustaining the verdicts below. State v. Atwood, 171 Ariz. 576, 596, 832 P.2d 593, 613 (1992), cert. denied, 506 U.S. 1084, 113 S. Ct. 1058, 122 L. Ed. 2d 364 (1993).

At approximately 3:30 a.m. on July 16, 1989, M,^[1] a nineteen-year-old woman, was assaulted in Tempe, Arizona. Upon returning home, M was surprised by a male assailant as she got out of her car. He put a handgun to her head and forced her into the yard of a nearby house. He then made her partially disrobe, fondled her breasts, and committed two separate acts of sexual assault. Afterwards, while the assailant was attempting to choke her, M bit him on the forearm. The assailant then struck M on the back of the head, causing severe scalp lacerations that required surgery. After warning her not to move, the assailant drove away in a car that M had seen parked nearby. After the assailant departed, M ran to her home, and her family called the police.

During a police interview, M described the assault and gave investigating officers a general description of her assailant and his vehicle. She described his car as a red Honda CRX with a grey out-of-state license plate with black lettering on it, bearing the numbers 939. She also remembered seeing an emblem on the rear bumper shaped like a map of Texas. M and a friend who had accompanied her earlier both said that, after leaving a Tempe night club, they saw a red Honda CRX driving the streets of Tempe shortly before the assault. M said that she later saw the CRX parked near her house as her attacker forced her into the neighboring yard where the assault occurred.

The following morning, after M's cousin spotted a red Honda CRX in the parking lot of a fast food restaurant near M's house, police linked defendant to the assault. Police observed that the car and license plate matched M's description. They later learned that defendant, a manager at the restaurant, was the owner of the car. Defendant told the officers that he had been with his co-workers at a party near the restaurant until 4:00 a.m. that morning. To corroborate this alibi, he produced a receipt showing that he had purchased gas from a Mesa service station at 4:20 a.m. When the officers asked defendant how he had received small scratch marks and a puncture wound on his forearm, he explained that he had burned his arm while cooking with hot grease at the restaurant.

Police contacted the three co-workers who hosted the party defendant attended the night of the assault. Each co-worker confirmed that defendant had left the party at about 4:00 a.m. At trial, however, these witnesses testified that defendant approached them before the police interview and suggested that they tell the police that he left the party at 4:00 a.m. Two of the witnesses recalled that defendant had actually departed much earlier, between 2:00 and 2:30 a.m.

At trial, the state also introduced evidence of a prior incident to establish defendant's identity as M's assailant. The victim of that incident, S, testified that, early on the morning of February 22, 1989, she was followed by a man driving a red Honda CRX as she drove through Chandler on her way home. The man followed S to a hotel, where she asked a security guard to call the police. A Chandler police officer testified that she stopped a CRX that matched the description of the reported vehicle and spoke with defendant, who was driving. Defendant denied following S, and told the officer that he was driving the streets of Chandler after work in order to familiarize himself with his new neighborhood.

The state also introduced testimony regarding evidence recovered from clothing worn by M at the time of the assault. A criminalist testified that one of four pubic hairs recovered from M's underpants matched all of the characteristics of a known sample of defendant's pubic hair. Blood group testing on questioned blood and semen samples was inconclusive in identifying defendant as the source. DNA testing of semen found on M's underpants, however, revealed that DNA on the questioned sample matched a known sample of defendant's DNA.

Defendant testified at trial on his own behalf. He denied following S or committing the assaults against M. He further denied that he asked his co-workers to lie about the time that he left the party. Defendant contended that his co-workers were subtly pressured by their employer to cooperate with the police. As a result, according to defendant, the witnesses changed their testimony in order to implicate defendant.

M also testified at trial. She made an in-court identification of defendant as her assailant. The defense, however, impeached M with evidence that two days after the assault she could not identify defendant's picture in a pre-trial photographic lineup.

The jury found defendant guilty of all the offenses charged in the indictment: (1) two counts of sexual assault, both class 2 felonies; (2) one count of kidnapping, a class 2 felony; (3) two counts of aggravated assault, all class 3 felonies; and (4) one count of sexual abuse, a class 3 felony. The jury also found that each offense was dangerous. The trial court than determined that defendant had a prior felony conviction in Texas for theft and that the offenses in this matter were committed while defendant was on probation for the Texas conviction.

Pursuant to Ariz. Rev. Stat. Ann. ("A.R.S.") section 13-604.02 (1989), the trial court imposed concurrent sentences of life imprisonment on each count. The trial court credited defendant with 198 days of presentence incarceration. The trial court also ordered defendant to pay $8,110.97 in restitution, a $600.00 felony penalty assessment ($100.00 per felony), and an $8.00 time payment fee. This appeal followed.

II. DISCUSSION

A. Admission of Evidence Regarding DNA "Match"

1. Background Regarding DNA Analysis

DNA testing relies upon the fundamental proposition that because any two human genomes differ at about 3 million sites, no two persons (except for identical twins) have the same DNA sequence. National Research Council, DNA Technology in Forensic Science 9 (National Academy Press 1992). Consequently, if enough sites of variation are examined, unique identification with DNA typing is theoretically possible. Id. The DNA typing systems used today, however, examine only a few sites of variation and have only limited ability for measuring the variability at each site. Id. Thus, a chance exists that two different persons have DNA patterns that match at the small number of sites examined. Id.

