The opinion of the Court was delivered by
POLLOCK, J.Defendant, Nathaniel Harvey, appeals directly from a judgment of conviction and sentence of death for the purposeful-or-knowing murder of Irene Schnaps. A jury originally convicted defendant of Schnaps’s murder and sentenced him to death in October 1986. This Court reversed that conviction because of errors in the admission of defendant’s confession and in the failure of the trial court to give a “Gerald charge.” State v. Harvey, 121 N.J. 407, 581 A.2d 483 (1990) (Harvey I), cert. denied, 499 U.S. 931, 111 S.Ct. 1336, 113 L.Ed.2d 268 (1991). The phrase “Gerald charge” refers to a charge that distinguishes murder when the defendant intended to kill from murder when the defendant intended only to cause serious bodily injury that resulted in death. State v. Gerald, 113 N.J. 40, 549 A.2d 792 (1988). Neither error occurred in the second trial.
In the absence of defendant’s confession, the State relied substantially on DNA evidence to establish that defendant was Sehnaps’s killer. Again, a jury convicted defendant and imposed the death penalty. On this appeal, defendant raises numerous points, including challenges to the admission of the DNA evidence and to the jury charge. After careful review of all of defendant’s arguments, we affirm his conviction and death sentence.
- I -
A. Discovery of the Body and the Crime Scene
Schnaps, age thirty-seven, lived alone in a ground-floor apartment at the Hunter’s Glen complex in Plainsboro, New Jersey. *138After she failed to appear at work on June 17, 1985, a concerned eoworker went to her apartment and entered through an unlocked doorway. On discovering Schnaps’s lifeless body, he immediately called for assistance.
Investigating police detected no signs of forced entry. The bedroom, however, was a scene of obvious struggle. Blood stains were on the carpet and throughout the room. Sehnaps’s naked body lay face-up on the floor. She had sustained severe head and facial wounds. Despite the extensive head wounds, no bloodstains were present on Sehnaps’s chest and stomach.
The matting of several small hairs to the victim’s body and the absence of blood on her torso suggested that someone had attempted to wipe the body clean. The carpeting around the body was wet from water. Schnaps’s back was covered with blood.
A white pillowcase bore a bloody sneaker-print with a chevron pattern and the letters “PON.” Although the bedding appeared clean, blood stained the mattress, the underlying box spring, a cardboard box that protruded from under the bed, and a towel.
The bedroom also included an empty Seiko-LaSalle watch box, an empty Olympus camera box, and an empty jewelry box. In the bathroom, the investigators found Schnaps’s poeketbook. The pocketbook was open and did not contain any money.
B. The Autopsy
On June 18, 1985, Dr. Marvin Shuster, the Middlesex County medical examiner, conducted an autopsy. He determined that Schnaps had sustained approximately fifteen blows to the head. The largest wound, six inches long and one inch wide, extended from the front of her forehead to the top of her head. In general, the skull wounds were either curving or linear. The curving wounds were likely caused by hammer blows, and the linear wounds could have been caused by an item akin to a tire iron, a two-by-four, or a dull hatchet or axe. Some of the blows fractured Schnaps’s skull and caused direct injury to the brain. Blows had *139been delivered from both the right and left sides, some from the front, but most from behind.
Triangular pressure marks appeared on both sides of the neck. Some of the victim’s teeth were knocked out, and her jaw was broken. The right sides of the neck, jaw, cheek, and forehead were bruised, and she was cut behind one ear.
Unable to attribute death to any particular wound, Dr. Shuster concluded that the combination of the blows had killed Schnaps. Sehnaps had bled profusely and died within a matter of minutes.
C. The Apprehension and Interrogation of Nathaniel Harvey
1. October 28, 1985
Throughout the summer and autumn of 1985, West Windsor police looked for the perpetrator of a series of unsolved burglaries and sexual assaults. Based on eyewitness descriptions, they believed that the perpetrator was a stocky black male, under five-feet three-and-a-half inches tall, who usually travelled on foot or by bicycle. The police also believed that the perpetrator of those other crimes might be responsible for Schnaps’s murder. Defendant fit the physical description.
On October 28, 1985, police investigating three burglaries arrested defendant after he was sighted standing with his bicycle at the edge of a soybean field in West Windsor. One of the burglary victims identified defendant at a subsequent “show-up.”
During questioning by West Windsor police on October 28, defendant confessed to committing a number of burglaries in West Windsor, as well as a sexual assault. Defendant also agreed to accompany the police on a car tour to point out the locations of his crimes.
2. October 29, 1985
At 10:00 a.m. on the following morning, defendant accompanied two detectives on a car tour of West Windsor. At 1:15 p.m., defendant consented to a search of his car and his Jamesburg apartment for evidence related to an unrelated sexual assault. *140Although defendant gave as his address his father’s apartment in Jamesburg, he lived with his estranged wife in West Windsor. Apparently, defendant feared that his wife would lose her welfare benefits if he revealed that he lived with her. After defendant signed the consent form, police transferred him to the Mercer County Detention Center.
While searching defendant’s car, the officers discovered two watches, including a Seiko-LaSalle like the one missing from Schnaps’s apartment. They notified the Plainsboro Police Department. After obtaining a search warrant, a Plainsboro officer seized the watch. The search of Harvey’s Jamesburg apartment did not yield any evidence.
3. October SO, 1985
Following defendant’s arraignment for the murder of Schnaps, investigating officers resumed questioning him. At one point, defendant said that “he would tell [them] about the murder but first wanted to speak to his father.” After defendant spoke with his father, police failed to administer new Miranda warnings. Shortly thereafter, defendant confessed to murdering Schnaps.
D. The First Trial
On November 19, 1985, a Middlesex County grand jury returned a three-count indictment charging defendant with the purposeful-or-knowing murder of Schnaps, contrary to N.J.S.A. 2C:11-3 (count one), second-degree robbery, contrary to N.J.S.A. 2C:15-1 (count two), and second-degree burglary, contrary to N.J.S.A. 2C:18-2 (count three). Two days later, on November 21, 1985, the Middlesex County Prosecutor filed a Notice of Aggravating Factors pursuant to Rule 3:13-4(a) and N.J.S.A. 2C:11-3c(2), making defendant’s case a capital prosecution. The State alleged the following aggravating factors:
1. The murder was outrageously or wantonly vile, horrible or inhuman in that it involved an aggravated battery to the victim
[N.J.S.A. 2C:11 — 3c(4)(c).]
*1412. The murder was committed for the purpose of escaping detection, apprehension, trial, imprisonment or confinement for robbery and burglary committed by the defendant
[N.J.S.A. 2C:11 — 3c(4)(f).]
3. The murder was committed while the defendant was engaged in the commission of or an attempt to commit, or flight after committing robbery and burglary
[N.J.S.A. 2C:11-3c(4)(g).]
The prosecution relied heavily on defendant’s confession. Harvey I, supra, 121 N.J. at 415-17, 581 A.2d 483. The jury found defendant guilty of purposeful-or-knowing murder, first-degree robbery, second-degree burglary, and felony murder, for which he had not been indicted. At a penalty-phase hearing, the same jury found the presence of all three alleged aggravating factors and returned a sentence of death.
E. Harvey I
On direct appeal, this Court reversed defendant’s conviction and remanded for a new trial. The Court held that the trial court’s jury instructions on murder did not comport with Gerald, supra, 113 N.J. 40, 549 A.2d 792, which required that a jury must be instructed separately on the crimes of intentional murder and serious-bodily-injury murder (SBI minder). At the time of Harvey’s trial, a conviction for SBI murder did not render a defendant death-eligible. In his confession, Harvey claimed that the victim struck him and that he then struck her only once. Relying in part on statements in his confession, the Court concluded that the evidence provided a rational basis for a jury to have concluded that defendant intended only to injure Schnaps. Harvey I, supra, 121 N.J. at 413, 581 A.2d 483.
The Court further held that Harvey’s confession had been procured in violation of State v. Hartley, 103 N.J. 252, 511 A.2d 80 (1986). Hartley provides that after invoking the right to silence, a defendant must receive new Miranda warnings before interrogation can resume. The Court ruled that, by asking to speak with his father, Harvey had invoked his right to silence. Harvey I, supra, 121 N.J. at 418-20, 581 A.2d 483. Consequently, the police *142should have advised him again of his constitutional rights before resuming interrogation. The failure of the police to abide by that bright-line test rendered defendant’s confession inadmissible. Id. at 422, 581 A.2d 483.
F. The Interim Between Trials
Faced with the prospect of retrying Harvey without his confession, the prosecution hired Cellmark Diagnostics Laboratory (Cell-mark) to conduct DNA tests on the blood recovered from the crime scene. Cellmark, the first commercial laboratory accredited by the American Society of Crime Laboratory Directors, conducts DNA tests exclusively. It analyzed a bloodstained section of Schnaps’s box spring, a bloodstained piece of cardboard, and a sample of both Schnaps’s and defendant’s blood.
G. The Retrial
Defendant’s retrial, from the pretrial motions to the return of the death sentence, lasted from November 25, 1992, to December 16, 1994. The trial court denied a motion for a new trial on January 30,1995.
1. Pretrial Motions
Following a hearing, the trial court denied defendant’s motion to suppress evidence seized from defendant’s car, ruling that defendant had consented to the search. The defendant also moved to exclude the State’s DNA evidence. After a three-day New Jersey Rule of Evidence 104 (Rule 104) hearing, the trial court denied defendant’s motion and held that the State’s DNA evidence was admissible.
2. Guilt Phase
After a lengthy jury-selection process, the guilt phase began on November 29,1994.
a. State’s Case
The State adduced evidence about the crime scene, including fifty-two photographs and various items of physical evidence. *143Investigating officers testified to the discovery of the bloody sneaker print, the empty Seiko-LaSalle watch box, the empty jewelry box, and the empty Olympus camera box, all of which were admitted into evidence.
Philip Beesley, a forensic scientist employed by the New Jersey State Police, testified that blood work done on control samples from both the defendant and Schnaps revealed that Schnaps’s blood was type “one plus, one minus” for the genetic marker PGM, and type “1” for the enzyme CA II. Defendant was type “one plus, one plus” for PGM and type “2-1” for CA II. Beesley further revealed that bloodstains found on the box spring and on the piece of cardboard tested as “one plus, one plus” for PGM and type “2-1” for CA II. He concluded that those stains were consistent with Harvey’s blood, therefore, not Schnaps’s. Beesley also testified that CA II of phenotype 2-1 is found only in African Americans.
Dr. Marvin Shuster testified about the nature of the wounds suffered by Schnaps and the cause of her death. See supra part I.B. Theodore Mozer, a forensic scientist employed by the New Jersey State Police, testified that one of the hairs recovered from Schnaps’s back did not belong to her. Mozer testified that this hair had “Negroid” characteristics that were consistent with a control hair taken from Harvey.
Mozer also examined two pairs of sneakers seized from Harvey’s ex-wife’s West Windsor apartment and the size 6 1/2 “Pony” sneakers that Harvey was wearing when he was arrested. Aided by six enlarged photographs of the bloody footprint left at the crime scene, Mozer explained that Harvey’s “Pony” sneakers were consistent with the sneaker impression. Although Harvey’s sneakers “could” have left the bloody mark, Mozer could not conclude definitively that they had done so.
In support of the admission of the DNA evidence, the State presented two witnesses from Cellmark, Julie Cooper, a senior molecular biologist, and Dr. Charlotte Word, a microbiologist and supervisor of forensic casework. They testified generally that *144DNA tests conducted on the blood samples recovered at the crime scene were genetically comparable to defendant’s DNA. Defendant’s genotypes for the genetic markers examined were common to only one-in-1,400 African Americans.
b. Defendant’s Case
Defendant did not testify.
His case consisted of only two witnesses. A witness from Seiko testified that it had made thousands of watches like the one seized from the trunk of defendant’s car. Dr. Robert Shaler, Director of Forensic Biology for the Office of the Chief Medical Examiner for the City of New York, testified that he believed that the DNA tests were “scientifically indefensible.” In the course of his testimony, he pointed out the imbalances in the dots on the strips. He found imbalances at the GYPA, HBGG, and GC loci. Dr. Shaler testified that at the GYPA locus an individual’s genes could cause a difference in dot intensity. He further disputed the one-in-1,400 calculation and asserted that he believed that the genetic makeup of the blood recovered from the crime scene could be found in approximately one in fifty to one in 200 African Americans.
c. The Verdict
In its jury instructions, the court included a Gerald charge. Consistent with State v. Purnell, 126 N.J. 518, 530-34, 601 A.2d 175 (1992), the court also charged on the unindicted count of felony murder. After deliberating for three and one-half hours, the jury returned its verdict finding defendant guilty of purposeful-or-knowing murder, felony murder, first-degree robbery, and second-degree burglary.
3. Penalty Phase
a. State’s Case
The State relied exclusively on the evidence adduced at the guilt phase to support proof of three aggravating factors: the murder involved aggravated assault of the victim, N.J.S.A. 2C:11-3c(4)(c); the murder was committed to escape detection, N.J.S.A. 2C:11-*1453e(4)(f); and the murder was committed during the course of a robbery and burglary, N.J.S.A. 2C:11-3c(4)(g).
b. Defendant’s Case
Pursuant to N.J.S.A. 2C:11-3c(5)(c) and (h), defendant alleged ten mitigating factors: the age of the defendant at the time of the murder; the defendant was traumatized at a young age when he witnessed the death of his older sister; defendant was uprooted from his home and sent to live with his grandparents who abused him; defendant suffered feelings of abandonment when his parents moved north and failed to send for him; he was exposed to domestic violence in the home of his grandparents; he was exposed to domestic violence in the home of his parents; he is a caring and loving father; his continuing relationship with his children including financial contributions; his relationship with his mentally disabled brother and his mentally disabled daughter; and all factors which relate to the defendant’s childhood and family background.
Professor Richard Moran, a criminologist specializing in the correlation between age and crime, testified that if defendant were sentenced to prison rather than death, he could not be eligible for parole prior to age 64, by which time he would be in the age group least likely to commit violent crime. Therefore, the chances of defendant committing another violent crime would be minute.
A forensic social worker testified about defendant’s social history. Defendant was raised in poverty. His often-absent father was a sharecropper and a migrant worker. As a two-year old, defendant had been injured in an automobile accident, but did not receive medical treatment. When defendant was four, he and his five-year old sister were left in an unheated home. While trying to light a stove, defendant’s sister spilled some kerosene on her nightgown. When defendant lit a match, she burned to death.
Defendant’s parents later moved to New Jersey, leaving defendant in the care of his grandparents for seven years. During his childhood, defendant was abused by both his grandfather and father.
*146Various family members testified that defendant was a loving and earing father, who also comforted his developmentally-disabled brother. Defendant’s family asked the jury not to sentence defendant to death.
c. The Verdict
After deliberating for two and one-half hours, the jury returned a unanimous verdict that defendant had committed the murder for the purpose of avoiding apprehension, N.J.S.A. 2C:11-3c(4)(f), and in the course of a robbery and burglary, N.J.S.A. 2C:11-3c(4)(g). It did not find aggravating factor N.J.S.A. 2C:11-3c(4)(c), that the murder involved aggravated assault to the victim. The jury further found that the aggravating factors outweighed all of the mitigating factors and that each aggravating factor alone outweighed all of the mitigating factors. The trial court sentenced defendant to death.
Later, the trial court sentenced defendant as a persistent offender on the non-capital counts of first-degree robbery and second-degree burglary. Accordingly, defendant received a sentence of life with a twenty-five year parole bar on the first degree robbery conviction to run consecutively to defendant’s death sentence for capital murder. On the burglary conviction, defendant was sentenced to a concurrent term of five years in prison with a two-and-one-half year parole bar. Those sentences were made consecutive to prison terms previously imposed on defendant for unrelated crimes. Thus, defendant’s aggregate sentence, irrespective of the death penalty, is life plus sixty-five years with a fifty-seven-and-one-half year parole disqualifier.
- II -
We first consider defendant’s contention that the trial court committed reversible error when it failed to instruct the jury in accordance with this Court’s later decision in State v. Mejia, 141 N.J. 475, 662 A.2d 308 (1995).
*147- A-
Evaluation of defendant’s Mejia argument begins with Gerald, supra, 113 N.J. 40, 549 A.2d 792. In Gerald, this Court held as a matter of state constitutional law that only those murderers who intended to kill were eligible for the death penalty. Those who intended to inflict only serious bodily injury were not death-eligible under N.J.S.A. 2C:11—3(a)(1) or (2), even if their actions resulted in their victim’s death. Id. at 69-70, 549 A.2d 792. Subsequent constitutional and statutory amendments have abrogated the Gerald rule by subjecting to the death penalty murderers who intended to cause only serious bodily injury. N.J. Const. art. I, ¶ 12 (1992); L. 1993, c. 111 (signed May 5, 1993). At the time of Sehnaps’s murder, however, only those who murdered with the intent to kill were death-eligible. Thus, the Gerald rule applies to the present case.