Nevertheless, even with existing technology, a match between two DNA patterns is strong evidence that the samples came from the same person. Id. But interpreting a DNA typing analysis requires a valid scientific method for estimating the probability that a random person by chance matches the sample at the sites of DNA variation examined. Id. "To say that two patterns match, without providing any scientifically valid estimate (or, at least, an upper bound) of the frequency with which such matches might occur by chance, is meaningless." Id.

In the DNA analysis employed in this case, the Federal Bureau of Investigation ("FBI") used a technique known as restriction fragment length polymorphism ("RFLP") analysis. The testing involved comparisons of a known sample of defendant's blood and a sample of semen found on M's underpants. The RFLP test does not detect the entire DNA strand. It focuses on discrete, limited segments of DNA detected by probes used in the test. And, because only 0.1 to 0.3 percent of approximately 3 billion positions on the human genome are variable, it is important that the probes used detect these variable or polymorphic sites, rather than monomorphic sites, which are the same for every human being. Washington v. Cauthron, 120 Wash.2d 879, 846 P.2d 502, 512-13 (1993) (citations omitted).

Each version of a polymorphic gene is known as an "allele." Id., 846 P.2d at 509. Ultimately, the matching process involves the visual comparison of bands on alleles, derived from known and questioned samples, as depicted on autoradiographs, or "autorads." In determining the probability that a defendant's DNA is the same as the sample taken from the crime scene, the expert relies on a previously constructed database. Id. at 513. The database is developed to test the probes used by the particular lab. In the absence of the database, one cannot determine if a probe is detecting an allele that is common to all human beings, or a hypervariable site on the DNA. Id. The more variations that a given site has, the more useful it is as an investigative tool. Id.

An autorad match is meaningless without the statistical evidence to validate the match. If the autorad reflects only sites on the DNA that are common to all human beings (monomorphic sites), the evidence obtained cannot be the basis for identifying the defendant. Id. Thus, the expert must also show that the alleles detected by the particular probes used are polymorphic. Id. This showing requires that a sufficient database — a large enough and truly random sample of the general population — serve as the foundation for the expert's conclusions. Id.

To make claims about probabilities that are of the high magnitudes usually seen in DNA cases, experts rely on the product rule. The scientist first collects the data pertinent to the allele being compared. Based on the statistics derived from the database, the expert determines the probability that a sample of known origin will match (on an autorad) the given sample. That probability will vary from one allele to another. Id.

Suppose, for example, that a pair of DNA samples match on two bands and that allele "A" may be found in 10 percent of the population and allele "B" in 50 percent of the population. Applying the product rule, an expert would conclude that the probability of a coincidental match on both alleles is 0.10 times 0.50 equals 0.05, or a five-percent probability. See State v. Bible, 175 Ariz. 549, 583, 858 P.2d 1152, 1180 (1993), cert. denied, ___ U.S. ___, 114 S. Ct. 1578, 128 L. Ed. 2d 221 (1994). This means that a one in twenty probability of a random match exists, leaving a nineteen in twenty chance that the samples came from the same person. Id.

Consequently, the frequency figures are vital components of the product rule. The problem is that the database for the frequency figures must be large enough to be statistically significant. Id. Any errors or shortcomings in the database may have a "profound and significant impact on the random match calculations." Id. This is one of the reasons our supreme court held in Bible that use of the "product rule" and "the resulting opinion of the odds against a random match were not derived by applying generally accepted scientific theory." Id. at 586, 858 P.2d at 1183.

2. Frye Hearing

Before trial, the trial court held a hearing to determine whether the DNA evidence met the standards of Frye v. United States, 293 F. 1013 (D.C. Cir.1923), which remains the test for the admission of new scientific evidence in Arizona. Bible, 175 Ariz. at 578-80, 858 P.2d at 1181-83. The parties presented the live testimony of four expert witnesses, deposition testimony of two FBI experts, and excerpts from testimony given in other cases involving litigation over the admissibility of DNA evidence.

After the hearing, the trial court issued a comprehensive fifteen-page order. The trial court found that the underlying theory of DNA testing met the Frye test of general acceptance in the relevant scientific community. The trial court further found that the procedures used by the FBI in comparing questioned and known DNA samples met the Frye standards. Accordingly, the trial court found admissible the FBI's conclusion that the DNA in the sperm sample found in M's underpants "matched" the DNA in defendant's blood sample.

The trial court also ordered, however, that evidence about the statistical significance of the match, as calculated from FBI databases, was inadmissible. The FBI calculated the probability that the match involving defendant occurred randomly was one in one million. The trial judge found that substantial disagreement existed in the scientific community about the FBI's method for determining the statistical probability that a match occurred randomly. The trial court further found that admitting evidence of the statistical probability of a random match would have an "overwhelming impact" on the jury and was therefore unfairly prejudicial under Rule 403, Arizona Rules of Evidence.

3. Issue on Appeal

While this matter was on appeal, our supreme court examined the DNA testing procedures of Cellmark Diagnostics Laboratory ("Cellmark"). Bible, 175 Ariz. at 590, 858 P.2d at 1193. The Bible court determined: (1) that the theory underlying DNA testing met Frye standards; (2) that Cellmark's criteria in declaring a match between questioned and known samples also satisfied Frye; and (3) that Cellmark's random match probability calculations were inadmissible at trial because they lacked general acceptance in the relevant scientific community. Id.

Although the DNA testing procedures and the random match probability calculations in this case were performed by the FBI, and thus fell outside the literal holding of Bible, neither party challenges the trial court's rulings on these issues. Instead, this appeal raises an issue expressly left open in Bible. That is, because Bible holds that random match probability calculations are inadmissible on Frye grounds, is testimony regarding the existence of a "match" between questioned and known DNA samples admissible?

In resolving this question, the trial court stated:

Testimony of a "match," the criteria for declaring a "match," the fact that the results show that a defendant cannot be excluded as the assailant, the fact of the uniqueness of DNA, witnesses' experience in finding random matches under the same circumstances, and related testimony would appear to be relevant and not create an undue danger of unfair prejudice to the defendants.

In deciding this issue, we must resolve two questions: (1) was any testimony regarding the declaration of a DNA match and its significance admissible, and (2) was the evidence introduced in this case proper? Defendant argues that all testimony regarding DNA testing should have been excluded because the evidence was "meaningless" in absence of the random match probability statistics that were excluded under Frye. The state disagrees, arguing in part that "[e]vidence of inclusion, as a DNA match is, is routinely admitted" in connection with other types of forensic analysis.

The state's proposition is not novel. Bible supports the conclusion that experts can properly testify that the declaration of a DNA match means that defendant could not be excluded as the contributor of the sample. There, our supreme court stated that evidence of Cellmark's match criteria was admissible, adding:

We emphasize, however, what this means and what it does not mean. If testing shows that samples do not match, then the conclusion is that they are from different individuals. If testing shows that the samples do match, the conclusion is that they may be from the same individual.

Bible, 175 Ariz. at 590, 858 P.2d at 1193. See also Minnesota v. Alt, 504 N.W.2d 38, 52-54 (Minn. Ct. App.), aff'd, 505 N.W.2d 72 (Minn. 1993) (finding that trial court properly limited testimony of FBI experts to statement that DNA testing did not exclude defendant as source).

Here, however, the state's experts concluded that a match of the known sample of defendant's DNA with the semen sample taken from M's underpants had far more significance than simply failing to exclude the defendant. Lawrence Pressley, the FBI's DNA expert, testified that from a comparison of autorads derived from three different probes, he found a match between the questioned sample and defendant's known sample. He further testified that the possibility of a random match was "rare" and that a match meant that "[e]ither you're brothers, identical twins, or that would be a very unique experience." The state's other trial expert, Mary Claire King, a professor of epidemiology at the University of California, reached a similar conclusion. She testified that by using RFLP analysis, "one can, by carefully choosing particular parts of the DNA that vary a lot between people, uniquely identify every person with just a sample of each person's DNA." Both experts also essentially testified that a "match" over three probes has always meant that the compared samples came from the same individual or from an identical twin.

This testimony should not have been admitted. In the absence of generally accepted population frequency statistics for determining the probability of a random match, the experts' opinions overstated the significance of the DNA test results. By informing the jury that a match over three probes was a "rare" event or that a DNA match could "uniquely identify" an individual, the experts effectively conveyed to the jury, at least implicitly, the random match probability statistics found inadmissible by the trial court.^[2] Moreover, as defendant argued below, cross-examination of the experts about the validity of their assumptions underlying the population frequency statistics was practically impossible without resorting to the precluded random match probability statistics. Arguably, estimates of the probability of a random match, even of one in one million, are less damaging to defendant than the experts' suggestions that the matches in this case were conclusive of identity. Thus, the trial court erred in admitting the state's expert testimony about the significance of the declared DNA match.

Our conclusion that this evidence was improperly admitted is supported by recent precedent from other jurisdictions. See California v. Barney, 8 Cal. App. 4th 798, 10 Cal. Rptr. 2d 731 (1992); Nelson v. Delaware, 628 A.2d 69 (Del. 1993); Massachusetts v. Lanigan, 413 Mass. 154, 596 N.E.2d 311 (1992); Massachusetts v. Curnin, 409 Mass. 218, 565 N.E.2d 440 (1991); New Hampshire v. Vandebogart, 136 N.H. 365, 616 A.2d 483 (1992); New Mexico v. Anderson, 115 N.M. 433, 853 P.2d 135 (Ct.App.), cert. granted, 848 P.2d 531 (1993); Cauthron, 846 P.2d 502. In these cases, the courts found that all DNA evidence, including the declaration of a match, is inadmissible in the absence of generally accepted population frequency statistics. Cf. United States v. Porter, 618 A.2d 629, 640-42 (D.C. 1992) (holding DNA evidence inadmissible in the absence of valid probability calculations, but remanding for trial court to consider whether consensus existed for a "conservative" statement of the probability of a coincidental match); Minnesota v. Bloom, 516 N.W.2d 159 (Minn. 1994) (statistical probability evidence derived by using the National Research Counsel's "limited ceiling principle" provided sufficient evidentiary foundation to allow expert opinion evidence regarding random match probability); but see Pennsylvania v. Crews, 536 Pa. 508, 640 A.2d 395 (1994) (trial court properly refused to admit statistical information regarding match probability; expert testimony, however, that the match of three out of four loci made it more probable than not that the sample was defendant's did not violate Frye and was properly admitted).