Under Gerald, the jury’s determination whether defendant killed with the intent to kill or merely with the intent to inflict serious bodily injury, became the linchpin of capital-punishment eligibility. If the evidence produced at trial provided even a rational basis for a jury to convict a defendant of SBI murder rather than intentional murder, the trial court was compelled to “instruct the jury to specify which, if [either], of those findings forms the basis for conviction.” State v. Coyle, 119 N.J. 194, 209, 574 A.2d 951 (1990). Under Gerald, the jury’s key role became to determine whether defendant’s intent was to kill or to inflict SBI. State v. Moore, 122 N.J. 420, 484, 585 A.2d 864 (1991).
In Mejia, supra, 141 N.J. at 481, 662 A.2d 308, this Court clarified Gerald by stating that a jury need not be unanimous on whether the defendant intended to kill or to injure seriously. The intent-to-kill requirement is not an element of murder, but a “triggering device” for the death-penalty phase of the trial. Id. at 486, 662 A.2d 308. Thus, unanimity is not required in making the Gerald determination. Id. at 487, 662 A.2d 308 (noting “unanimity requirement extends only to verdicts adverse to the defendant”). *148A jury can return a valid guilty verdict for purposeful-or-knowing murder even if it cannot agree that defendant killed intentionally. Such a verdict, however, will not support the imposition of the death penalty. Id. at 486, 662 A.2d 308. In Mejia, the record provided “a rational basis for a jury to find that defendant intended only to cause serious bodily injury.” Id. at 481, 662 A.2d 308. The trial court’s failure to instruct the jury about the possibility of returning a non-unanimous verdict on the defendant’s intent thus constituted plain error.
In State v. Harris, 141 N.J. 525, 549, 662 A.2d 333 (1995), however, we held that a failure to give a non-unanimous Mejia charge could be considered harmless when the evidence of intent to kill was overwhelming and no rational basis existed for concluding that defendant had intended to inflict only serious bodily injury.
Although defendant’s case was tried seven months before the issuance of Mejia and Harris, defense counsel requested a charge on a non-unanimous verdict. Before us, defendant argues that the absence of a non-unanimous verdict charge coerced the jury into returning a verdict of capital murder. The initial inquiry, then, is whether the evidence provided a rational basis to find that defendant intended to inflict only serious bodily, and not to kill.
- B -
Failure to charge in accordance with Gerald/Mejia requires the reversal of a death sentence if the record below contains evidence that is “minimally adequate to provide a rational basis for the jury to hold a reasonable doubt that the defendant intended to cause death.” Mejia, supra, 141 N.J. at 489, 662 A.2d 308; see also State v. Pennington, 119 N.J. 547, 561, 575 A.2d 816 (1990) (characterizing rational-basis standard as a “low threshold”); State v. Pitts, 116 N.J. 580, 615, 562 A.2d 1320 (1989) (same). Accordingly, a rational basis may exist even though a jury likely would reject the defendant’s serious-bodily-injury theo*149ry. Mejia, supra, 141 N.J. at 489, 662 A.2d 308; see State v. Dixon, 125 N.J. 223, 254, 593 A.2d 266 (1991) (“The error was not harmless because there was evidence in this case that could have sustained an SBI ... verdict. We do not suggest that such a verdict was likely, but merely that if the jury returned that verdict, the court could not reject it.”). Satisfaction of the “minimally adequate/rational basis” standard, however, requires more than a mere “scintilla of the evidence.” Mejia, supra, 141 N.J. at 489, 662 A.2d 308; State v. Crisantos, 102 N.J. 265, 278, 508 A.2d 167 (1986).
We have found harmless error in eases where defendants’ actions have been “so wantonly brutal that the jury could have concluded only that the defendant intended to cause death.” Mejia, supra, 141 N.J. at 488, 662 A.2d 308; see, e.g., Harris, supra, 141 N.J. at 550, 662 A.2d 333 (victim was handcuffed and lying on ground when defendant shot him in back of head); State v. Bey, 129 N.J. 557, 579, 610 A.2d 814 (1992) (Bey III) (defendant stomped on victim with sufficient force to crush her chest), cert. denied, 513 U.S. 1164, 115 S.Ct. 1131, 130 L.Ed.2d 1093 (1995); State v. Biegenwald, 126 N.J. 1, 18, 594 A.2d 172 (1991) (Biegenwald IV) (defendant fired four gunshots into victim’s head); State v. McDougald, 120 N.J. 523, 558-60, 577 A.2d 419 (1990) (defendant slashed victims’ throats, bludgeoned one with baseball bat, and expressed intent to kill both before and after killings); State v. Hightower, 120 N.J. 378, 412-14, 577 A.2d 99 (1990) (Hightower I) (defendant shot victim at close range in chest, neck, and head, and then dragged victim into freezer); State v. Rose, 120 N.J. 61, 63-64, 576 A.2d 235 (1990) (Rose II) (defendant fired twelve-gauge shotgun point-blank into victim’s stomach); State v. DiFrisco, 118 N.J. 253, 571 A.2d 914 (1990) (DiFrisco I) (defendant shot victim four times in head and admitted intention to kill), cert. denied, — U.S. -, 116 S.Ct. 949, 133 L.Ed.2d 873 (1996); Pitts, supra, 116 N.J. at 614-20, 562 A.2d 1320 (defendant threatened to kill victims two days before murders, inflicted twenty-five to thirty stab wounds with combat knife, cut one victim’s throat twice, and paused to take victim’s pulse to verify death); State v. *150Hunt, 115 N.J. 330, 374-77, 558 A.2d 1259 (1989) (defendant stated intent to kill immediately prior to stabbing victim twenty-four times).
- C -
Schnaps was brutally murdered. Her killer struck her fifteen times in the head with a blunt instrument. He struck her with sufficient force and frequency to fracture her skull in several places, knock out her teeth, and break her jaw. He further applied sufficient pressure to her neck to cause severe bruising. Reviewing that medical evidence in Harvey I, supra, we wrote that “such repeated blows can support a jury finding of intentional murder.” 121 N.J. at 413, 581 A.2d 483.
Relying substantially on the assertions in defendant’s confession, however, we concluded that the evidence as a whole required reversal. We reached that conclusion because “the record provided a rational basis for the jury to find that [defendant] intended to cause only serious bodily injury.” Id. at 414, 581 A.2d 483 (internal quotations omitted). Contrary to the dissent’s assertions, all of the “pertinent facts” contained in Harvey’s confession were not before the jury in the retrial. Post at 299, 699 A.2d at 686. Not before the jury was defendant’s confession that he had struck the victim only once in response to being punched in the nose. Harvey I, supra, 121 N.J. at 412, 581 A.2d 483. That confession provided a rational basis for a juror in the first trial to have concluded that Harvey intended only to injure his victim and not kill her. The absence of Harvey’s confession is a critical distinction between the record in Harvey I and that in the present case.
- D -
In determining whether a rational basis existed for a Gerald charge, we are confined to the evidence in the record. See Bey III, supra, 129 N.J. at 581, 610 A.2d 814 (“[W]e examine scrupulously the evidence that was adduced at trial to see whether *151the jury had a rational basis for finding that the defendant could have intended only serious bodily injury.”); Dixon, supra, 125 N.J. at 253, 593 A.2d 266 (noting appellate role is to ask if jury answered question on death-eligibility). We may not consider the evidence in the first trial. In brief, defendant’s confession, which was admitted in the first trial but excluded from the second, is not part of the record.
Absent the confession, the evidence does not provide a rational basis for the supposition that defendant struck his victim to ward off her attack. The dissent attempts to distinguish Sehnaps’s murder from related cases where we have found harmless trial courts’ failure to provide a non-unanimity instruction. According to the dissent, repeated bludgeoning to the head by a hammer-like instrument, such as the one used by defendant, “is not like a gun fired at close range into a person; the victim’s death is not so assured.” Post at 302, 699 A.2d at 687 (citing Hightower I, supra, 120 N.J. at 412-14, 577 A.2d 99 (finding that defendant shot victim with a .32-caliber handgun from close range in the chest, neck, and head)). Incomprehensibly, the dissent also concludes that, unlike other cases, “there is no evidence that defendant took further steps to assure Ms. Schnaps’s death.” Post at 302, 699 A.2d at 687 (citing Hightower I, supra, 120 N.J. at 413, 577 A.2d 99 (finding that defendant dragged victim into freezer); Pitts, supra, 116 N.J. at 618, 562 A.2d 1320 (noting that defendant took victim’s pulse)). We disagree. The inescapable inference created by the objective facts, especially the severity and number of Schnaps’s wounds, is that defendant intended to kill her.
- E -
Nor is defendant’s argument strengthened by the trial court’s delivery of a Gerald charge. Given this Court’s reversal of the conviction in the first trial, the inclusion of a Gerald charge in the second trial is understandable, if unnecessary. In reversing the conviction in the first trial, which included defendant’s confession, we did not predetermine the need for such a charge in a retrial in *152which the confession was excluded. Likewise, the inclusion of charges on the lesser-included offenses of aggravated and reckless manslaughter do not compel the need for a Gerald/Mejia charge. Those charges reflect the trial court’s caution when instructing the jury.
- F -
We are unpersuaded also by defendant’s contention that he was entitled to a Mejia charge. Defendant bases that contention on the State’s acknowledgment that when defendant entered the victim’s apartment, he intended to commit only burglary, not murder. Similarly unpersuasive is defendant’s attempt to derive support from the State’s argument, made pursuant to N.J.S.A. 2C:11-3c(4)(c), that defendant, when committing aggravated assault on the victim, inflicted some blows not to kill, but to cause pain. Even if defendant did not intend to kill Schnaps when he first entered her apartment, the evidence reveals that he changed his mind once he began to assault her. Whether this occurred in a “brief moment,” as the dissent argues, or not is irrelevant. Post at 301, 699 A.2d at 686.
Furthermore, a Gerald/Mejia charge is not required simply because the prosecution relies on N.J.S.A. 2C:11-3c(4)(c). Overwhelming evidence establishes that even if defendant struck the first few blows merely to cause pain, he struck the others to kill. The trial court did not err in failing to tell the jury that it could return a non-unanimous verdict on whether defendant intended to kill. Defendant’s attack was “so wantonly brutal that the jury could have concluded only that the defendant intended to cause death.” Mejia, supra, 141 N.J. at 488, 662 A.2d 308.
- III -
Defendant contends that the trial court’s instruction on murder improperly precluded the jury from considering the non-capital offense of felony murder. Specifically, defendant alleges that the instruction and the verdict sheet prevented the jury from *153considering felony murder until after it first had found him guilty of purposeful-or-knowing murder. Defense counsel did not object to the charge or the verdict sheet. Hence, the issue arises as a matter of plain error under Rule 2:10-2. The issue is whether the error “is clearly capable of producing an unjust result.” We recognize that the charge was flawed, but conclude that the error was not capable of producing such a result.
- A-
Defendant was indicted “for purposely or knowingly causing Schnappes [sic] death or serious bodily injury resulting in death.” The indictment also included counts for second-degree robbery and second-degree burglary. In addition, the State relied on aggravating factor N.J.S.A. 2C:11-3c(4)(g) (the murder was committed during the course of a robbery and a burglary). Although defendant was neither indicted for, nor charged with felony murder, N.J.S.A. 2C:11-3a(3), the court, in accordance with Purnell, supra, 126 N.J. at 530-34, 601 A.2d 175, instructed the jury on felony murder. A conviction for felony murder, however, does not render a defendant eligible for the death penalty. N.J.S.A. 2C:11-3c; Dixon, supra, 125 N.J. at 255, 593 A.2d 266. In Purnell, this Court held that the death penalty could not be imposed for a murder that was committed in the course of a felony if the jury was not permitted to consider the non-capital verdict of felony murder. Although the trial court here gave a felony-murder charge, the verdict sheet makes clear that the jury could not have considered a verdict on felony murder without first finding defendant guilty of purposeful-or-knowing murder.
- B -
The court instructed the jurors that page one of the verdict sheet provided them with four choices when considering the murder charge. They could find defendant not guilty, guilty of murder, guilty of aggravated manslaughter, or guilty of reckless *154manslaughter. Page one did not present felony murder as an option.
The court instructed the jurors that if they found defendant not guilty of murder, they were to ignore pages two and three of the verdict sheet, and proceed directly to page four, where they were to consider the robbery count. Only if the jury found defendant guilty of murder were they to proceed to pages two and three. Page two directed the jury to determine whether defendant had committed the Wiling by his own conduct and purposely or knowingly. If the jury answered affirmatively, the verdict sheet directed them to page three. That page directed the jury to the Gerald issue, whether defendant had acted with the intent to Wll or to inflict serious bodily injury. It also directed the jury to consider whether defendant was guilty of felony murder. Thus, unless the jurors first found defendant guilty of purposeful-or-knowing murder, they would not have reached the felony-murder alternative. In effect, the verdict sheet distracted the jury from convicting the defendant of felony murder, but not purposeful-or-knowing murder.
In a capital case, “the jury must be given every opportunity to convict of the charge not carrying the death penalty.” Mejia, supra, 141 N.J. at 484, 662 A.2d 308 (quoting John M. Cannel, New Jersey Criminal Code, Annotated, Comment 13 to N.J.S.A. 2C:1-8(e)). Here, the combination of the jury charge and the verdict sheet led the jury away from rendering a non-capital verdict of felony murder. That was error.
On the facts of this case, however, the flaws in the charge and verdict sheet do not constitute plain error. Defendant has not advanced any plausible version in which the jury could have convicted him of felony murder without also convicting him of purposeful-or-knowing murder. Nor can we ascertain any such version from the record. In sum, the error was not clearly capable of producing an unjust result.
*155- IV-
Following a Rule 104 hearing, the trial court permitted the prosecution to admit into evidence the results of a DNA test kit known as the “PM” or “polymarker” test. The kit, which is manufactured by Roche Laboratories and marketed by the Per-kin-Elmer Corporation, is sold under the trade name “AmpliType PM.” This polymarker evidence was an important link in tying defendant to the crime. We accept the admission into evidence of results of polymarker testing, which is used primarily on small samples of genetic material, such as blood stains.
On this record, we conclude that the scientific community generally accepts polymarker testing, including dot-intensity analysis. Generally speaking, dot-intensity analysis is a means of identifying a single sample from two sources, such as a blood stain that contains the blood of two people. Such samples may be present at the scene of a violent crime.
The State’s experts identified the genetic markers for the victim and the perpetrator. Through polymarker testing they determined that the box-spring sample, which contained the victim’s blood, also contained blood with the same genetic markers as defendant’s blood. The experts concluded that defendant’s genotype for the genetic markers was shared by only one-in-1,400 African Americans. We hold that the trial court did not err in admitting the testimony of the State’s experts about the results of the DNA tests. The weight of that evidence was for the jury.
DNA testing is an evolving science. The general acceptance or rejection of a test may change over time. Even a test that is accepted generally, moreover, may attract critics. One generally accepted DNA test involves restriction fragment length polymorphism (RFLP) analysis. Because the blood sample in this case was so small, the State’s experts could not use RFLP analysis. Instead, they used tests based on a Polymerase Chain Reaction (PCR): the HLA DQ ALPHA (DQ Alpha) and polymarker (PM) tests.
*156We begin with a basic explanation of DNA. The explanation necessarily uses technical terms and describes scientific methods. Our purpose is to discuss the basic concepts to the extent necessary for comprehension of the trial court’s decision to admit the DNA evidence.
In the course of our discussion, we shall review objections raised by defendant and the dissent to the admission of the DNA evidence. Generally, the defense repeats arguments rejected by the trial court. The dissent, however, raises several objections not raised by the defense either at trial or on this appeal. Both the defense and the dissent share the objective of precluding the admission of DNA evidence proving that defendant murdered Irene Schnaps. Our scrutiny of the record leads to the conclusion that the trial court did not err in admitting the DNA evidence. The weight of the evidence was a matter for the jury.
- A-
1. Deoxyribonucleic Acid (DNA)
Deoxyribonucleic acid (DNA) is a molecule of genetic materials shaped like a double-helix or spiral ladder. In every person, each cell with a nucleus contains a copy of that person’s DNA. Thus, DNA serves as a blueprint for the human body.
The sides of the DNA helix or ladder are composed of two chains comprised of sugars and phosphates. Rungs or steps connect the two sides of the ladder. The rungs consist of pairs of molecules called “bases” or “nucleotides,” which consist of four types: adenine (A), cytosine (C), guanine (G), and thymine (T). Nucleotides from separate DNA strands bond in a specific order to form the rungs that connect the sides of the DNA ladder. C bonds only with G, and A bonds only with T. Thus, for example, if the nucleotides on one strand are CGAT, the corresponding nucleotides on the attached strand will be GCTA.
The order of the base pairs along the DNA molecule comprises an individual’s genetic code. Human DNA contains approximately *157three to four billion base pairs, known as the “genome.” These base pairs govern the production of bodily proteins.
A gene is a sequence of nucleotides on a DNA strand responsible for producing a particular protein. The sequence of the nucleotides can vary. The possible sequences or variations are called “alleles.” Thus, an allele is simply a version of a gene.
A gene’s position on a chromosome is its locus. In different individuals, genes may be “polymorphic,” meaning that they may take different forms or contain different sequences of base pairs. The polymorphic genes, which vary from one person to another, provide the basis for DNA identification. Most DNA has no known function, but even non-functional DNA remains important in forensic analysis.
During mitosis, or cell division, each chromosome is copied. The paired nucleotides separate, dividing the chromosome’s DNA molecule into two separate strands. Free-floating nucleotides attach to the exposed nucleotides of the separated strands in accordance with the G-C, A-T rule. Thus, each strand reconstitutes identical DNA molecules. When the cell divides, these two identical chromosomes enter newly-created “daughter” cells. Each new cell has the identical genetic composition as the original cell.
All cells contain the same chromosomal composition. No two individuals, except identical twins, have the same nucleotide sequences throughout their DNA. DNA testing conducted on cells from various parts of the same body, whether blood, skin, semen, saliva, or hair will yield the same results. As in this case, DNA analysis can help identify donors of genetic material, such as blood.
2. Restriction Fragment Length Polymorphism, (RFLP)
At present, the most widely accepted DNA test is the RFLP analysis. See, e.g., Fishback v. People, 851 P.2d 884, 892 (Colo.1993) (holding that “no serious dispute exists as to whether the *158techniques involved in RFLP analysis are generally accepted”); State v. Moore, 268 Mont. 20, 885 P.2d 457, 468 (1994) (concluding that “the theory underlying DNA and RFLP technology is generally not open to serious attack and [] such evidence is widely admitted in various state and federal courts and jurisdictions”); State v. Streich, 163 Vt. 331, 658 A.2d 38, 48 (1995) (noting that “we cannot find any recent decision under any standard of admissibility which refuses to admit the DNA match result based on” the RFLP technology). Recently, in State v. Marcus, the Appellate Division recognized that the scientific community generally accepts RFLP analysis. 294 N.J.Super. 267, 683 A.2d 221 (1996). As Judge Skillman stated, “DNA testing by the RFLP method is generally accepted and is sufficiently reliable to warrant its admission in criminal cases.” Id. at 285, 683 A.2d 221. RFLP was not the DNA analysis employed in this ease. A brief description of RFLP, however, may be useful as background information.
RFLP focuses on non-functional regions of DNA known as variable-number tandem repeats (VNTRs). In these regions, which typically range from 500 to 10,000 pairs of nucleotides, a core sequence of approximately fifteen to thirty-five base pairs is repeated many times consecutively along the chromosome. The number of repeats varies among individuals. At a given locus or site on a chromosome, sequences with different numbers of repeated units are known as VNTR alleles. Because different VNTR alleles contain different numbers of repeats, these alleles can be identified by their lengths. National Research Council, The Evaluation of Forensic DNA Evidence 14-15 (1996) (NRG Report).
In RFLP analysis, the recovered DNA sample and the sample from the suspect are treated with a restriction enzyme, which seeks out a specific nucleotide pattern on the DNA helixes. It then fragments the molecules at those sites. Because of VNTRs, the locations of these sites, and the lengths of the resulting fragments, differ among individuals. Through a process called “gel electrophoresis” the DNA fragments are sorted by size and *159split into single strands. These strands bond to a nylon membrane, where a specially treated and radioactively-tagged single strand of DNA, called a “genetic probe,” is applied. The genetic probe bonds with a targeted VNTR sequence.
The nylon membrane is then placed in contact with a, piece of X-ray film. The radioactivity of the probes exposes the film, producing a pattern of bands, like the bar-code on a box in a supermarket, where the probes have attached to VNTRs. This bar-code image is called an “autoradiograph” or “autorad.”
Fragments from different donors contain different numbers of repeat units, with a corresponding variation in the lengths of the fragments. Typically, radioactive probes need days or even weeks to expose the film. Id. at 18. Generally speaking, RFLP testing is time-consuming and may require months for a complete analysis. Ibid.
Comparison of the location of the bands reveals whether the targeted VNTR in the subject’s DNA matches the DNA from the recovered genetic material. That analysis can lead with a high degree of certainty to a correlation between the DNA samples.
The next step involves analysis of population statistics, which reveals the likelihood of a random match between the samples. Using single-locus probes, the probability of finding a random match between unrelated individuals on all bands of a DNA fingerprint is less than one in ten million. Using one multi-locus probe, the probability is about one in thirty-three billion. Thomas M. Fleming, Annotation, “Admissibility of DNA Identification Evidence,” 84 A.L.R.4th 313, 324 (1991).
One problem with RFLP testing is that it requires a large quantity of high-quality genetic material. For example, it requires at least a quarter-sized blood stain or a dime-sized semen stain. Unless those samples are recovered when relatively fresh, they will degrade into fragments too small for RFLP analysis. Id. at 320. Cellmark attempted RFLP testing in this case. The samples, however, were too degraded to permit RFLP analysis. *160Thus, Cellmark turned to a newer technology, which involved Polymerase Chain Reaction (PCR).
3. PCR
When, as here, the quantity or quality of genetic material recovered from a crime scene is insufficient to allow RFLP analysis, forensic scientists have used the PCR process to amplify the DNA to produce an amount suitable for testing. The PCR process can copy a segment of DNA millions of times. NRC Report, supra, at 22-23. With the resulting genetic product, scientists can conduct “allele-specific probe” analysis. Fleming, supra, 84 A.L.R.4th at 322-23.
The PCR process copies DNA fragments similar to the way DNA replicates itself during mitosis. Through heating the DNA sample in a thermal cycler, the process separates the helix into separate strands. Primers composed of short DNA segments are added to define the target sequence of DNA. Then, a basic solution containing the enzyme DNA polymerase and the four basic nucleotides are added to the primed DNA sample. The added nucleotides pair with the exposed nucleotides on the separated target-strands in accordance with the G-C, A-T pairing rule. From the original DNA segment, two identical segments result. The thermal cycler runs through its cycle approximately thirty-two times, amplifying the original sample by a factor of two billion. Currently, PCR technology effectively amplifies only small regions of DNA. Accordingly, PCR cannot be used to amplify longer VNTRs for RFLP testing. NRC Report, supra, at 69-70.
PCR-based testing methods have several advantages over RFLP analysis. They are relatively simple processes and can yield results within a short period of time, often within twenty-four hours. Of particular importance to the present case, the PCR process also makes possible DNA tests on small amounts of genetic material.
A disadvantage of PCR-based tests, however, is that the identified genes have fewer alleles than VNTRs. Hence, scientists must *161examine more loci to produce the same amount of information about the likelihood that two individuals share a profile. Id. at 71. Also, some of the loci examined by PCR-based tests are functional genes. Unlike non-functional VNTR markers, functional genes are more susceptible to natural selection, a susceptibility that might undermine their usefulness in matching DNA samples. Ibid. Contamination also is of concern in PCR testing. The technology is so efficient that even small contaminants can be replicated along with the targeted DNA. Ibid. Cellmark used two types of PCR-based tests in defendant’s case: the HLA DQ Alpha (DQ Alpha) and polymarker (PM) tests.
a. HLA DQ Alpha Test
The DQ Alpha test reveals an individual’s form of alleles for the human-leukocyte-antigen DQ Alpha gene. The purpose of the DQ Alpha test is to idéntify the genotype or the two alleles that comprise the DQ Alpha gene present in the DNA sample. That result is then compared with the DQ Alpha genotype of the suspect. If the genotypes match, then the suspect cannot be excluded as a possible donor. Genetics population databases then produce the frequency with which the suspect’s genotype appears in the population. Although eight alleles have been identified at the DQ Alpha locus, only six are commonly used in forensic work. Ibid. Each of those six alleles can be distinguished by specific enzyme probes. Ibid. The six alleles for DQ Alpha are denominated as 1.1, 1.2, 1.3, 2, 3, and 4. For the DQ Alpha gene, there are twenty-one possible pair combinations or genotypes.
To interpret the results, the test uses a test strip with six chemical dots. Each dot consists of a specific enzyme probe that selectively binds to one of the six DQ Alpha alleles. Because the probes, rather than the DNA, are fixed on the membrane, this is called a “reverse” blot. Ibid. This test strip is then immersed in a solution containing the PCR product. The alleles for DQ Alpha present in the PCR product then attach to their corresponding enzyme probe on the test strip. Where the alleles bond, the dots *162turn blue. Two of the six dots will turn blue to indicate which two alleles constitute an individual’s genotype. If an individual is homozygous, meaning that the two DQ Alpha alleles are identical, only one dot will turn blue.
The DQ Alpha test performed on defendant’s blood revealed that his genotype for DQ Alpha is 4,4 or homozygous. Approximately 17% of the entire population (about one-in-six people) and 11.9% of the African American population (about one-in-eight) share that genotype.
DQ Alpha tests performed on a blood stain from the box-spring fabric indicated the presence of the DQ Alpha alleles 2 and 4. Schnaps’s genotype for DQ Alpha was 2,4.
If the blood stain on the box spring were from a single donor, defendant could have been excluded because he does not possess the 2 allele. Schnaps, whose genotype matched the alleles, however, could not be excluded. If, however, the blood on the box spring is from a mixed sample, i.e., from more than one donor, then defendant cannot be excluded. The reason is that the 4 allele, which defendant possessed, was present in the blood stain. Based on other evidence, the prosecution established that the box-spring stain was a mixed sample.
Courts in New Jersey and in other jurisdictions have admitted results of PCR technology and the DQ Alpha test as scientifically reliable. See State v. Dishon, 297 N.J.Super. 254, 277, 687 A.2d 1074 (App.Div.), certif. denied, 149 N.J. 144, 693 A.2d 112 (1997) (finding that testimonial hearing was unnecessary to establish the general acceptance of DQ Alpha DNA testing); State v. Williams, 252 N.J.Super. 369, 381, 599 A.2d 960 (Law Div.1991) (holding that “overwhelming and persuasive evidence” pointed to general acceptance of PCR process and DQ Alpha test); State v. Haddock, 257 Kan. 964, 897 P.2d 152 (1995) (admitting results of DQ Alpha test as scientifically reliable); People v. Palumbo, 162 Misc.2d 650, 618 N.Y.S.2d 197 (Sup.Ct.1994) (same); State v. Gentry, 125 Wash.2d 570, 888 P.2d 1105, cert. denied, — U.S. -, 116 S.Ct. 131, 133 L.Ed.2d 79 (1995) (same).
*163Defendant does not contest the admissibility of the PCR/DQ Alpha test results. He objects, however, to the admissibility of the results of the polymarker test, the validity of which the dissent concedes. Post at 236, 699 A.2d at 654. Because of defendant’s objection, we briefly discuss the polymarker test.
b. The Polymarker (PM) Test
The PM test works like the DQ Alpha test, but instead of testing for the composition of one gene, it tests for six genes. The six genes tested in the PM test are: LDLR flow density lipoprotein receptor), GYPA (glycophorin A), HBGG (hemoglobin gamma-globulin), D7S8, and GC (Group Component). Each of those genes consists of combinations of either two or three different alleles. A blue-dot test, similar to that used in the DQ Alpha test, determines the genotype for each gene. As explained by Ms. Cooper, one of the State’s experts, the PCR test begins by amplifying the amount of DNA. Then, the DNA is passed over a polymere test strip. When the DNA finds its type on that locus, the dot changes color. Each combination of alleles is associated with a population frequency that is expressed as a percentage.
After ascertaining a sample’s genotypes and population frequencies for each of the five individual genes, a mathematical formula known as the “product rule” reveals the likelihood that another individual in the relevant population would share the test subject’s genotype for all five targeted genes. The product rule, which gives the profile frequency in a population as a product of coefficients and allele frequencies, rests on the assumption that a population can be treated as a single, randomly mating unit. NRC Report, supra, at 5. Under the product rule, the population frequencies for each of the six genotypes are multiplied by one another.
Defendant’s genotype for LDLR is present in 56% of the African-American population, his GYPA in 50%, his HBGG in 27%, his D7S8 in 45%, his GC in 17%, and his DQ Alpha in 11.9%. Applying the product rule, the prosecution’s expert testified that *164one-in-170 Afincan Americans would share defendant’s genotype for the LDLR, GYPA, HBGG, D7S8, and GC loci. The product rule equation would be: .56 x .5 x .27 x .45 x .17 = .0058, or, expressed as a ratio, about one-in-170 African Americans. By including defendant’s DQ Alpha genotype frequency in the product rule equation along with the polymarker loci, only one-in-1400 African Americans would share defendant’s composite genotype for all six genes. The product rule equation would be: .56 x .5 x .27 x .45 x .17 x .119 = .00069. Expressed as a ratio, the result is about one-in-1400 African Americans.
4. The Dot-Intensity or Association-of-Alleles Technique
In addition to his general objection to the polymarker test, defendant contends that the polymarker test cannot accurately analyze a mixed-blood sample. At trial, the experts described this method of analysis of interpreting the polymarker test on a mixed-blood sample as involving “the association of alleles.” Scientific articles describe it as “dot-intensity” analysis, a description that we adopt generally in this opinion.
At the pretrial hearing, Dr. Word, one of the State’s experts, testified that the polymarker test is designed so that if a single blood source contributes a pair of heterozygous alleles (i.e., A, B or B, C but not A, A), the two alleles will turn blue in equal intensities. If, however, two or more blood donors contribute the same allele, that dot would be more intense or a darker blue than if an individual donor contributed only one such allele. The same result would follow if one blood donor contributed two of the same allele so as to result in a pair of homozygous alleles (i.e., A, A or B, B but not A, B). In sum, the determination whether a sample is from more than one source of blood depends on the intensity of the color of the dots.
Cellmark performed PM tests on three samples: (1) on a bloody towel, assumed to contain blood from the victim only; (2) on a sample of defendant’s blood; and (3) from a portion of the box-spring cover that contained a mixture of blood. The polymarker *165test on the blood removed from the box spring revealed the presence of both possible alleles (A and B) for LDLR, GYPA, and D7S8, and all three possible alleles (A, B, and C) for HBGG and GC. An individual can possess, at most, two different alleles. Consequently, the presence of three different alleles at the HBGG and GC loci demonstrated that the DNA sample from the box spring was a mixture of blood from more than one individual.
Dr. Word testified that the PM test conducted on the mixed-blood sample from the box spring revealed two distinct sources. Two sets of alleles caused the color imbalances in the dots from the box-spring sample. On the test strip, some of the blue dots appeared darker, but others were lighter. None of the dots, however, was lighter than the control dot. Those dot-intensity imbalances resulted from the presence of some alleles in pairs and other singly.
Using dot-intensity analysis, Dr. Word explained that a theoretical subtraction of the victim’s blood from the PM results of the box-spring sample revealed the genotype of the second subject. She explained that if all three alleles for GC were present (A, B, and C) and the A allele-dot was darker on the test strip, the sample contained two A alleles, one B allele, and one C allele. The genotypes of the contributing donors, then, had to be either AB and AC, or AA and BC. No other combinations consisted of two A alleles, and only one each of B and C.
Based on dot-intensity analysis of the blood stain from the box spring, the State’s experts concluded that the blood could have been a mixture of defendant’s blood with that of Schnaps’s. Cellmark made only two assumptions in its analysis: (1) that Schnaps was a donor to the blood on the box-spring sample; and (2) that the blood was a mixture of two people.
The PM test for GC on the box spring revealed the presence of the A, B, and C alleles. The A dot was darker than the other two. That difference indicated that the composition of the GC in the sample consisted of two A alleles, plus the B and C alleles. If the sample had two donors, the possible combinations were AA and *166BC, or AB and AC. Schnaps was AC type for GC. The remaining donor therefore had to be AB. Defendant’s genotype was AB.
The D7S8 test on the box-spring sample revealed the A and B alleles, with the A dot being darker. The only possible combination, then, was AA and AB. Schnaps was AA type for D7S8. The remaining donor had to be AB. Defendant’s genotype was AB.
The HBGG test revealed the A, B, and C alleles at equal intensities. Because no dot was darker, Dr. Word testified that the possible combinations could be only AA and BC, AB and CC, or AC and BB. Schnaps was BB type for HBGG. Cellmark concluded that the remaining donor, then, had to be AC. Defendant’s genotype was AC.
The GYPA test revealed A and B alleles with the A dot being darker. Schnaps was AA type for GYPA. The other donor had to be AB. Defendant was AB.
The LDLR test revealed A and B alleles with the B dot being darker. Schnaps was AB type for LDLR. The other donor had to be BB. Defendant’s genotype for LDLR was BB.
Based on those test results, the State’s experts concluded that neither Harvey nor Schnaps could be excluded as donors to the box-spring blood sample. As with the RFLP test, infra part rV.A.2., the second step involves analysis of population statistics. Our discussion of the statistics in the present ease is at infra part VI.
- B -
1. Standard of Review
The first question concerns the standard of appellate review of a trial court’s decision on the admissibility of DNA evidence. Generally, appellate courts review a trial court’s determination of the admissibility of evidence for an abuse of discretion. State v. Conklin, 54 N.J. 540, 545, 258 A.2d 1 (1969). The question remains, however, whether the abuse-of-discretion stan*167dard should apply to decisions concerning the admission of novel scientific evidence. We begin by recognizing that the best time to make the record on admission of such evidence is in a Rule 104 hearing.