The issue was squarely addressed in Cauthron. There, as in this case, the state introduced no evidence regarding the statistical significance of the declared DNA match. Nevertheless, expert witnesses testified that the match conclusively established that the defendant was the source of known and questioned samples. In finding such testimony inadmissible, the Washington Supreme Court relied on prior cases and on the National Research Council's Report, DNA Technology in Forensic Science. The Washington court stated:

The [National Research Council's] views support the conclusion reached in the courts:

To say that two patterns match, without providing any scientifically valid estimate (or, at least an upper bound) of the frequency with which such matches might occur by chance, is meaningless.

Because the testimony presented did not include the background probability information, it was insufficient. The [National Research Council] recommends that: "[r]egardless of the calculated frequency, an expert should-given with the relatively small number of loci used and the available population data-avoid assertions in court that a particular genotype is unique in the population." Testimony of a match in DNA samples, without the statistical background or probability estimates, is neither based on a generally accepted scientific theory nor helpful to the trier of fact.

Cauthron, 846 P.2d at 516 (citations omitted); accord Alt, 504 N.W.2d at 52-53. Thus, Cauthron directly supports our conclusion in this case.

The state cites Minnesota v. Johnson, 498 N.W.2d 10 (Minn. 1993) and Delaware v. Pennell, 584 A.2d 513, 515 (Del. Super. Ct. 1989), in support of the proposition that, even in the absence of admissible population frequency statistics, the court did not err in admitting evidence that a match uniquely identifies an individual. These cases do not support the state's argument. In Johnson, 498 N.W.2d at 14-15, the Minnesota Supreme Court upheld the admission of testimony that alleles were present in at least three to twenty-seven percent of the samples in the FBI database. The court did so, however, in light of a determination that such testimony met Frye standards. Id. The court did not endorse expert opinion that a match of alleles indicated they came from a unique source. Similarly, in Pennell, 584 A.2d at 519, the Delaware Superior Court permitted testimony regarding a DNA match, but it has since been overruled by the Delaware Supreme Court. See Nelson, 628 A.2d at 75 ("We hold that DNA matching evidence is inadmissible in the absence of statistical interpretation of the significance of the declared match.").

4. Harmless Error Analysis

We begin our harmless error analysis by observing that, even in absence of the DNA evidence, there is strong evidence of defendant's guilt. M saw defendant's automobile at the scene of the assault and identified the first three numbers on the vehicle's license plate. M's description of the assailant fit defendant's general description. M bit her assailant's arm during the assault, and defendant had a wound on his arm consistent with a bite mark. Also, on one other occasion, approximately five months before the assault on M, defendant stalked another teenage female in a similar manner. And, a pubic hair obtained from M's pubic area matched defendant's pubic hair. Finally, defendant's co-workers testified that he asked them to tell police that he was at a party until 4:00 a.m. on the night of the assault when he in fact left the party around 2:30 a.m.

Nevertheless, the sufficiency of the evidence to convict is not the focus of the harmless error analysis. As our supreme court has reminded us, "we do not, and cannot, find harmless error based on our idea of guilt or innocence or whether there is sufficient proper evidence to convict." Bible, 175 Ariz. at 590, 858 P.2d at 1193. To affirm defendant's conviction, we "must be confident beyond a reasonable doubt that the error had no influence on the jury's judgment." Id. at 588, 858 P.2d at 1191.

Here, in the absence of the testimony about the significance of a DNA match, the jury would have been required to decide whether M was more credible than defendant. Defendant testified that he did not assault M. And, although M positively identified defendant at trial, her pre-trial identification of defendant was questionable. At trial, M testified on direct examination that two of the six photographs in a photo lineup presented to her two days after the assault looked like her assailant. She also testified that one of the photos was that of defendant. On cross-examination, however, M said that she identified either Exhibit 35 or Exhibit 39 from the photo lineup as the photo of defendant. Neither of these exhibits were photos of defendant. Instead, Exhibit 34 from the photo lineup was defendant's photograph. Detective Gary Remeikis, who presented the photo lineup to M, verified that defendant's photograph was not among the two photographs M identified the day following the assault.

The state's theory of the case was that a DNA match meant that defendant was the assailant. In closing argument, the prosecutor emphasized this point by stating that the expert testimony regarding a match "doesn't not just exclude Mr. Hummert, it not only includes him, it tells you Mr. Hummert was the rapist." (Emphasis added.) This, of course, was not true. An argument that the random match probability constitutes a "guilt probability" is incorrect and misleading. "[T]he DNA random match probability `says nothing about guilt or innocence.'" Bible, 175 Ariz. at 582 n. 18, 858 P.2d at 1185 n. 18.

Against this backdrop, we conclude that the jury was undoubtedly influenced by the statements of the two experts that the declared match of DNA samples indicated a unique source. Accordingly, we reverse defendant's convictions and sentences. On retrial, the state may introduce evidence regarding a match of DNA samples only if it is accompanied by the explanation that such a match signifies that defendant is not excluded as the donor. Any further statement regarding the significance of the match is inadmissible in the absence of random match probability data meeting Frye standards.