The party offering the evidence bears the burden of proof. Windmere, Inc. v. International Ins. Co., 105 N.J. 373, 378, 522 A.2d 405 (1987). Trial lawyers must make myriad choices in deciding how best to present or refute novel scientific evidence. Those choices construct the universe of discourse within which the trial court decides whether the scientific community considers the evidence acceptable generally. The choices also influence the presentation of the evidence to the jury and the record on appeal.
Unlike many other evidentiary issues, whether the scientific community generally accepts a methodology or test can transcend a particular dispute. People v. Miller, 173 Ill.2d 167, 219 Ill.Dec. 43, 61, 670 N.E.2d 721, 739 (1996) (McMorrow, J., concurring), cert. denied, — U.S. -, 117 S.Ct. 1338, 137 L.Ed.2d 497 (1997). In determining the general acceptance of novel scientific evidence in one case, the court generally will establish the acceptance of that evidence in other cases. Jones v. United States, 548 A.2d 35, 40 (D.C.1988). Notwithstanding the trial court’s better position to shape the record and make factual determinations, appellate courts retain an important residual role for questions concerning the admission of scientific evidence. Like trial courts, appellate courts can digest expert testimony as well as review scientific literature, judicial decisions, and other authorities. To the extent that general acceptance focuses on issues other than a witness’s credibility or qualifications, deference to the trial court is less appropriate. Miller, supra, 219 Ill.Dec. at 60-62, 670 N.E.2d at 738-40 (McMorrow, J., concurring).
When reviewing a decision on the admission of scientific evidence, an appellate court should scrutinize the record and independently review the relevant authorities, including judicial opinions and scientific literature. In the rapidly changing world of modem science, continuing research may affect the scientific *168community’s acceptance of a novel technology. By reviewing post-trial publications, an appellate court can account for the rapid pace of new technology. The continuing review also recognizes that general acceptance may change between the time of trial and the time of appellate review. State v. Bible, 175 Ariz. 549, 858 P.2d 1152, 1189 n. 33 (1993), cert. denied, 511 U.S. 1046, 114 S.Ct. 1578, 128 L.Ed.2d 221 (1994); see also Hadden v. State, 690 So.2d 573, 579 (Fla.1997) (finding that an appellate court “should consider the issue of general acceptance at the time of appeal rather than at the time of trial”). But see Lindsey v. People, 892 P.2d 281, 290-91 n. 25 (Colo.1995) (reasoning that Frye v. United States, 293 F. 1013 (D.C.Cir.1923), “requires nothing more than general acceptance at the time the evidence is admitted”). Moreover, by examining such additional information, an appellate court can prevent any injustice rendered by the admission or exclusion of the evidence at the trial level. Bible, supra, 858 P.2d at 1189 n. 33 (stating that if “the result obtained is the product of invalid scientific theory, there is no good reason to accept it simply because we were fooled at the inception of the inquiry”).
On this appeal, we do not decide whether a different standard of appellate review should apply to a trial court’s decision to admit or exclude expert testimony in civil cases, where the focus is not on whether the scientific evidence is generally accepted, but rather whether it derives from a reliable methodology supported by some expert consensus. Landrigan v. Celotex Corp., 127 N.J. 404, 417, 605 A.2d 1079 (1992); Rubanick v. Witco Chem. Corp., 125 N.J. 421, 449, 593 A.2d 733 (1991). In Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), the United States Supreme Court adopted essentially the same standard for federal courts in both civil and criminal cases. The question of the appropriate standard of appellate review for federal courts is currently before the United States Supreme Court. Joiner v. General Elec. Co., 78 F.3d 524 (11th Cir.1996), cert. granted, — U.S. -, 117 S.Ct. 1243, 137 L.Ed.2d 325 (1997). We restrict our analysis to the standard of review of evidence that the trial court has found to be generally accepted.
*1692. Standard for Admission of Expert Testimony
New Jersey Rule of Evidence 702, which is virtually identical to former Evid. R. 56(2), governs the admission of expert testimony. The rule provides:
If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education may testify thereto in the form of an opinion or otherwise.
In effect, this rule imposes three basic requirements on the admission of expert testimony:
(1) the intended testimony must concern a subject matter that is beyond the ken of the average juror;
(2) the subject of the testimony must be at a state of the art such that an expert’s testimony could be sufficiently reliable; and
(3) the witness must have sufficient expertise to explain the intended testimony.
[State v. Kelly, 97 N.J. 178, 208, 478 A.2d 364 (1984); N.J.R.E. 702, 1991 Supreme Court Committee Comment.]
Defendant does not contest that DNA testing is beyond the ken of the average juror. Likewise, the defense does not dispute the qualifications of Ms. Cooper or Dr. Word as experts in the field of DNA testing. The sole issue is whether the scientific community sufficiently accepted the DNA tests to justify admission of the testimony of the State’s experts.
In criminal cases we continue to apply the general acceptance or Frye test for determining the scientific reliability of expert testimony. In Frye, supra, the court wrote:
[W]hile courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs.
[293 F. at 1013-14 (emphasis added).]
In 1993, the United States Supreme Court abandoned Frye's general-acceptance standard as the exclusive test for admitting scientific testimony in favor of the more relaxed standards of Federal Rule of Evidence 702. Daubert, supra, 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469. Even before the United States Supreme Court decided Daubert, this Court had relaxed the test *170for admissibility of scientific evidence in toxic-tort cases. Landrigan, supra, 127 N.J. at 404, 605 A.2d 1079; Rubanick, supra, 125 N.J. at 421, 593 A.2d 733. We have been cautious in expanding the more relaxed standard to other contexts. State v. Fertig, 143 N.J. 115, 126, 668 A.2d 1076 (1996); State v. Spann, 130 N.J. 484, 509-10, 617 A.2d 247 (1993); State v. J.Q., 130 N.J. 554, 572-73, 617 A.2d 1196 (1993). Thus, the test in criminal cases remains whether the scientific community generally accepts the evidence. Spann, supra, 130 N.J. at 509, 617 A.2d 247; Windmere, supra, 105 N.J. at 386, 522 A.2d 405.
A proponent of a newly-devised scientific technology can prove its general acceptance in three ways:
(1) by expert testimony as to the general acceptance, among those in the profession, of the premises on which the proffered expert witness based his or her analysis;
(2) by authoritative scientific and legal writings indicating that the scientific community accepts the premises underlying the proffered testimony; and
(3) by judicial opinions that indicate the expert’s premises have gained general acceptance.
[Kelly, supra, 97 N.J. at 210, 478 A.2d 364 (citing State v. Cavallo, 88 N.J. 508, 521, 443 A.2d 1020 (1982)).]
The burden to “clearly establish” each of these methods is on the proponent. Williams, supra, 252 N.J.Super. at 376, 599 A.2d 960.
Courts have applied this test in various contexts to evaluate the reliability of scientific evidence. See, e.g., Kelly, supra, 97 N.J. at 209, 478 A.2d 364 (admitting expert testimony relating to battered woman’s syndrome); State v. Zola, 112 N.J. 384, 412-13, 548 A.2d 1022 (1988) (admitting expert testimony that modified-chemical test detected presence of saliva on victim), cert. denied, 489 U.S. 1022, 109 S.Ct. 1146, 103 L.Ed.2d 205 (1989); Windmere, supra, 105 N.J. at 373, 522 A.2d 405 (concluding that voice-print evidence does not derive from reasonably reliable scientific method); Romano v. Kimmelman, 96 N.J. 66, 82, 474 A.2d 1 (1984) (holding breathalyzer scientifically reliable); State v. Hurd, 86 N.J. 525, 432 A.2d 86 (1981) (admitting hypnotically refreshed testimony when subjected to strict safeguards ensuring reliability of hypnotic *171procedure); State v. King, 215 N.J.Super. 504, 518-20, 522 A.2d 455 (App.Div.1987) (finding isoenzyme test, which detects presence in blood of six distinct enzyme systems, held scientifically reliable); Williams, supra, 252 N.J.Super. at 378-83, 599 A.2d 960 (holding PCR/DQ Alpha test and Gm/Km blood tests sufficiently reliable to be admitted at trial).
3. General Acceptance
Proof of general acceptance within a scientific community can be elusive. Windmere, supra, 105 N.J. at 379, 522 A.2d 405. Satisfying the test involves more than simply counting how many scientists accept the reliability of the proffered technology. Williams, supra, 252 N.J.Super. at 375, 599 A.2d 960. Proving general acceptance “entails the strict application of the scientific method, which requires an extraordinarily high level of proof based on prolonged, controlled, consistent, and validated experience.” Rubanick, supra, 125 N.J. at 436, 593 A.2d 733. Essentially, a novel scientific technique achieves general acceptance only when it passes from the experimental to the demonstrable stage. Windmere, supra, 105 N.J. at 378 n. 2, 522 A.2d 405.
General acceptance, however, does not require complete agreement over the accuracy of the test or the exclusion of the possibility of error. See Richard J. Biunno, Current N.J. Rules of Evidence, Comment 4 to N.J.R.E. 702; State v. Johnson, 42 N.J. 146, 171, 199 A.2d 809 (1964). Neither is it necessary to demonstrate that the techniques, methodology, and procedures are infallible. Nor is it necessary that acceptance within the scientific community be unanimous. State v. Tate, 102 N.J. 64, 83, 505 A.2d 941 (1986). Every scientific theory has its detractors. Windmere, supra, 105 N.J. at 379, 522 A.2d 405. Here, the State’s burden is to prove that the polymarker test and the interpretation of its results are non-experimental, demonstrable techniques that the relevant scientific community widely, but perhaps not unanimously, accepts as reliable.
*1724. Polymarker Test
A proponent of scientific evidence can prove general acceptance through expert testimony, publications, or judicial opinions.
a. Expert Testimony
At the pretrial hearing, Dr. Word testified that the scientific community generally accepts the polymarker test as scientifically reliable. The dissent agrees. Post at 236, 699 A.2d at 654. Dr. Word explained that the polymarker test is technologically and procedurally similar to the well-established PCR/DQ Alpha test. Cellmark, moreover, conducted validation studies on the PM test. The studies, which were conducted in accordance with protocols established by the Technical Working Group on DNA Analysis Methods (TWGDAM), revealed that the PM test reliably produced accurate results. Approximately thirty to forty laboratories in the United States use the PM test either for casework or validation studies. Six of the better-known laboratories independently had tested the polymarker and likewise concluded that it was reliable. Those laboratories included: the Federal Bureau of Investigations (FBI), the Department of Justice, the Georgia Bureau of Investigations, the California Laboratory in Berkeley, the Center for Blood Research in Boston, the Regional Crime Laboratory at Indian River Community College, and Roche Biomedical Laboratories. Of all the laboratories using the PM kit, none had ever reported receiving an incorrect result from a properly performed test.
Defendant challenged the general acceptance of the polymarker test, asserting that some laboratories still subject it to validation studies. Defendant’s expert at the Rule 104 hearing, Dr. Robert Shaler, testified that polymarker testing was not yet appropriate for use in casework and that Cellmark’s validation procedures were flawed.
The trial court concluded:
Based upon the expert testimony of Dr. Charlotte Word, and the scientific literature admitted into evidence, PCR/PM has been sufficiently validated to be *173used in casework; has been accepted by the relevant scientific community; and produces uniform and reasonably reliable results.
Our review of the record leads us to conclude that the scientific community has generally accepted the polymarker test. Dr. Word established that independent tests have validated the polymarker’s reliability and that highly-regarded laboratories have employed the test in casework. Only defendant doubts the reliability of the PM test.
b. Publications
Scholarly and scientific publications, moreover, approve the polymarker test. In support of the polymarker test, the State submitted one published article, George Herrin, Nicola Fildes & Rebecea Reynolds, “Evaluation of the AmpliType PM DNA Test System on Forensic Case Samples,” 39 J. Forensic Sci. 1247 (1994) (Herrin article), and two articles that since have been published: Nicola Fildes & Rebecca Reynolds, “Consistency and Reproducibility of AmpliType PM Results Between Seven Laboratories: Field Trial Results,” 40 J. Forensic Sci. 279 (1995) (Roche field study); Bruce Budowle, et al., “Validation and Population Studies of the Loci LDLR, GYPA, HBGG, and D7S8, and Gc (PM Loci), and HLA-DQx Using A Multiplex Amplification And Typing Procedure,” 40 J. Forensic Sci. 45 (1995) (FBI Report). All three of those documents concluded that the PM test was reliable for casework. Further, the State submitted a list of forty-four presentations, posters, lectures, seminars, and workshops in which forensic scientists discussed issues regarding polymarker-related research, testing, and results. Dr. Word had attended fifteen to twenty such lectures and had lectured on the reliability of the polymarker test at many of the meetings.
Defendant argues that the scholarly works were too few to support a conclusion of general acceptance. Further, defendant maintains that lectures do not provide effective forums for peer review. In addition, defendant alleges that Cellmark violated the national standards for DNA testing by failing to publish both the *174results of its validation studies and the types of primers and probes that it uses.
The articles and lectures, however, support general acceptance of the polymarker test. Admittedly, prior cases have involved greater bodies of accumulated scholarship. See, e.g., Kelly, supra, 97 N.J. at 211, 478 A.2d 364 (noting that there were “at least five books and almost seventy scientific articles and papers about the battered-woman’s syndrome”); Williams, supra, 252 N.J.Super. at 382, 599 A.2d 960 (noting that six testifying experts had authored, cumulatively, close to 600 articles, many on PCR technology). We recognize, additionally, the correlation between the number of published articles and the general acceptance of a subject. Yet, we never have required a specific number of articles to satisfy the test of general acceptance. Rather, our focus always has been whether existing literature reveals a consensus of acceptance regarding a technology. In Windmere, supra, 105 N.J. at 383-84, 522 A.2d 405, for example, although ample literature existed on the subject of voiceprint analysis, the existing “journals [were] in disarray.” By contrast, all of the written works on the polymarker test agree that the test is reliable. Defendant produced no documents, published or pre-publication, suggesting anything to the contrary.
Further, “[u]nder appropriate circumstances, speeches, addresses, and other similar sources may be used to demonstrate the acceptance of a premise by the scientific community.” Kelly, supra, 97 N.J. at 211 n. 17, 478 A.2d 364; see also State v. Anderson, 118 N.M. 284, 881 P.2d 29, 42-43 (1994) (rejecting claim that presentations, as opposed to formal publications, were inadequate as peer review). From the forty-four lectures and presentations, it appears clear that the forensic-science community has kept abreast of developments and has had adequate opportunity for peer review of PM testing.
Since the conclusion of defendant’s trial, moreover, the National Research Council, the members of which are drawn from the councils of the National Academy of Sciences, the National Acade*175my of Engineering, and the Institute of Medicine, published The Evaluation of Forensic DNA Evidence (referred to supra as NBC Report), an update to their 1992 report, DNA Technology in Forensic Science. The report concludes that PCR-systems, including the polymarker, are “ready to be used.” NRC Report, supra, at 119. The report further confirms that the polymarker has been “validated with tests for robustness with respect to environmental insults,” and that “substantial information on population frequencies” exists for the polymarker loci. Id. at 72. We conclude that existing literature, combined with research shared at lectures and symposia, supports a finding that the polymarker test is generally accepted as reliable by the forensic science community.
c. Judicial Opinions
At the time of the Rule 104 hearing, both the State and the defense were unaware of any judicial opinion discussing polymarker evidence. See Wilkerson v. Pearson, 210 N.J.Super. 333, 336, 509 A.2d 818 (Ch.Div.1985) (holding that absence of judicial opinions demonstrating acceptance by other courts of particular type of scientific technique should not, by itself, foreclose finding of general scientific acceptance and reliability). Before the Rule 104 hearing, however, a New York court had admitted polymarker evidence. People v. Morales, N.Y.L.J., Oct. 26, 1994, at 34 (N.Y.Cty.Ct.1994), aff'd, 227 A.D.2d 648, 643 N.Y.S.2d 217, appeal denied, 677 N.E.2d 301 (1996). In Morales, experts from the Center for Blood Research Laboratories, Yale University School of Medicine’s Department of Genetics, and the Office of the Chief Medical Examiner for New York testified in support of admission of the evidence. Curiously, the witness from the New York Medical Examiner supporting the admission of the polymarker evidence was Dr. Shaler, the same expert who testified against admission of polymarker evidence in the present case. The New York court concluded that “the People have met their burden in establishing that the PCR tests at issue here are sufficiently established to gain general acceptance in the scientific community and satisfy the standard of reliability.” Ibid.