Because this matter must be tried again, we address the remaining issues raised by defendant on appeal.

B. Evidence of Other Acts

1. Evidence that Defendant Previously Followed S

Defendant argues that the trial court erred in admitting evidence that he followed S through Chandler during the early morning of February 22, 1989. He contends that because the state offered this evidence to prove his bad character, it was inadmissible under Rule 404(b), Arizona Rules of Evidence. We conclude that this evidence was properly admitted under Rule 404(b) to prove the identity of M's assailant.

To be admissible under the 404(b) identity exception, the state must show: (1) that the defendant committed the prior act; and (2) that "the prior act was not too remote in time, was similar to the offense charged and was committed with a person similar to the prosecuting witness in the case being tried." Bible, 175 Ariz. at 575, 858 P.2d at 1178 (quoting State v. Roscoe, 145 Ariz. 212, 217, 700 P.2d 1312, 1317 (1984), cert. denied, 471 U.S. 1094, 105 S. Ct. 2169, 85 L. Ed. 2d 525 (1985)). Here, it is undisputed that defendant was driving the car identified by S as the car that was following her on February 22, 1989. Thus, defendant clearly committed the prior act. Because the prior incident occurred less than five months prior to the assault on M, it was not too remote in time. Further, numerous similarities exist between the two incidents: both involved teenage women that were driving alone; both occurred around 3:00 a.m.; both occurred in the same general area of Maricopa County; both involved persistent following of the women over a distance of seven to twelve miles despite intervening traffic stops and turns; and both involved a red Honda CRX bearing a silver emblem in the shape of a map of Texas. Thus, evidence of the prior incident involving S and defendant satisfies the test for admission under the identity exception to Rule 404(b).

Defendant contends that the incidents were not sufficiently similar to justify admission of S's testimony. He argues that defendant's CRX followed S without making any evasive maneuvers while the CRX observed by M before she was assaulted turned away from her direction of travel both times it was clearly observed. We reject this argument. "Absolute identity in every detail cannot be expected. Where an overwhelming number of significant similarities exist, the evidence of the prior act may be admitted." Roscoe, 145 Ariz. at 218, 700 P.2d at 1318. Because of the numerous similarities between the two incidents, the distinction drawn by defendant is clearly immaterial.

Defendant also argues that S's testimony gave the impression that he was a "bad man" and therefore her testimony should have been precluded as unfairly prejudicial under Rule 403, Arizona Rules of Evidence. We disagree. Rule 403 provides that relevant evidence may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice. "`[T]he appellate court "must look at the evidence in a light most favorable to its proponent, maximizing its probative value and minimizing its prejudicial effect."'" State v. Castro, 163 Ariz. 465, 473, 788 P.2d 1216, 1224 (App. 1989) (citations omitted). Applying this standard, we find that the probative value of the evidence that defendant stalked S is not substantially outweighed by any unfair prejudice.

Furthermore, the trial judge gave a limiting instruction to the jury about the effect of such evidence. "A timely instruction from the judge usually cures the prejudicial impact of evidence unless it is highly prejudicial or the instruction is clearly inadequate." United States v. Berry, 627 F.2d 193, 198 (9th Cir.1980). Here, the trial court instructed the jurors that evidence of defendant's prior act should be considered only in determining defendant's identity and not his character. We conclude that this instruction was more than adequate to cure any prejudicial effect the evidence might have had. Accordingly, the trial court did not err in admitting the prior act evidence.

2. Exclusion of Evidence About Mesa Sexual Assault

Defendant argues that the trial court erred by excluding evidence of a similar sexual assault in Mesa from which defendant was excluded as the assailant. We agree. We conclude that sufficient similarities exist between the assault on M and the Mesa assault to warrant the admission of the Mesa assault evidence under Rule 404(b). Accordingly, the trial court should have admitted this evidence.

The Mesa assault occurred approximately seven weeks before the assault on M. In that incident, the victim was assaulted just outside of her residence, at approximately 2:45 a.m. The assailant approached the victim as she got out of her car and pointed a long-barreled handgun at her head. He then grabbed her by the hair with his other hand and led her to the front yard of a neighboring residence. As in this case, the assailant penetrated the victim's vagina with his finger and with his penis. The Mesa assailant asked his victim, "How come you aren't wet?" This was similar to the statement made by M's assailant, "Get wet for me." In both incidents, the assailants asked their victims, "Are you on the pill?" And, both assailants inflicted head wounds on their victims. Finally, the victim of the Mesa assault gave a description of her assailant that was similar to the description of M's assailant.

Defendant, however, was eliminated from consideration as a suspect in the Mesa assault. This occurred because he was excluded by the victim from a photo lineup. In addition, a hair comparison tended to eliminate defendant as the assailant. The foreign hair found on the victim could not have come from defendant.

Defense counsel admitted that there were some differences between the Mesa incident and the assault on M. He noted that the Mesa assailant tried to force the victim into her car after the assault and then fled the scene on foot. Another dissimilar fact was that the Mesa assailant attempted vaginal penetration from behind the victim. M's assailant did not attempt vaginal penetration from this position.

Although the trial court observed "that there are some striking similarities between the two cases which would cause, ... a fact finder to be suspicious or to suspect that the same person committed both offenses," it excluded the evidence. The trial court noted that "[t]he cases say there has to be something inherent in ... the Mesa rape case incident to connect that assailant with this offense here.... And I don't think that can be said." The trial court relied on State v. Oliver, 169 Ariz. 589, 590-91, 821 P.2d 250, 251-52 (App. 1991), which stated:

A defendant can always show that some person other than the defendant committed the crime. Thus, where there is evidence of third-party culpability that raises a reasonable doubt as to defendant's guilt, it should be admitted. However, it is within the sound discretion of the trial court to refuse to admit such evidence when it simply affords a possible ground of suspicion against another. The evidence must show that the evidence has an inherent tendency to connect the other person with the actual commission of the crime.

(Citations omitted) (emphasis added). See also State v. Williams, 133 Ariz. 220, 230, 650 P.2d 1202, 1212 (1982). Here, the trial court indicated that the inherent tendency to connect the Mesa assault with this offense was absent because there were "numerous differences" between the incidents. We disagree.

Rule 404(b) governs our analysis. Other courts have addressed, under rules similar to Rule 404(b), defense use of other crime evidence to prove misidentification. See, e.g., United States v. Spencer, 1 F.3d 742, 745 (9th Cir.1992) (holding that reverse 404(b) evidence inadmissible because crime was not distinctive); United States v. Stevens, 935 F.2d 1380, 1401-06 (3rd Cir.1991) (finding that exclusion of other crime evidence offered to prove misidentification was error); New Jersey v. Garfole, 76 N.J. 445, 388 A.2d 587 (1978); cf. United States v. Armstrong, 621 F.2d 951, 953 (9th Cir.1980) (similar result; no analysis under Rule 404(b)); United States v. Robinson, 544 F.2d 110, 112-13 (2nd Cir.1976), cert. denied, 434 U.S. 1050, 98 S. Ct. 901, 54 L. Ed. 2d 803 (1978) (same). The relevance of such "reverse 404(b)" evidence was explained in Stevens:

The syllogism goes as follows. In view of the many parallels between the two crimes, one person very likely committed both; and because [defendant] was exonerated by [the victim of the other offense] ... [defendant] was not that person. The critical question, of course, is one of similarity.

935 F.2d at 1401.

Several cases hold that other crime evidence may be admitted on a lower showing of similarity in a "reverse 404(b)" case because the element of prejudice to the defendant need not be considered. See, e.g., id. at 1404; Garfole, 388 A.2d at 591. We need not decide here, however, whether a more generous standard should govern the admission of other crime evidence offered by the defense. In light of the numerous similarities between the Mesa assault and the assault on M, we find that the slight differences between the two incidents were not significant enough to justify exclusion of the Mesa assault evidence. As we noted earlier, "Absolute identity in every detail cannot be expected. Where an overwhelming number of significant similarities exist, the evidence of the prior act may be admitted." Roscoe, 145 Ariz. at 218, 700 P.2d at 1318.

The trial court also based its decision to exclude the evidence on Rule 403. The trial court stated that the "dangers" outweighing the probative value of the evidence included "delay in the trial of the case, not knowing where the admission of that evidence might lead in terms of further witnesses and exhibits, trying of a second sexual assault case, ... confusing the jury, and misleading the jury concerning the true issues in this case." Again, we disagree.

In ruling on the admissibility of other crime evidence, the trial court properly considered Rule 403. Atwood, 171 Ariz. at 638, 832 P.2d at 655. Nevertheless, exclusion of relevant evidence under Rule 403 is to be employed sparingly as it is an extraordinary remedy. United States v. Meester, 762 F.2d 867, 875 (11th Cir.), cert. denied, 474 U.S. 1024, 106 S. Ct. 579, 88 L. Ed. 2d 562 (1985). Here, the evidence should not have been excluded under Rule 403. The trial court's conclusion that the jury would have been "confused" or "misled" is not supported by the record. Furthermore, any confusion the jury might have had about the evidence could have been cured with a limiting instruction informing the jury of the purpose of the evidence. Id.; Ariz.R.Evid. 105. Consequently, the potential for confusion did not support wholesale exclusion of the evidence. Cf. Castro, 163 Ariz. at 473, 788 P.2d at 1224.

Furthermore, the only delay that is a bar to the admission of such evidence is undue delay. Hill v. Bache Halsey Stuart Shields, Inc., 790 F.2d 817, 826 (10th Cir.1986). Here, the record does not indicate that presentation of the facts of the Mesa assault would have unduly delayed the trial, which had already consumed fourteen days by the time of the trial court's ruling. Although the trial court may have had legitimate concerns that delay would further disrupt the jurors' lives, that concern should not have been dispositive.

The trial court's concern regarding "the trying of a second sexual assault case" may have stemmed from the prosecutor's avowal that the detective who investigated the Mesa assault doubted the victim's claim that she had been sexually assaulted. According to the prosecutor, the detective based her opinion on the fact that no semen was found in a sexual assault examination and that no unidentified fingerprints were located on areas of the victim's car that the assailant supposedly touched. The prosecutor said that the detective "felt" the victim made the false allegation to get attention from her boyfriend.