*176Since defendant’s trial in the present case, at least six other courts have held that polymarker testing is scientifically reliable. United States v. Beasley, 102 F.3d 1440, 1448 (8th Cir.1996), cert. denied, — US. -, 117 S.Ct. 1856, 137 L.Ed.2d 1058 (1997) (holding that DQ Alpha and polymarker testing are sufficiently reliable under Daubert and have achieved general acceptance within relevant scientific community); United States v. Shea, 957 F.Supp. 331, 338 (D.N.H.1997) (finding PCR testing, including polymarker testing, reliable under F.R.E. 702); United States v. Lowe, 954 F.Supp. 401, 418 (D.Mass.1996) (finding that polymarker and another PCR-based test, D1S80, are sufficiently reliable under Daubert); Brodine v. State, 936 P.2d 545, 550-51 (Alaska.Ct.App.1997) (finding polymarker testing generally accepted in scientific community); People v. Pope, 284 Ill.App.3d 695, 220 Ill.Dec. 309, 314, 672 N.E.2d 1321, 1326 (1996) (finding that DQ Alpha and polymarker typing are generally accepted in scientific community under Frye); Keen v. Commonwealth, 24 Va.Ct.App. 795, 485 S.E.2d 659, 664 (1997) (rejecting defendant’s challenges to the polymarker test). In Pope, supra, the Illinois Court of Appeals found polymarker testing generally accepted in the scientific community even when the Frye hearing in that ease involved the testimony of only one witness, the State’s expert. Pope, supra, 220 Ill.Dec. at 314, 672 N.E.2d at 1326. Admission of the polymarker test in other jurisdictions supports our conclusion that the trial court correctly admitted the evidence in the present case.
We thus conclude that the trial court did not err in admitting expert testimony on the results of the polymarker test. We are satisfied that the polymarker technology is scientifically reliable and that Cellmark conducted the tests in accordance with established procedures.
5. Dotr-Intensity Analysis
Defendant asserts that even if the results of the polymarker test are admissible, dot-intensity analysis, the State’s method of interpreting the mixed-blood sample, is scientifically unreliable. The *177dissent likewise rejects dot-intensity analysis. Indeed, the dissent goes so far as to assert that if dot-intensity analysis is rehable, “the results obtained would be inconsistent with defendant being the murderer.” Post at 261, 699 A.2d at 667.
At the Rule 104 hearing, dot-intensity analysis was presented as an integral part of polymarker testing. The Rule 104 hearing involved consideration of oral testimony and written evidence on the polymarker test and dot-intensity analysis. Both the defense and the State considered dot-intensity analysis as an application of the polymarker test, not as an issue apart from polymarker testing. In holding that the results of polymarker testing were admissible, the trial court implicitly approved dot-intensity analysis, which was the only use of the polymarker test on which the State relied.
Based on the record, as well as on posttrial publications and judicial opinions, we conclude that the trial court correctly allowed the State’s experts to testify about dot-intensity analysis. In so concluding, we recognize that a court must examine each step of a scientific process or technique. Kelly, supra, 97 N.J. at 210, 478 A.2d 364. Thus, we independently examine dot-intensity analysis to determine whether it has obtained sufficient acceptance to justify admission of its results into evidence.
a. Expert Testimony
At the pretrial hearing, the defense offered the testimony of Dr. Shaler and a written report by Dr. Blake, of Forensic Science Associates, both of whom disagreed with Dr. Word on the propriety of dot-intensity analysis. Dr. Shaler disputed Cellmark’s theory that the dot-intensity imbalances on the polymarker test strip represented the presence of allele pairs. He testified that such color imbalances occurred even on the DNA test strips of the single-donor control samples of defendant’s and Schnaps’s blood. For example, Dr. Shaler claimed to have detected intensity imbalances on defendant’s GYPA and HBGG loci and on the victim’s GC locus. According to the defense, those variances in the dot *178intensities in Harvey’s and Schnaps’s control strips destroyed the integrity of the dot-intensity analysis of the mixed-blood sample taken from the box spring. The dissent agrees, asserting that “[t]o the degree that there are naturally occurring variances in dot intensities that cannot be predicted prior to the analysis, dot-intensity analysis disintegrates as a reliable and useful test.” Post at 169, 699 A.2d at 620-21. According to Dr. Shaler, the PM test of the box spring supported a calculation only of “all the second donor types which would be possible” to fill out the necessary alleles.
Defendant also asserts that color imbalances in the polymarker test strips may have been caused not by allele pairs, but by human errors such as contamination or manufacturing defects. Dr. Shaler, at the pretrial hearing and at trial, faulted Cellmark for not repeating the polymarker test procedure despite the possibility of error or contamination. He further stated that contamination, a manufacturing defect, or improper laboratory procedures caused the dot imbalances.
Our previous holding that the polymarker test is scientifically reliable, supra part IV.B.4.d., leads us to the conclusion that the foregoing challenges to dot-intensity analysis regarding Cellmark’s performance of the polymarker test, concern not the admissibility, but the weight of the evidence. See Marcus, supra, 294 N.J.Super. at 291, 683 A.2d 221 (holding that interpretation of extra bands on autorads developed from bloodstains, “like an expert’s ability to perceive an abnormality on an x-ray, is a matter within the province of the jury”); Fishback, supra, 851 P.2d at 893 (reasoning that defendant’s challenge to techniques of RFLP analysis, including interpretation of autorads, concerns weight and not admissibility of DNA typing evidence under Frye); State v. Schweitzer, 533 N.W.2d 156, 160 (S.D.1995) (reasoning that DNA-expert’s conclusions regarding results of DNA test were issue of weight for jury to consider); State v. Kalakosky, 121 Wash.2d 525, 852 P.2d 1064, 1072 (1993) (holding that defendant’s assertions that specific laboratory procedures utilized to analyze DNA sam*179pie were flawed, goes to weight of evidence, not admissibility). As such, the ultimate determination of these issues was properly left to the jury.
The State’s experts, Dr. Word and Ms. Cooper, differed from the defense expert, Dr. Shaler, in the interpretation of the dot intensities on the control strips. Defense counsel vigorously questioned the State’s experts whether the differences in the dots of the victim’s and defendant’s blood samples undermined the dot-intensity analysis of the sample from the box spring. On cross-examination, Ms. Cooper disputed Dr. Shaler’s claim that dot-intensity imbalances existed at the HBGG and GC loci on the defendant’s and Schnaps’s test strips. Ms. Cooper stated that “[t]he only place where I really see any type of a slight difference would be at the GYPA locus for Mr. Harvey.” Moreover, neither Ms. Cooper nor Dr. Word found any evidence that Cellmark had made a mistake in the conduct of the polymarker test. According to Ms. Cooper, the intensity imbalance on defendant’s GYPA locus was probably due to a variant allele present in some African Americans. Ms. Cooper explained that the difference in dot intensities at the GYPA locus was not due to errors in testing or to a manufacturing defect. Indeed, nothing in the record supports the defense’s conjecture that contamination or manufacturing defects caused any differences in dot intensities.
The dissent goes beyond the defendant’s arguments. For example, the dissent repeatedly concludes that the polymarker test kit used by the State was “designed solely to determine the presence or absence of certain alleles,” so that it at most can “potentially show that a mixture exists in certain circumstances.” Post at 244, 259, 699 A.2d at 658, 666.
The State, however, presented expert testimony at the Rule 104 hearing demonstrating that differences in intensities of the test-strip dots could reveal the blood donors to a mixed-blood sample. Dr. Word testified that the Perkins-Elmer Amplitype PM test strip was designed so that a sample from a single individual would produce similar dot intensities within each locus for which the *180individual is heterozygous. If two or more blood donors contributed the same allele, or one blood donor contributed two of the same allele so as to result in a pair of homozygous alleles, that dot would be more intense than if one of that allele was contributed by an individual donor. Dr. Word further testified that Cellmark’s validation studies included a mixture-dilution study. That study provided Cellmark with information on the dilutions that were capable of detection and showed that the mixtures could be interpreted in many cases.
Contrary to the dissent’s characterization, see post at 255, 699 A.2d at 664, the testimony of the State’s expert supports the general acceptance of dot-intensity analysis. We conclude that the trial court reached the right decision in not preventing the jury from hearing the State’s DNA testimony.
The dissent nonetheless argues that certain omissions in the expert testimony of both parties reveal that dot-intensity analysis is unreliable. It maintains that the State’s inability to use dot-intensity analysis for the DQ Alpha test proves that the analysis does not work. It further states that the “DQ Alpha marker was retested using the PM testing strips.” Post at 265 n. 14, 699 A.2d at 668 n. 14. The record, however, contains no evidence of two sets of DQ Alpha test results. Likewise, the record contains no discussion of the application of dot-intensity analysis to the results of the DQ Alpha test. At trial, the defense did not challenge the absence of any such analysis. Hence, the record does not reveal whether dot-intensity analysis can be performed on a DQ Alpha test or whether the test for the DQ Alpha allele differs from the polymarker test. The Roche field study states that “[O]n DQA1 DNA probe strips, the probes are not as well balanced and the signal intensities cannot be used as easily to associate alleles from different sources” and that “it was easier to determine the major genotype on the PM DNA probe strips than the DQA1 DNA probe strips because the relative intensities of the dots could be compared.” Fields, supra, 40 J. Forensic Sci. at 284; see also Edward Blake, et al., “Polymerase Chain Reaction (PCR) Amplifi*181cation and Human Leukocyte Antigen (HLA)-DQa Oligonucleotide Typing on Biological Evidence Samples: Casework Experience,” 37 J. Forensic Sci. 700, 706 (May 1992) (reasoning that for DQ Alpha test results, a “mixture cannot necessarily be detected, however, if two contributors to a mixture contribute no more than two alleles in total and contribute approximately equivalent amounts of DNA”); NRC Report, supra, at 130 (“Mixed samples can also lead to more complicated calculations with DQA, where some alleles are inferred by subtraction.”).
Finally, the dissent claims that the polymarker results from the HBGG marker of the box-spring sample prove that dot-intensity analysis does not work. Post at 263-265, 699 A.2d at 667-668. The HBGG marker results for the box spring were “A,B,C,” with all three alleles balanced in intensity. Schnaps’s type at the HBGG marker was “B,B.” According to dot-intensity analysis, the B allele should be more intense. Neither at trial nor on appeal, however, has defendant challenged the polymarker results or the dot-intensity analysis of the box-spring sample at the HBGG marker. Consequently, the State never has had an opportunity to meet the challenge. Perhaps defense counsel elected not to challenge the findings that the HBGG locus for strategic reasons. On the record before us, the unchallenged results of the dot-intensity analysis at the HBGG marker do not justify a reversal of defendant’s conviction. In an application for post-conviction relief, defense counsel may make such use of the polymarker results at the HBGG marker as is appropriate.
b. Publications
At the Rule 104 hearing, the State introduced three’ articles: the Roche field study, the FBI Report, and the Herrin article. See supra Part IV.B.4.b. All three articles discuss dot-intensity analysis. Not one states that dot-intensity analysis is unreliable. The articles go no further than to express caution when interpreting the results through dot-intensity analysis.
*182The Roche field study was designed to determine the genotypes of either the major contributor or minor contributor, or both, to a mixed-blood sample. For example, in a mixed sample, where the intensity of the dots corresponding to the minor component was greater than or equal to the control dots, the study could not identify the major or minor component. Consequently, the study placed such a result in the uninterpretable category.
Apparently, the Roche field study was not a study of dot-intensity analysis for mixtures where the relative amounts of DNA for each donor were equal. The study, however, concluded that analyzing the balance of dots on a PM strip could be a “valuable asset of the system for the analysis of mixtures” and directed individual laboratories to develop policies for the interpretation of mixed-blood samples “based on experience and case history information.” Id. at 284-85. Here, Cellmark developed such interpretive methods.
The dissent contends that the FBI Report “flatly contradicts the assumptions of dot-intensity analysis.” Post at 280, 699 A.2d at 676. We read the FBI Report differently. The FBI Report stated that “[a]s a general rule, samples with two contributors of known PM types displayed PM types as predicted.” FBI Report, supra, at 49 (emphasis added). Each contributed to the profile and when the samples with two contributors of known PM types shared alleles in common the dot intensities were increased compared to alleles that were not shared. Ibid. Because the FBI, in its study, already knew the PM types of each donor, dot-intensity analysis was not necessary. Thus, contrary to the dissent’s assertion, the FBI Report did not refuse to draw any conclusions from dot-intensity imbalances. Post at 257, 699 A.2d at 665. Rather, the FBI Report acknowledged that mixtures of donors sharing the same alleles resulted in an increased dot intensity.
The dissent argues further that the Herrin article “did not conclude that dot-intensity analysis was sufficiently reliable to warrant its use in an actual criminal case.” Post at 258, 699 A.2d at 665. In fact, the Herrin article supports dot-intensity analysis, *183especially where the victim’s known allele profile at the relevant locus is homozygous:
Comparison of the dot intensity within loci allowed deduction of types at some loci even when separation of DNA from sperm and non-sperm cells was not complete. This feature may prove to be very useful in routine casework____In this case the non-sperm fraction HBGG locus produced 3 dots with the A allele being approximately twice as intense as the B and C alleles (indicating that at least 2 A alleles were present). By comparison with the known sample from the victim in this case (HBGG AA), it is possible to assign the A alleles to the victim, and know that the B and C alleles originated with another individual(s).
[Herrin et al, supra, 39 J. Forensic Sci. at 1249-50 (emphasis added).]
Recent publications enhance the acceptance of dot-intensity analysis. The NRC Report states:
Mixed samples are sometimes found in crime situations — for instance, blood from two or more persons at the scene of a crime____ In many cases, one of the contributors — for example, the victim — is known, and the genetic profile of the unknown is readily inferred. In some cases, it might be possible to distinguish the genetic profiles of the contributors to a mixture from differences in intensities of bands in an BFLP pattern or dots in a dot-blot typing; in either case, the analysis is similar to the unmixed case.
[NRC Report, supra, at 129 (emphasis added).]
Moreover, Dr. Word recently has published an article validating the use of the PM kit for use on mixed-blood samples. Charlotte J. Word et al., “Summary of Validation Studies from Twenty-Six Forensic Laboratories in the United States and Canada on the Use of the AmpliType PM PCR Amplification and Typing Kit,” 42 J. Forensic Sci. 39 (1997).
Our review of the relevant literature leaves the impression that dot-intensity analysis is an esoteric topic of interest to a limited scientific community. Within that limited community, however, the analysis has been sufficiently accepted to permit the jury to hear about it. We recognize that, as occurred in the trial below, dot-intensity analysis provides an opening for cross-examination and contradictory expert testimony. That opening, however, should not deprive the jury from hearing testimony about the analysis. We also recognize that other eases considering the general acceptance of scientific innovations have involved a greater number of scientific articles. See supra Part IV.B.4.b. No *184published article, however, rejects dot-intensity analysis. As previously discussed, the trend supports acceptance of the test.
c. Judicial Opinions
Although no court has discussed dot-intensity analysis at length, the United States District Court for the District of New Hampshire recently accepted as reliable under Daubert the premise underlying this interpretive method. In Shea, supra, the defendant argued that the polymarker test should be excluded because PCR cannot reliably detect mixtures of more than one person’s DNA. 957 F.Supp. at 339. The district court rejected that argument, finding persuasive the government expert’s testimony that “an examiner -will be able to identify a mixture from observable differences in the relative strengths of the signals indicated on the PCR test strips, except in extremely unusual circumstances.” Id. at 340.
d. New Jersey Rule of Evidence 403
Finally, the dissent argues that even if the DNA evidence derives from a generally accepted methodology, the evidence’s prejudicial effect outweighs its probative value. Post at 270, 699 A.2d at 671. In particular, the dissent challenges the assumption that only two people provided blood to the box-spring sample and that the victim’s and suspect’s blood were present in equal amounts. The argument is more theoretical than real. Nothing in the record suggests the presence of more than two persons at the crime scene.
In general, whether evidence should be excluded under New Jersey Rule of Evidence 403 because its prejudicial effect outweighs its probative value is an issue remitted to the discretion of the trial court. State v. Wilson, 135 N.J. 4, 20, 637 A.2d 1237 (1994). Only when the trial court commits a clear error of judgment does an appellate court disturb the trial court’s decision. State v. DiFrisco, 137 N.J. 434, 496, 645 A.2d 734 (1994) (DiFrisco II), cert. denied, — U.S. -, 116 S.Ct. 949, 133 L.Ed.2d 873 *185(1996); State v. Koedatich, 112 N.J. 225, 313, 548 A.2d 939 (1988) (Koedatich I), cert. denied, 488 U.S. 1017, 109 S.Ct. 813, 102 L.Ed.2d 803 (1989).
We find no error in the admission of the DNA evidence. The probative effect of the evidence is that the DNA test results showed that defendant could not be excluded as a contributor to the box-spring blood sample. Without specifying how the prejudicial effect of the evidence outweighs its probative value, the dissent suggests possible avenues of inquiry that defense counsel might have pursued. Suffice it to say that defense counsel made a different choice, one that involved the cross-examination of the State’s experts concerning the dot-intensity analysis of the blood. We perceive no reason to disturb the trial court’s ruling admitting the DNA evidence.
- V-
We next evaluate defendant’s argument that the trial court violated his constitutional right to present a defense by restricting examination of expert witnesses to potential flaws in the administration of the polymarker test.
- A-
After holding that the polymarker kit was admissible, the trial court precluded further challenges to admissibility at trial. The court’s ruling, however, did not prevent defendant from introducing evidence relevant to the weight or credibility of the expert testimony.
In accordance with that ruling, the trial court prohibited defendant from attempting before the jury (1) to cross-examine the State’s experts on Cellmark’s validation studies and (2) to elicit testimony from Dr. Shaler that the FBI did not use the polymarker test on mixed-blood samples. Defendant claims that the trial court committed reversible error when it barred him from cross-*186examining on these issues. To evaluate the defense’s argument, it is necessary to review the scope of the expert-witness testimony.
- B -
The State’s first DNA expert, Ms. Cooper, explained the results of the DQ Alpha and Polymarker tests conducted on crime scene samples and blood samples taken from defendant. Ms. Cooper also explained that before employing the PM test in its lab, Cellmark conducted a validation study.
The defense cross-examined Ms. Cooper on the dot-intensity imbalances in defendant’s PM test. Defense counsel sought to elicit that Cellmark had not conducted the PM test under the proper conditions and had failed to follow prescribed procedures. Ms. Cooper conceded that dot imbalances could result from improper hybridization or wash temperatures, improper salt-solution concentrations, or improperly-timed stringent washes. She disputed that discernible dot imbalances appeared on defendant’s HBGG and GC loci and explained that variations could occur at the GYPA locus. Because the positive control conducted with the test appeared normal, she did not investigate possible causes for the imbalances. Defense counsel also questioned Ms. Cooper on the possibility that by lifting the lid of the thermal cycler during the amplification process, she could have affected the accuracy of the tests.
Dr. Word testified that she reviewed all of Ms. Cooper’s work, and that she agreed with all of Ms. Cooper’s results. At one point during direct examination, the prosecutor inquired about a list of presentations on the PM test. Concerned that the prosecutor was trying to bolster the reliability of the PM test, the trial court precluded the inquiry.
On cross-examination, defendant pressed Dr. Word on the dot imbalances and on the possibility that Cellmark had not followed correct testing procedures. Dr. Word acknowledged that Cell-mark did not know how many people contributed to the mixed-blood sample recovered from the box spring.
*187On redirect, Dr. Word testified that Cellmark had performed validation studies with the PM test. When the prosecutor tried to inquire into the nature and conclusions of Cellmark’s validation study, defense counsel objected successfully.
According to the defense expert, Dr. Shaler, Cellmark’s PM test results were flawed and Cellmark’s method of interpreting the test was “scientifically indefensible.” Dr. Shaler emphasized the presence of dot imbalances to the jury and faulted Cellmark’s technicians for not doing the tests twice. He also raised the possibility that Cellmark’s test strips were defective.
Dr. Shaler further testified that he did not believe in dot-intensity interpretation, stating that he had a “philosophical difference” with the State’s experts. Accordingly, he disputed Cell-mark’s thesis that the defendant could be linked to the box-spring stain. Defense counsel requested a side-bar conference because he wanted to ask Dr. Shaler about Cellmark’s validation work. The court responded that it deemed inappropriate the inquiry about validation because the inquiry was directed at the admissibility, not the weight, of the evidence.
Later, on redirect, Dr. Shaler restated his view that associating alleles was an unscientific method of analyzing a mixed-blood sample. Defense counsel then presented Dr. Shaler with an article asserting that the FBI did not use the dot-intensity analysis technique. The State objected. At side-bar, defense counsel explained that he did not believe that the article went to admissibility, but rather to the weight of Dr. Shaler’s testimony that associating alleles was not a valid interpretive technique. The prosecutor responded that the article went to admissibility and that Dr. Shaler already had testified to his belief that other scientists did not believe in dot-intensity analysis. The trial court sustained the State’s objection.
- C -
The Sixth Amendment to the United States Constitution and Article 1, paragraph 10 of the New Jersey Constitution *188guarantee a criminal defendant’s right to confront witnesses. The right to cross-examine is an essential element of that right. State v. Budis, 125 N.J. 519, 530-31, 593 A.2d 784 (1991). A criminal defendant, therefore, may cross-examine an expert witness for the State on the facts, methodology, and rationale underlying the expert’s opinion. State v. Martini, 131 N.J. 176, 264, 619 A.2d 1208 (1993) (Martini I), cert. denied, — U.S. -, 116 S.Ct. 203, 133 L.Ed.2d 137 (1995); see also N.J.R.E. 705 (“The expert may ... be required to disclose the underlying facts or data on cross-examination.”). The scope of cross-examination, however, rests within the sound discretion of the trial court. Martini I, supra, 131 N.J. at 263, 619 A.2d 1208. Thus, a defendant’s constitutional right to confrontation does not guarantee unlimited cross-examination of a witness. Delaware v. Fensterer, 474 U.S. 15, 20, 106 S.Ct. 292, 294, 88 L.Ed.2d 15, 19 (1985).
Here, the trial court properly ruled that defendant could not relitigate admissibility issues that the court had determined. The court limited the parties to presenting evidence that went to the weight of the polymarker test results.
1. Validation Studies
Without merit is defendant’s argument that the trial court violated his confrontation rights by refusing to allow him to cross-examine Cellmark’s validation studies. Defense counsel never attempted to cross-examine the State’s experts on that matter. To the contrary, defense counsel objected when the prosecutor attempted to elicit testimony from Dr. Word on Cellmark’s validation trials. The dissent therefore misleads when it states that the trial court did not permit defendant to examine the scientific experts on Cellmark’s validation studies. Post at 276, 699 A.2d at 674.
The trial court properly precluded defendant from eliciting testimony through his own expert about Cellmark’s validation studies. The issue of Cellmark’s validation studies on the poly-marker more properly concerns the reliability of the polymarker. *189As such, the issue goes to the admissibility of the test and not its weight.
We are convinced the trial court did not compromise the defendant’s confrontation rights. As the record reveals, defense counsel cross-examined the State’s experts on their methodologies and on the results from the PM tests.
2. FBI Report
The trial court should have allowed defendant to elicit testimony from Dr. Shaler about an article indicating that the FBI did not use the association-of-alleles or dot-intensity technique. The error, however, was harmless. R. 2:10-2.
As previously explained, supra part IV.B.5., some of defendant’s criticisms regarding the dot-intensity technique bore on the weight, not the admissibility of the DNA evidence. Thus, the court should have permitted defendant to elicit testimony about the article, which went only to the weight of the State’s evidence that it could determine defendant’s identity from the mixed-blood sample taken from Schnaps’s box spring.
The error, however, was not clearly capable of producing an unjust result. R. 2:10-2. Dr. Shaler tried to prove that dot-intensity analysis was unscientific. He also testified that other forensic scientists agreed. Through his testimony, the jury knew that some scientists questioned the validity of dot-intensity analysis.
Knowledge of the FBI’s position would have added only limited weight to defendant’s argument. As previously discussed, supra part IV.B.5., the FBI article did not, as the dissent states, contradict dot-intensity analysis. Post at 256-258, 280, 699 A.2d 664-665, 676. Rather, the FBI Report urged caution when interpreting evidentiary samples that potentially may be from more than one donor. FBI Report, supra, at 52. The article also states, however, that mixtures of donors sharing the same alleles resulted in an increased dot intensity. Id. at 49.
*190- VI -
Defendant next argues that Cellmark improperly computed the one-in-1,400 number associated with the prosecution’s DNA evidence. Specifically, defendant alleges that Cellmark improperly calculated its product-rule computation; failed to inform the jury of the margin of error associated with such statistical calculations; and improperly assumed that the contributor to the mixed-blood sample was an African American. In addition to echoing the defendant’s criticisms, the dissent urges the necessity of an independent hearing for the admission of the statistical evidence. We find, however, that the trial court did not err in admitting statistic evidence tending to show that defendant could not be excluded as a contributor to the blood found at the crime scene.
- A-
Cellmark concluded that Harvey’s genotype for all of the six genes studied in both the PM and the DQ Alpha tests was common to only one-in-1,400 African Americans. In reaching that conclusion, Cellmark consulted databases of genetic frequencies for each genotype and multiplied those figures together in accordance with the product rule.
Cellmark used two population-frequency databases. One derived from Cellmark’s own study of fifty African Americans, some of whom were Cellmark employees, and from “paternity work” blood samples done at the lab. Cellmark also referred to a database compiled by Roche derived from a sample of 100 African Americans. Defendant does not challenge the adequacy of Cell-mark’s databases. Indeed, Dr. Shaler used the same databases in his calculations.
As mentioned above, defendant’s DNA tests revealed the following results:
LDLR: Genotype BB = 56% of African Americans
GYPA: Genotype AB = 50% of African Americans
HBGG: Genotype AC = 27% of African Americans
D7S8: Genotype AB = 45% of African Americans
*191GC: Genotype AB = 17% of African Americans
DQ Alpha: Genotype 4,4 = 11.9% of African Americans
According to the prosecution’s experts, someone with the foregoing genetic make-up could not be excluded as a donor to the mixed-blood sample recovered from the victim’s box spring.
To determine the percentage of African-American individuals who would share defendant’s genotype for all six genes, Cellmark multiplied genotype frequencies by one another in accordance with the product rule. See supra part IV.A.3.b. Application of the product rule indicated that defendant’s genotype for all six genes was common to approximately one-in-1,400 African Americans.
On appeal, defendant alleges that Cellmark’s one-in-1,400 figure was incorrect because it did not include the percentage of African Americans who could have matched the mixed-blood sample. Defendant argues that this alleged error arose because the DQ Alpha percentage used in Cellmark’s product-rule calculation was wrong. The dissent agrees. Post at 283-285, 699 A.2d at 678-679.
The DQ Alpha test conducted on the box-spring sample revealed the presence of the 2 and 4 alleles. It was not possible to associate alleles with the DQ Alpha test. The parties agree that the DQ Alpha test revealed all the possible genotypes that could have combined to reveal the presence of the 2 and 4 alleles: 2,2; 2,4; 4,4. Because defendant’s DQ Alpha genotype was 4,4, he could not be excluded as a potential donor if two people had contributed to the box-spring stain.
Defendant argues that the population frequencies for each of these three genotypes (2,2; 2,4; and 4,4) should have been added together before inserting a DQ Alpha statistic into the product-rule calculation. According to defendant, individuals who share defendant’s five polymarker genotypes and who have a DQ Alpha genotype of either 2,2; 2,4; or 4,4 — not just 4,4 — could not be excluded from the mixed-blood, box-spring sample.
*192Evidence for the entire population revealed the population frequencies for only two of these three possible genotypes: 9% of the population had the 2 and 4 alleles, and 17% had the 4 and 4 alleles. On that basis, defendant argues that at least 26% of the population (17% + 9%) had a DQ Alpha genotype that would correspond to the box-spring sample. Therefore, defendant alleges that Cellmark should have used a DQ Alpha value of 26% rather than 17% (adjusted to 11.9%, taking race into account) when it conducted the product-rule calculation. Defendant contends that substituting the 26% DQ Alpha value yields a result of .0015 or one-in-666. Thus, defendant posits that one-in-666 individuals could have contributed genetic material consistent -with that detected in the box-spring sample. In sum, defendant concludes that the State unfairly prejudiced him by stating that one-in-1400, instead of one-in-666 African Americans, could have provided the sample.
The prosecution did not present the one-in-1400 figure to prove the percentage of African Americans whose genetic compositions could be comparable with the PM/DQ Alpha profile of the box-spring sample. Nor did the prosecution, as the dissent states, “assert[] that the blood recovered from the scene revealed the ‘4,4’ DQ-Alpha genotype.” Post at 283, 699 A.2d at 678. Rather, the prosecution used the DNA evidence to demonstrate that defendant could not be excluded as a donor to the mixed-blood sample and that one-in-1400 African Americans shared his composite genotype.
On the State’s case, Ms. Cooper testified:
Prosecutor: Now, using the product rule, if you use the four, four, based upon your calculation and did as you indicated with a polymarker, what percentage do you come up with?
Cooper: For African-Americans having the polymarker types we have indicated and an individual having a four, four DQ Alpha type, the number is approximately 1 in 1400 in African-Americans.
Prosecutor: Let’s mark this down the bottom. You can excuse one thousand three hundred and ninety-nine African-Americans from having these characteristics, is that correct?
*193Cooper: Again, these numbers are just giving you an idea how rare or common, how these different types occur in a given population. And if you have a random population of approximately 1400 people, the chances are at least 1 of those people would have those particular polymarker and DQ Alpha type.
So described, the one-in-1400 evidence was relevant to show that defendant’s relatively rare composite genotype could not be excluded by the PM/DQ Alpha tests as a contributor to the box-spring sample. On cross-examination, moreover, Ms. Cooper testified that defendant’s specific DQ Alpha marker of 4,4, which occurred in seventeen percent of the population, constituted millions of people.
In addition, Ms. Cooper testified about statistics that did not include Harvey’s DQ Alpha results. She stated that multiplying the frequency calculations of each of Harvey’s PM loci showed that approximately one-in-170 African Americans shared Harvey’s genotype at the LDLR, GYPA, HBGG, D7S8, and GC loci.
Defendant now argues that the prosecution should have presented a ratio of the individuals whose polymarker genotypes would match defendant’s and whose DQ Alpha genotype was either 2,2; 2,4; or 4,4. That ratio, according to defendant, is approximately one-in-666. Defense counsel adduced no testimony in support of that ratio at trial. It is too late in the proceeding for defendant to insist that the State should have presented statistical evidence that defendant now believes would have been helpful at trial.
In addition, we note that the trial court, pursuant to defense counsel’s request, instructed the jury that statistical evidence does not relieve the jury of its obligation to determine the ultimate question of guilt. Specifically, the trial court stated:
Now, you heard testimony from DNA experts, from both the State and the Defense. Some of the testimony was expressed to you in terms of percentages. You must determine whether the State, again, has proven each and every element of the these [sic] charges beyond a reasonable doubt, and no testimony couched in terms of percentages or probabilities can relieve you of your obligation to make that determination.
At trial, the State derived its statistical evidence from genetics-population databases submitted to the trial court at the Rule 104 *194hearing. Dr. Word testified that the databases were compiled from Cellmark’s own study of fifty African Americans and a' database created by Roche Molecular Systems, which was derived from a Roche study of 100 African Americans. She further explained that Cellmark’s polymarker databases were “essentially the same” as those used by other prominent forensic laboratories around the country. Dr. Word also testified to the use of the product rule to derive a composite frequency for defendant’s six-loci genotype.
At the Rule 104 hearing, the defense’s expert, Dr. Shaler, did not contest the accuracy of Cellmark’s databases. Instead, he relied on them in his own statistical calculations. Although the defense still does not criticize the databases, it contends that the State’s population databases were not large enough to be an accurate sampling of the population. Post at 291, 699 A.2d at 682.
In addition, the defense never sought to introduce evidence disputing the general acceptance or the mathematical soundness of the product rule. Even on this appeal, the defense does not challenge the use of the product rule. Indeed, the defense states that “everyone agrees that the product rule is absolutely valid.” The dissent, however, maintains that the trial court erred in failing to hold an admissibility hearing on the reliability of the product rule. Post at 293, 699 A.2d at 682. In particular, the dissent argues that the State never established the independence of the loci tested, post at 290, 699 A.2d at 681, and that the use of the “unmodified” product rule was erroneous, post at 286, 290, 699 A.2d at 679, 681.
On this record, the dissent’s arguments are particularly unpersuasive. First, publications reveal that statistical databases used by other laboratories for DQ Alpha and PM loci are often constructed with small groups of blood donors. See, e.g., Kwang Man Woo & Bruce Budowle, “Korean Population Data on the PCR-Based Loci LDLR, GYPA, HBGG, D758, Gc, HLA-DQA1 and D1S80,” 40 J. Forensic Sci. 645 (1995) (obtaining blood from 116 *195unrelated Korean individuals); Jeanne M. Hayes et al. & Maya Freund, “Arab Population Data on the PCR-Based loci: HLADQA1, LDLR, GYPA, HBGG, D758, Gc and D1S80,” 40 J. Forensic Sci. 888 (1995) (collecting ninety-four blood samples from unrelated Arabs); see also Pope, supra, 220 Ill.Dec. at 312, 672 N.E.2d at 1324 (finding that “[b]ecause the PCR polymarker systems are two allele or three allele, the FBI may rely on a smaller database. Thus, the PCR Caucasian database contains only 145 individuals”).
Second, the trial court found that the underlying databases were reliable. In light of that finding, questions regarding the size of a database go to the weight of the evidence, not its admissibility. See People v. Adams, 195 Mich.App. 267, 489 N.W.2d 192, 198 (1992) (holding that, in face of expert testimony that database was valid, questions concerning size of database and statistical conclusions drawn from it went to weight, not admissibility, of evidence), modified in part on other grounds, 441 Mich. 916, 497 N.W.2d 182 (1993); State v. Copeland, 130 Wash.2d 244, 922 P.2d 1304, 1320-21 (1996) (holding that questions regarding size of FBI’s population databases was matter of weight once principle that frequency calculations could be made from an adequate database was determined to be generally accepted). Even on appeal, the defense does not challenge the size of Cellmark’s databases. See Fishback, supra, 851 P.2d at 893 (admissibility established because evidence uncontradieted).