Although the Mesa detective's opinion on the truth of the victim's claim was not admissible at trial, see State v. Lindsey, 149 Ariz. 472, 475, 720 P.2d 73, 76 (1986), the trial court was not limited by the rules of evidence in determining the admissibility of the other crime evidence. Ariz.R.Evid. 104(a). Nevertheless, if the evidence of the Mesa assault is offered on retrial, the trial court should proceed with caution in premising exclusion of this evidence on the Mesa detective's opinion. Because Rule 404(b) assumes that the prior incident actually occurred, facts that cast doubt on its occurrence — for example, a recantation by the victim — could be weighed by the trial court in determining the admissibility of the prior incident.

It is the basis for the detective's opinion, however, rather than the fact of that opinion that the court must assess in determining whether the evidence is admissible. In the context of determining admissibility, "the judge's preliminary inquiry `should be limited to asking whether evidence in the record ... would permit a reasonable person to believe' the evidence on the preliminary questions." State v. Plew, 155 Ariz. 44, 50, 745 P.2d 102, 108 (1987) (quoting State v. LaGrand, 153 Ariz. 21, 28, 734 P.2d 563, 570, cert. denied, 484 U.S. 872, 108 S. Ct. 207, 98 L. Ed. 2d 158 (1987)); accord State ex rel. McDougall v. Superior Court, 172 Ariz. 153, 155-56, 835 P.2d 485, 487-88 (App. 1992). Specifically, the trial court should determine if sufficient evidence exists that would permit a reasonable person to believe that the Mesa victim was truly assaulted. If this test is met, the Mesa assault evidence should be admitted for the jury to weigh together with all other admissible evidence. See McDougall, 172 Ariz. at 156, 835 P.2d at 488 ("The jury occupies the unique province of weighing the credibility, veracity, and reliability of evidence and of resolving the conflicting inferences drawn from the evidence.").

C. Admission of In-Court Identification

Defendant claims that the trial court erred in permitting M to make an in-court identification of defendant because the identification was based entirely on M having seen defendant at trial seated next to defense counsel at the defense table. Defendant argues that the in-court identification by M was unduly suggestive under State v. Dessureault, 104 Ariz. 380, 453 P.2d 951, cert. denied, 397 U.S. 965, 90 S. Ct. 1000, 25 L. Ed. 2d 257 (1969). We disagree. We conclude that the confrontation of a defendant during trial by an identifying witness is not an unduly suggestive identification procedure under Dessureault. Accordingly, the trial court properly admitted M's in-court identification.

Before trial, defendant requested that the trial court hold a Dessureault hearing if M indicated during trial that she was able to identify defendant. Defendant did not contend that M had been subjected to a suggestive pre-trial identification procedure. Instead, defendant contended that because M would see him seated at the defense table with counsel during her testimony, her in-court identification would be tainted. Defendant then requested that the court determine whether any in-court identification of him by M would be based on a source independent of the earlier photo array and M's observation of him in the courtroom.

When M indicated that she could identify defendant as her assailant, the trial court recessed trial and heard additional testimony about the propriety of an in-court identification. The trial court concluded that no determination regarding the source of the identification was necessary because the confrontation between the witness and defendant at trial was not an unduly suggestive procedure under Dessureault. Nevertheless, the trial court found that any in-court identification of defendant by M had a reliable independent source — "her claimed memory of the events of the evening in question."

We agree with the trial court that the in-court identification by M was not an unduly suggestive procedure requiring application of Dessureault. As a matter of federal due process, the United States Supreme Court has "never set any guidelines for in-court identification procedures nor indicated that in-court identification must be made in a way that is not suggestive." United States v. Domina, 784 F.2d 1361, 1368 (9th Cir.1986), cert. denied, 479 U.S. 1038, 107 S. Ct. 893, 93 L. Ed. 2d 845 (1987). Similarly, no Arizona case has interpreted Dessureault to require a hearing prior to an in-court identification in the absence of a prior suggestive, out-of-court procedure.

Furthermore, our supreme court's decision in State v. Meeker, 143 Ariz. 256, 693 P.2d 911 (1984), strongly supports our conclusion that Dessureault does not apply in these circumstances. There, the defendant claimed that his counsel was ineffective by failing to secure an in-court lineup to minimize the chance of misidentification by witnesses. Meeker further argued that the in-court identifications were unduly suggestive because he was seated with defense counsel at a table that had a sign stating "defense." The Meeker court rejected the claim, stating:

We find no requirement for "in court identifications" that the accused be surrounded by persons nearly identical in appearance. Furthermore, merely [having] a sign stating "defense" is not unduly suggestive.

Id. at 265, 693 P.2d at 920.

Defendant cites State v. McCall, 139 Ariz. 147, 677 P.2d 920 (1983), cert. denied, 467 U.S. 1220, 104 S. Ct. 2670, 81 L. Ed. 2d 375 (1984), and State v. Rosthenhausler, 147 Ariz. 486, 711 P.2d 625 (App. 1985), as support for his proposition that M's in-court identification was unduly suggestive. We find that McCall and Rosthenhausler are distinguishable from this case. In McCall, our supreme court held that a preliminary hearing which produced a positive identification of defendant was suggestive. In that case, however, the suggestive circumstances were extreme:

[Defendant] was present and sitting next to his attorney at the defense table. He was wearing handcuffs, leg manacles, and a white T-shirt with "MCSO" printed across the front. Two uniformed deputies were in the courtroom; one sat directly behind the appellant while one sat behind the prosecution.

McCall, 139 Ariz. at 154, 677 P.2d at 927. Likewise, in Rosthenhausler, the court found that a positive identification made at a Dessureault hearing was suggestive because the defendant was the only person in the court-room wearing handcuffs. Here, when M made her in-court identification, defendant was not handcuffed, shackled, or dressed in jail clothes. Although defendant was seated at the defense table, this fact by itself does not render an in-court identification unduly suggestive. Thus, McCall and Rosthenhausler do not apply to this case.

D. Grand Jury Testimony

Defendant claims that the indictment must be dismissed because it was partially based on perjured, material testimony given by Detective Remeikis before the grand jury. We disagree. Because no evidence exists that the testimony given by Remeikis could be characterized as perjurious, we conclude that the indictment need not be dismissed.

Before trial, defendant moved to remand this case to the grand jury for a new finding of probable cause, arguing that Remeikis misled the grand jury on three points. Defendant claimed that Remeikis falsely testified: (1) that M's assailant wore "boxer-type underwear," arguably similar to the white shorts defendant wore to the party, even though M had said only that her assailant wore "white underwear"; (2) that defendant was twenty-three years old even though his true age was twenty-eight; and (3) that defendant had no children when he actually had two. The trial court denied the motion to remand. On cross-examination at trial, Remeikis acknowledged that each of the statements at issue was incorrect.

In general, to obtain review of the denial of a motion for a new finding of probable cause, a defendant must seek relief by special action before trial. State v. Verive, 128 Ariz. 570, 575, 627 P.2d 721, 726 (App. 1981). An exception to this rule exists, however, when a defendant "has had to stand trial on an indictment which the government knew was based partially on perjured, material testimony." State v. Gortarez, 141 Ariz. 254, 258, 686 P.2d 1224, 1228 (1984). The exception adopted in Gortarez was derived from the Ninth Circuit's decision in United States v. Basurto, 497 F.2d 781 (9th Cir.1974). There, the court ordered the dismissal of an indictment after the defendant established that a witness had informed the government before trial that the witness had given perjured testimony before the grand jury. Id. at 784.

Here, because defendant has not established any evidence that the testimony given by Remeikis was perjurious, we decline to dismiss the indictment. "A person commits perjury by making a false, sworn statement in regard to a material issue, believing it to be false." A.R.S. § 13-2702 (1989) (emphasis added). No evidence exists in this record that Remeikis believed the statements were false when he made them to the grand jury. Defense counsel essentially conceded in his closing argument that Remeikis did not intentionally lie, stating:

Detective Remeikis with a mindset that, hey, I got a job to do, we've got an indictment to return here, consciously, subconsciously intending to lie? I don't suggest that.

(Emphasis added.) Furthermore, no evidence exists that the state knew that Remeikis's statements were perjurious. Thus, this case does not fall within the Gortarez exception. Accordingly, the indictment is proper.

E. Reasonable Doubt Instruction

Defendant claims that the trial court's jury instruction defining "reasonable doubt" was fundamental error because it improperly excluded "possible doubt" from the definition. Our supreme court recently held that giving such an instruction is not fundamental error. State v. West, 176 Ariz. 432, 862 P.2d 192 (1993); accord State v. Portillo, 876 P.2d 1151, 1155-56 (App. 1994); State v. Duzan, 176 Ariz. 463, 862 P.2d 223 (App. 1993). Thus, the reasonable doubt instruction given in this case was proper.

III. CONCLUSION

Pursuant to A.R.S. section 13-4035 (1989), we have reviewed the entire record for fundamental error. Apart from the error in admitting testimony about the statistical significance of a DNA match, we find no fundamental error.^[3] We reverse defendant's convictions and sentences and remand the matter for a new trial consistent with this opinion.

WEISBERG, P.J., and CONTRERAS, J., concur.

NOTES

[1] To protect the privacy of victims and witnesses, we refer to them by initials only. State v. Bartlett, 164 Ariz. 229, 230 n. 1, 792 P.2d 692, 693 n. 1 (1990), vacated on other grounds, 501 U.S. 1246, 111 S. Ct. 2880, 115 L. Ed. 2d 1046 (1991).

[2] The trial court appeared to conclude that the testimony was unobjectionable because the experts couched their opinions in terms of their personal experience as scientists. We believe that this approach is nothing more than a subtle evasion of Frye. If such testimony were admissible, any expert could express an opinion regarding matters not generally accepted by the relevant scientific community merely by framing it in terms of his or her own observations.

[3] We are unable determine from the record whether the trial court properly assessed the $600.00 felony penalty assessment in conformity with State v. Gordon, 161 Ariz. 308, 778 P.2d 1204 (1989), and State v. Alexander, 175 Ariz. 535, 858 P.2d 680 (App. 1993). Because we are reversing and remanding for a new trial, however, this point is moot.