The dissent’s contention that the trial court on its own motion should have conducted an independent Rule 104 hearing on the product rule also is without merit. The product rule assumes that each of the multiplied events (i.e., the matches at each allele) has occurred independently of one another. The State’s witnesses testified at the pretrial hearing and at trial that the population frequencies of the genes examined in the polymarker and DQ Alpha tests satisfied that requirement. Each derives from a gene located on a different part of the chromosome. Defendant never challenged the independence of the loci. At the pretrial hearing, *196Dr. Shaler used the same version of the product rule as the State to compute his statistical figures. Furthermore, he testified that those loci were “independently inherited.” See also FBI Report, supra, at 53 (“The distribution of the genotype frequencies for the various PM loci (as well as HLA-DQa) meet HWE [Hardy-Weinberg expectations], and there is little evidence for association of alleles across loci (for the PM loci, HLA-DQa, and D1S80) for our African American, Caucasian, southeastern Hispanic, and southwestern Hispanic population databases. The data demonstrate that valid estimates of a multiple locus profile frequency can be derived for .identity testing purposes using the product rule under the assumption of independence.”).
Although we have not previously considered whether the product rule has sufficient general acceptance so as to be admissible, two recent decisions by the Appellate Division accepted DNA statistical evidence calculated using the product rule. See Marcus, supra, 294 N.J.Super. at 287, 683 A.2d 221 (finding that “it is even clearer now than at the time of the Frye hearing that the use of the product rule in DNA analysis is generally accepted in the scientific community”); Dishon, supra, 297 N.J.Super. at 284-85, 687 A.2d 1074 (holding that use of product rule to calculate frequency of occurrence in Caucasian and Black populations is scientifically valid and that debate concerning ethnic substructures went to weight, rather than admissibility, of evidence). In addition, many other jurisdictions accept the product rule as scientifically reliable. See, e.g., Fishback, supra, 851 P.2d at 893-94 (holding that calculations done by applying product rule were generally accepted in relevant scientific community); Clark v. State, 679 So.2d 321, 321 (Fla.App.1996) (reasoning that “product rule calculations are appropriate as a matter of scientific fact and law”); Pope, supra, 220 Ill.Dec. at 315-16, 672 N.E.2d at 1327-28 (holding that use of product rule was generally accepted in scientific community); People v. Chandler, 211 Mich.App. 604, 536 N.W.2d 799, 803 (1995) (holding that “product rule method of DNA statistical evidence is now generally accepted in the relevant scientific community”); State v. Kinder, 942 S.W.2d 313, 317 *197(Mo.1996) (finding that product rule was generally accepted in scientific community); Copeland, supra, 922 P.2d at 1319 (concluding that “use of the product rule in establishing statistical probabilities of a genetic profile frequency in the human population is generally accepted within the relevant scientific community and that a significant dispute no longer exists in this matter”).
All of the foregoing decisions rely on the 1996 NRC Report and the FBI Report, both of which concluded that the product rule was appropriate for calculating frequency probabilities. See FBI Report, supra, at 50-51 (“These examples demonstrate that the use of the product rule would provide a valid estimate of a multiple loci frequency for forensic purposes.”); NRC Report, supra, at 5 (“In general, the calculation of a profile frequency should be made with the product rule.”).
The dissent, however, asserts that a remand is necessary because the product rule used by the State and the defendant was “unmodified.” Post at 287-291, 294, 699 A.2d at 680-681, 683. Specifically, the dissent notes that the NRC has recommended a “factor-of-ten” adjustment to the product rule, a modification of the product rule not employed here by either party. Post at 287, 699 A.2d at 680. That adjustment to the product rule can account for uncertainties in estimated frequencies. NRC Report, supra, at 33-34. Thus, for example, “[i]f the calculated probability of a random match between the suspect and evidence DNA is 1/(100 million), we can say with confidence that the correct value is very likely between 1/(10 million) and 1/(1 billion).” Id. at 34.
The dissent cites out-of-state decisions raising questions about population substructuring and the failure of the parties to use the “ceiling principle.” Post at 287-291, 699 A.2d at 680-681 (citing, e.g., State v. Johnson, 186 Ariz. 329, 922 P.2d 294 (1996) (finding product rule admissible only with use of ceiling principle); United States v. Porter, 618 A.2d 629, 643 (D.C.App.1992) (same); Commonwealth v. Lanigan, 413 Mass. 154, 596 N.E.2d 311, 314-16 *198(1992) (same)). In its 1996 report, however, the NRC dispelled many of those concerns.
In 1992, the NRC in its report, DNA Technology in Forensic Science 10-12 (1992), questioned the reliability of population sub-structuring:
The validity of the multiplication rule depends on the assumption of absence of population substructure. Population substructure violates the assumption of statistical independence of alleles. In a population that contains groups each with different allele frequencies, the presence of one allele in a person’s genotype can alter the statistical expectation of the other alleles in the genotype. For example, a person who has one allele that is common among Italians is more likely to be of Italian descent and is thus more likely to cany additional alleles that are common among Italians. The true genotype frequency is thus higher than would be predicted by applying the multiplication [product] rule using the average frequency in the entire population____The key question underlying the use of the multiplication rule — i.e., whether actual populations have significant substructure for the loci used for forensic typing — has provoked considerable debate among population geneticists. Some have expressed serious concern about the possibility of significant substructure.
To account for substructuring, the 1992 report recommended the use of the celling principle to calculate frequency statistics. The ceiling principle assumes the existence of some degree of population substructure and generates more conservative population frequency statistics than the product rule. Marcus, supra, 294 N.J.Super. at 282, 688 A.2d 221. In its 1996 report, however, the NRC “concluded that use of the ceiling principle overstates the effect of population substructure in calculating the population frequencies of a combination of matching DNA print patterns and consequently its use for forensic purposes is unnecessary.” Ibid.; NRC Report, supra, at 35.
We conclude that the scientific community generally accepts the product rule. Whether a party uses a “factor-of-ten” adjustment as the NRC recommends, or the “ceiling principle,” or no adjustment may be a legitimate subject of scientific debate. Indeed, the FBI, unlike the NRC, does not recommend a specific adjustment to the product rule. FBI Report, supra, at 50-52. General acceptance, however, does not require scientific consensus. We agree with the Appellate Division’s reasoning in Marcus:
*199The State [ ] may present evidence [to the jury] of population frequencies calculated by use of the product rule, the ceiling principle, or any other method that has a legitimate scientific basis. The defendant remains free to present conflicting expert opinion testimony regarding population frequency calculations. A point ultimately may be reached where there is such widespread agreement among experts in the field regarding the proper method of calculating population frequencies that only statistics generated by that methodology should be presented to the jury. However, until such a scientific consensus is established, this remains a legitimate subject for expert testimony at trial.
[Marcus, supra, 294 N.J.Super. at 288, 683 A.2d 221.]
Thus, the dissent’s concerns regarding the use of the “unmodified” product rule concern its weight, not its admissibility. See Shea, supra, 957 F.Supp. at 343 (finding that whether adjustments to product rule suggested in 1996 NRC Report were sufficiently conservative was issue of weight, not admissibility). Notably, the defense did not challenge the unmodified product rule, but rather used it in its own statistical analysis.
Moreover, application of the “factor-of-ten” to the polymarker results does not necessarily help defendant’s case. The dissent claims that, based on the State’s one-in-170 number “as many as one in seventeen people may share the DNA characteristics of the blood found at the scene.” Post at 289, 699 A.2d at 680. The “factor-of-ten,” however, is a confidence limit in both directions. Thus, using the one-in-170 number, as many as one-in-1700 hundred people may also share the DNA characteristics of the blood found at the scene.
As one court has stated:
There are doubtless many formulas and principles which experts use ... to arrive at their ultimate opinions. The determination of which factors, formulas, or calculations are necessary, either singly or in conjunction with each other, to form an expert opinion is within the knowledge and judgment of the expert, and, again is a subject which can be approached and examined in the cross-examination or by bringing forward other expert witnesses.
[Jenkins v. State, 627 N.E.2d 789, 794 (Ind.1993), cert. denied, 513 U.S. 812, 115 S.Ct. 64, 130 L.Ed.2d 21 (1994).]
Defendant had ample opportunity at trial to present the jury with a statistical interpretation of the DNA test results. In fact, Dr. Shaler explained his belief that, without associating alleles and taking into account all possible second donors to the box-spring *200sample, approximately one-in-100 people could have contributed a non-excludable composite genotype, a more favorable figure than the one-in-666 figure that defendant now propounds. Defendant had the opportunity to challenge the State’s statistical evidence, to present his own evidence, and to argue to the jury. We conclude that the trial judge did not err in permitting the jury to evaluate the State’s statistical evidence.
Our reading of the record reveals that the dissent attributes greater significance to the statistical evidence than the evidence received at trial. The summations did not belabor the statistical evidence. On summation, defense counsel argued:
The percentage that eventually came out as a result of this test, I believe is one out of 400 and there was another number, one out of 170.
For its part, the State invited the jury to deliberate based on Dr. Shaler’s one-in-100 calculation, if it found that calculation to be more credible:
Prosecutor: Remember the figures that we have, one in a hundred and seventy, one in fourteen hundred.
Evaluate them. If you want to use them, I suggest you should use the one in fourteen hundred, or if you choose, use the figure given by Dr. Shaler, one in a hundred. Exclude ninety-nine percent.
Further, the trial court instructed the jury that it was not bound by any expert opinion, that it could assign whatever weight it deemed appropriate to the expert evidence, and that testimony couched in terms of percentages or probabilities could not reheve the State of its burden of proof. The court left the weight of the statistical evidence to the jury. That defendant belatedly has thought of an alternative manner in which to present statistical evidence concerning the DQ Alpha results does not constitute reversible error.
- B -
Likewise without merit is defendant’s objection to the admission of the State’s statistical evidence through experts who were not statisticians. As previously indicated, defendant’s own *201expert, Dr. Shaler, did not take issue with Cellmark’s databases or its mathematical formula.
Moreover, the competency of a witness to testify as an expert is an issue remitted to the sound discretion of the trial court. See supra Part IV.B.1. Absent a clear abuse of that discretion, an appellate court will not interfere with the exercise of that discretion. Henningsen v. Bloomfield Motors, Inc., 32 N.J. 358, 411, 161 A.2d 69 (1960). We cannot say that the trial court’s decision to allow Dr. Word’s testimony at the pretrial hearing and Ms. Cooper’s testimony at trial regarding the frequency of genetic markers in the population was an abuse of discretion.
- C -
Also without merit is defendant’s objection to the admission of the State’s statistical evidence on the basis that it was introduced without telling the jury of the “confidence intervals” or “margins of error” associated with statistical analysis. In support of that argument, the defense attaches to its brief a letter from Dr. Eleanor Feingold, a biostatistieal professor at Emory University. Dr. Feingold’s letter asserts that the Cellmark data used to calculate defendant’s number for inclusion on the polymarker test has a range of between one-in-122 and one-in-297, at a 95% confidence level.
Pursuant to Rule 2:5-4(a), the State argues that the letter from Professor Feingold in defendant’s appendix concerning confidence intervals should be stricken from the record along with the portions of defendant’s brief relying on the letter. The State contends that such material is outside the record on appeal. The defense counters that Dr. Shaler alluded at trial to confidence intervals and that Dr. Feingold’s letter is simply an “illustration of the importance of confidence intervals to a fair understanding of statistical evidence.”
We grant the State’s motion to strike. An appellate court, when reviewing trial errors, generally confines itself to the *202record. See County of Bergen v. Borough of Paramus, 79 N.J. 302, 309-10 n. 2, 399 A.2d 616 (1979) (holding that appellant’s appendix, which contained numerous documents not offered into evidence, clearly violated R. 2:5-4); Ambassador Ins. Co. v. Montes, 76 N.J. 477, 481-82, 388 A.2d 603 (1978) (reasoning that Court was limited to review of insurance policy introduced into evidence and that insurance company could not supplement record on appeal). As previously explained, an appellate court may review scientific literature and judicial opinions, including those published after trial, to determine whether a technique is generally accepted. That practice, however, does not constitute an invitation for the parties to supplement the record with additional expert testimony. The place to introduce expert testimony is at trial, where the expert is subject to cross-examination, the opposing party can introduce contradictory expert testimony, and the trial court can assess the experts’ credibility.
Moreover, even if we were to accept Dr. Feingold’s letter, it would not alter our conclusions concerning the statistical evidence. As stated previously, the issue of the proper confidence level, if any, is one of weight, not admissibility. See supra Part VI.A. Defendant had ample opportunity to present such evidence at trial.
- D -
Defendant challenges the results of Cellmark’s DNA tests because the tests assume the source of the mixed-blood sample was an African American. Defendant states that Cellmark’s assumption was unwarranted and contrary to the presumption of innocence. He argues that his genotype for the D7S8 gene is less common in African Americans (45%) than in Caucasians (50%). From that premise, he concludes that analyzing the blood samples on the assumption that the suspect was black made the inclusion ratio appear more incriminating.
The allegation is without merit. The statistics were expressed in terms of the African-American population because defendant *203was African-American. Significantly, considerable other evidence indicated that the perpetrator of the Sehnaps murder was African-American. An African-American hair was recovered at the scene. Non-DNA blood typing revealed the presence of the CA II enzyme, which is present only in African Americans. We conclude that that evidence justifies the premise that the perpetrator was an African American. Legitimate reasons existed for expressing the percentage figures in terms of the African-American population. In sum, the DNA tests were not tainted by racism.
-VII -
We next consider defendant’s argument that the trial court erred in failing to grant a mistrial when a prosecution witness testified that another suspect had passed a polygraph examination.
- A-
Early in the investigation, police considered Peter Stohwasser, a neighbor of Sehnaps, as a possible suspect. Defendant contended at trial that the police failed to investigate thoroughly Stohwasser and prematurely dismissed him as a suspect.
To that end, defense counsel cross-examined Philip Beesley, a forensic scientist for the New Jersey State Police. Beesley acknowledged that during the time when Stohwasser was a suspect, Beesley did not conduct any of the comparison studies on Stohwasser’s blood that he later ran on defendant’s blood. Nor did Beesley receive any blood samples from Stohwasser.
Following Beesley’s testimony, Investigator O’Brien explained why the police had dismissed Stohwasser as a suspect. Pursuant to a search warrant, police searched Stohwasser’s apartment and seized various items for bloodstain analysis. Investigating police eventually eliminated Stohwasser as a suspect, however, for numerous reasons. First, the blood stains on the seized items did *204not match the blood type of the victim. Second, Stohwasser did not possess any footwear that matched the sneaker print left at the crime scene. Finally, Stohwasser is Caucasian and could not have been the source of the “Negroid hair” found at the crime scene.
On redirect, the State elicited from O’Brien that Stohwasser had taken and passed a polygraph test. At that point, defense counsel objected. The trial court sustained the objection and informed the jury to “disregard the last question and the last answer.”
The following day, the defense moved for a mistrial on the basis of the polygraph reference. The defense unsuccessfully argued that mention of the polygraph raised an inference that the police had not eliminated defendant as a suspect because either he refused to take a lie detector test or he took the test and failed.
At the charge conference the polygraph issue re-emerged. The trial court reprimanded the State and proposed a corrective jury instruction:
Now, I prepared a proposed instruction on my own, thinking about this issue, which I will eventually tell the jury that a polygraph examination is not admissible in evidence. While it may be a legitimate investigatory tool, and I would have told the jury or [sic] my proposed instruction that I thought of was simply to suggest that the question was not designed to bring out the truth of the result, but simply to show why the police eliminated him as a suspect.
I would also — my proposed instruction will tell the jury that they are to draw no adverse inference against Mr. Harvey, because when the Stonehauser [sic] request was made, Mr. Harvey was not in custody. That was my proposal.
You want me not to raise it because of your concern of possible impact, I certainly won’t do it.
All I am saying is that in my view of my notes on this issue and my feeling of the case, so to speak, I am satisfied tht [sic] the question and answer are simply something not capable of producing any prejudice to the defendant whatsoever.
Defendant declined the instruction.
- B -
Defendant now argues that the reference to the polygraph results of an earlier suspect should have resulted in a mistrial *205because it violated his Fifth Amendment right to silence and prejudiced him in the eyes of the jury.
The decision to grant or deny a mistrial is entrusted to the sound discretion of the trial court, State v. DiRienzo, 58 N.J. 360, 383, 251 A.2d 99 (1969), which should grant a mistrial only to prevent an obvious failure of justice. State v. Rechtschaffer, 70 N.J. 395, 406, 360 A.2d 362 (1976). An appellate court should defer to the decision of the trial court, which is in the best position to gauge the effect of the allegedly prejudicial evidence. State v. Winter, 96 N.J. 640, 647, 477 A.2d 323 (1984). Thus, an appellate court will not disturb a trial court’s ruling on a motion for a mistrial, absent an abuse of discretion that results in a manifest injustice. DiRienzo, supra, 53 N.J. at 383, 251 A.2d 99. Here, the reference to the polygraph test did not refer directly either to defendant or a testifying witness, but to the polygraph results of an unindicted suspect.
In a similar context, the Ninth Circuit has found the reference to the polygraph test of another suspect to constitute harmless error. See United States v. Candoli, 870 F.2d 496, 505 (9th Cir.1989) (finding that district court’s refusal to strike references to polygraph examination was not prejudicial because it did not materially affect verdict); United States v. Hall, 805 F.2d 1410, 1417 (10th Cir.1986) (testimony that defendant had failed two polygraphs submitted to explain police’s failure to conduct more complete investigation). But see State v. Moss, 180 W.Va. 363, 376 S.E.2d 569 (1988) (holding that admission of testimony that State had dismissed indictment against earlier suspect because he successfully passed polygraph was reversible error when prosecutor repeatedly referred to polygraph on direct and in summation).
On this record, the reference to an unindicted suspect’s polygraph results does not constitute reversible error. The reference came in response to defendant’s attack on the conduct of the police’s murder investigation. Further, the trial court immediately sustained the defendant’s objection and instructed the jury to disregard the reference to the polygraph results. The State, *206moreover, presented substantial evidence explaining why it eliminated Stohwasser as a suspect. Any prejudice to defendant was minimal.
-VIII -
Defendant argues that the trial court committed reversible error by inadequately protecting the jury from prejudicial pre- and midtrial publicity. Specifically, defendant alleges that the trial court erred in refusing: (1) to dismiss the entire pool of qualified jurors as being tainted by pretrial publicity; (2) to question individually jurors regarding their exposure to press coverage of the trial; (3) to question jurors on whether they had been exposed to publicity of the contemporaneous murder trial of State v. Johnson; and (4) to sequester the jury during the penalty-phase deliberations.
- A-
The first phase of jury selection lasted from February 2 to March 31, 1994. Each potential juror filled out an eleven-page questionnaire. On the specific issue of pretrial publicity, the questionnaire contained five relevant questions:
37. Have you seen anything on television or heard anything on the radio about this case? _Yes _No.
38. Have you read anything about this case in the newspaper? _Yes _No.
39. Other than what the judge said to you today, has anyone else talked to you about this case? _Yes _No.
40. Have you heard anybody discussing this case or the people involved in it, either here today or at anytime previously? _Yes _No.
41. Do you know anything about this case other than what you have heard in court today? —Yes _No.
*207The trial court questioned each juror individually concerning their answers. The court also afforded both the defense and the prosecution opportunities to question the jurors. Not one juror was excused because of exposure to prejudicial press coverage. Defendant did not object to the court’s handling of the publicity issue. At the close of the February-March 1994 selection phase, and without objection from defendant, the trial court qualified forty-seven jurors.
An adjournment to enable defendant to examine the prosecution’s DNA evidence delayed resumption of the jury-selection process until October 18,1994. Using the same process, the court qualified six more jurors. Again, no juror was excused on the basis of exposure to publicity.
On October 24, 1994, the jurors who had been selected during the February-March phase returned to court. The trial court asked whether, during the interim, they had heard anything about either the parties or the proposed evidence that would impair their ability to serve impartially. Only one prospective juror indicated that he had read about the ease. He told the court that he had learned that this case was a retrial and that defendant previously had been sentenced to death. The trial court excused that juror and told the other jurors to return October 28, at which time the court resumed voir dire with the remaining jury pool.
On October 25, 1994, defense counsel notified the court of the existence of three recently published newspaper articles that discussed defendant’s case. The articles mentioned defendant’s suppressed confession and the previously-imposed death sentence. At defendant’s request, the court inquired of potential jurors about their knowledge of the articles. From October 25 through October 27, in response to specific questions from the court, no juror revealed that he or she had read the subject articles. Concerned about the jurors selected in February-March and on October 18 and 20, who had not been present in the courtroom since October 24, defense counsel requested on October 28 that *208the court similarly question those jurors individually. The court denied the request. Instead, the court questioned the previously-selected jurors as a group and individually questioned only those jurors who indicated that they were familiar with the newspaper articles. In addition, the court gave defense counsel the option of requesting additional peremptory challenges.
When the forty-seven previously qualified jurors were brought into the courtroom, the court asked whether any of them had “read, heard, overheard, or been told anything about this matter whatsoever?” Twelve jurors raised their hands. Of these twelve jurors, two had overheard discussions of the case, two had heard radio broadcasts, and seven admitted having read about the case in the newspaper. One of the jurors who raised her hand had not heard about the case but wanted to inform the court that she knew a sheriffs officer. Ultimately, all twelve of those jurors were excused.
In light of the number of jurors who had indicated that they had been exposed to pretrial publicity, defendant moved either to dismiss the entire pool of the remaining, previously-qualified jurors or for individual questioning of the jurors. Defendant argued that, despite the jurors’ denials, a substantial likelihood existed that the remaining jurors had been exposed to the circulating articles and broadcasts.
The trial court denied both motions, explaining that the questioning had effectively identified tainted potential jurors. The court reminded the remaining jurors of the importance of avoiding news coverage and of their ongoing obligation to report exposure to any such coverage. It then instructed the qualified jurors to return on November 29, 1994, for the commencement of the trial.
To replace the jurors excused on October 28, the court continued jury selection on November 2, 3, 4, 7, 9, 10, and 15, 1994. During those sessions, the court individually questioned each potential juror about his or her familiarity with defendant’s case. None of the potential jurors revealed any exposure to pretrial *209publicity. By the close of jury selection on November 15, 1994, fifty jurors had been qualified.
On the morning of the first day of the trial, November 29,1994, defendant moved to sequester the jury or to strike the entire panel and to start the selection process anew. Defendant informed the court that the morning editions of the Star Ledger, News Tribune, and Home News carried articles on the court’s DNA rulings. Two of the articles mentioned defendant’s suppressed confession. The court denied the defendant’s motions for sequestration and to strike the jury, but agreed to ask all of the qualified jurors whether any of them had read anything about the ease.
The court’s inquiry to the panel of qualified jurors revealed that two had been exposed to press coverage. The court excused both jurors. Defendant renewed his motion to strike the panel and sequester the new jury. Again, the court denied the motions.
The jury was then selected from the pool of qualified venire persons. Defendant used only eighteen of his twenty peremptory challenges, thus mooting the court’s earlier offer to consider defense requests for additional challenges. At the end of the day on November 29, the court instructed the jury to avoid exposure to any press coverage.
On December 1, 1994, the third day of trial, defendant brought to the attention of the court that three newspapers had published new articles about the case. Five days later, a juror informed the court that he had heard a coworker discussing the case and that he had discovered that two of his eoworkers were related to defendant. The court excused the juror. Thereafter, the court denied defendant’s motion for sequestration.
Following defendant’s conviction on December 13, 1994, the court began the penalty phase on December 15,1994. On December 14, defendant again moved for sequestration because of continuing publicity. The court denied that motion and defendant’s *210request that the court again question individually the jurors on their exposure to the press coverage.
At the beginning of the second day of the penalty phase, defendant complained about three more damaging articles in the morning newspapers. The court declined a defense request to interview jurors individually about the articles. In response to questions posed to the entire panel, no juror indicated that he or she had been exposed to press coverage.
- B -
The Sixth and Fourteenth Amendments to the United States Constitution and Article I, paragraph 10 of the New Jersey Constitution guarantee the right to a fair trial by an impartial jury. Irvin v. Dowd, 366 U.S. 717, 722, 81 S.Ct. 1639, 1642, 6 L.Ed.2d 751, 755 (1961); Koedatich I, supra, 112 N.J. at 267, 548 A.2d 939; State v. Williams, 93 N.J. 39, 61, 459 A.2d 641 (1983) (Williams I). The guarantee protects the defendant from substantial pre- and midtrial publicity. Sheppard v. Maxwell, 384 U.S. 333, 362-63, 86 S.Ct. 1507, 1522-23, 16 L.Ed.2d 600, 620 (1966); Williams I, supra, 93 N.J. at 60, 459 A.2d 641. In death-penalty cases, the trial court has a heightened duty “to preserve the integrity of the jury and minimize the danger that prejudice will infiltrate the adjudication process.” Williams I, supra, 93 N.J. at 63, 459 A.2d 641.
The means for ensuring jury impartiality in the face of pretrial publicity is interrogation of the jury. Patton v. Yount, 467 U.S. 1025, 1038-39, 104 S.Ct. 2885, 2892-93, 81 L.Ed.2d 847, 858 (1984); State v. Jackson, 43 N.J. 148, 203 A.2d 1 (1964), cert. denied sub nom. Ravenell v. New Jersey, 379 U.S. 982, 85 S.Ct. 690, 13 L.Ed.2d 572 (1965); State v. Gary, 229 N.J.Super. 102, 111, 550 A.2d 1259 (App.Div.1988). In performing that function, trial courts possess considerable discretion. To protect juries from the taint of pretrial publicity, particularly in capital cases, courts should “exercise extraordinary care in the voir dire of potential jurors and [can] excuse for cause any juror who has been exposed *211to sensational prejudicial publicity, especially where such exposure is repeated and involves patently inadmissible evidence.” Williams I, supra, 93 N.J. at 68-69, 459 A.2d 641. Even capital defendants, however, are not entitled to jurors who are totally ignorant of the facts and issues of their cases. Koedatich I, supra, 112 N.J. at 268, 548 A.2d 939; State v. Sugar, 84 N.J. 1, 23, 417 A.2d 474 (1980); Gary, supra, 229 N.J.Super. at 110, 550 A.2d 1259.
The appellate standard for reviewing a voir dire procedure is whether, despite the trial court’s efforts, there still existed a “realistic likelihood of prejudice resulting from pretrial publicity.” Williams I, supra, 93 N.J. at 63, 459 A.2d 641.
Preliminarily, an appellate court must distinguish “between cases in which the trial atmosphere is so corrupted by publicity that prejudice may be presumed, and cases in which pretrial publicity, while extensive, is less intrusive, making the determinative issue the actual effect of the publicity on the impartiality of the jury panel.” State v. Biegenwald, 106 N.J. 13, 33, 524 A.2d 130 (1987) (Biegenwald II). Our examination of the record does not support a presumption of prejudice to defendant.
When a court cannot assume prejudice, the inquiry to determine the existence of a realistic likelihood of prejudice is whether under the totality of the circumstances the voir dire resulted in a fair and impartial jury. Biegenwald IV, supra, 126 N.J. at 22-23, 594 A.2d 172. In making that determination, an appellate court should show appropriate deference to the trial court’s assessment of “matters of credibility, judgment and discretion which should not ordinarily be disturbed on appeal.” Gary, supra, 229 N.J.Super. at 111, 550 A.2d 1259; see also State v. Singletary, 80 N.J. 55, 63-64, 402 A.2d 203 (1979); Jackson, supra, 43 N.J. at 160, 203 A.2d 1.
We conclude that the trial court did not err by failing to strike the qualified jury pool. The court examined all potential jurors through questionnaires and individual questioning. It spe*212cifleally questioned jurors on learning of the October 25 newspaper articles and questioned collectively the previously-qualified jurors. The court excused jurors who revealed prejudicial exposure to press reports. Furthermore, the court informed the jurors of their ongoing duty to avoid contact with newspaper articles and broadcasts. We conclude that a substantial likelihood did not exist that media coverage had prejudiced the remaining jurors.
- C -
We also reject defendant’s argument that the trial court committed reversible error by not interrogating the previously-qualified jurors after publication of each new series of newspaper articles. After considering the totality of the circumstances, the trial court effectively informed the jurors of their duty to remain impartial and removed tainted jurors from the panel. The collective questioning elicited admissions of taint from twelve jurors. On balance, we conclude that the trial court adequately responded to problems posed by press coverage of the trial.
- D -
Defendant also alleges that the trial court erred by failing specifically to inquire of jurors about another Middlesex County murder prosecution, State v. Johnson, which was also the subject of press coverage during defendant’s trial. The Johnson case bore superficial similarities to defendant’s case in that it also involved an African-American male accused of murdering a white woman. Local newspapers widely reported Johnson’s statement that “blacks should kill white women so they can’t reproduce.”
In response to defendant’s concerns about the Johnson case, the trial court asked the penalty-phase jury collectively whether it had read reports about defendant’s case or “anything else” that would affect their impartiality. Defendant complains that the trial court should have asked the jurors specifically about the Johnson case. We disagree.
*213Defendant did not express concern over the Johnson case until December 15, 1994, the eve of the commencement of the penalty phase. Furthermore, defendant did not request questioning of the jurors about the Johnson ease before the guilt phase. Before starting the penalty-phase deliberations, the court asked the jurors whether they had learned about defendant or heard anything else that could affect their impartiality in deciding defendant’s sentence. Although the court did not specifically mention Johnson, the court broadened the inquiry to ascertain whether the jurors had read any prejudicial reports from any source outside the record.
The present case differs from State v. Jasuilewicz, 205 N.J.Super. 558, 501 A.2d 583 (App.Div.1985), certif. denied, 103 N.J. 467, 511 A.2d 649 (1986), where the defendant killed his mother by stabbing her twenty-one times. At trial, Jasuilewicz had relied on the defense of insanity. Jury selection began in the wake of news accounts of the acquittal by reason of insanity of John Hinckley, who had shot President Ronald Reagan. Concerned about taint from the Hinckley verdict, Jasuilewicz repeatedly requested a voir dire concerning the effect on each juror of the Hinckley publicity. Id. at 567, 501 A.2d 583. The trial court, however, refused defendant’s request. The Appellate Division found that the trial court’s refusal to interrogate about the Hinckley publicity constituted reversible error. Id. at 569, 501 A.2d 583.
In Jasuilewicz, the trial court failed not only to ask searching questions, but failed generally to ask any questions about the jurors’ potential exposure to the Hinckley case. Ibid. In contrast, the trial court here, although it did not specifically mention the Johnson case, adequately interrogated the jury about its exposure to another murder prosecution.
Unlike Jasuilewicz, who was concerned about the reaction of the jury to the insanity defense in Hinckley, defendant here did not voice a concern over the Johnson ease until the day before the beginning of the penalty phase. Thus, the court did not question jurors about the Johnson case before the guilt phase. Further*214more, the trial court in the present case, unlike the court in Jasuilewicz, responded to defendant’s concerns.
- E -
Defendant alleges error in the trial court’s failure to sequester the jury, especially during the penalty phase. Defendant also asks this Court to rule that jury sequestration should be mandatory in all capital cases.
Jury sequestration is a decision generally left to the discretion of the trial court. State v. Moriarty, 133 N.J.Super. 563, 569, 338 A.2d 14 (App.Div.1975), certif. denied, 68 N.J. 172, 343 A.2d 459 (1975). The practice places a great burden on the judicial system and on individual jurors. Accordingly, courts use it only in extraordinary circumstances. R. 1:8-6(a). We are unconvinced that the publicity surrounding defendant’s trial was so pervasive and prejudicial that the trial court abused its discretion by denying defendant’s motion.
- IX -
Defendant next alleges various incidents of prosecutorial misconduct. Specifically, defendant argues that the prosecutor: exceeded the bounds of proper summation; improperly questioned the State’s expert on hair analysis by asking questions the answers to which were inadmissible and beyond the scope of examination as limited by the trial court; and indirectly introduced inflammatory evidence.
- A-
Defendant contends that the prosecutor committed misconduct in summation by urging the jury to apply the product rule to the entirety of the State’s case.
In her summation, defense counsel stressed that reasonable doubt still existed about the identity of Irene Schnaps’s murderer. Counsel claimed that the initial investigation was marred by *215shoddy police work and that all of the State’s evidence, including the DNA test results, could not narrow the pool of potential suspects below the thousands.
The prosecution countered in its summation that the evidence established beyond a reasonable doubt that defendant was guilty of the murder of Irene Schnaps. The prosecutor argued to the jury:
Use the product rule ... and figure out in your own mind the likelihood that one person out of one hundred — it could be thousands — would have size six and a half shoe [sic], would have sneakers with the thread [sic] that size and wear [sic] match, blood stained pillow at the scene, multiply it by the fact that the Negroid hair matched the Negroid hair found at the scene, match the exemplar from Mr. Harvey, and multiply that by the watch.
You have taken a circumstantial — you started with everybody in the world and you narrowed that circumstantial by what is available. These are the items that were available at the scene. This is what the State used. I suggest to you, they are more than sufficient to show you beyond a reasonable doubt Mr. Harvey committed this offense.
********
This is not speculation, this is not guesswork, this is fact. The probabilities of a person having those things, all these things, are minuscule, well beyond a reasonable doubt, and the person that had these characteristics, whether it be hair, blood, shoe size, shoes and watch is minuscule. That person is Mr. Harvey____
Following the State’s summation, defense counsel objected to the prosecutor’s comment that the probabilities were “minuscule” that all the pieces of evidence, when taken together, allowed room for a reasonable doubt. Counsel argued that the prosecution was “trying to assess a mathematical standard” in determining the existence of reasonable doubt. Further, counsel requested the court to instruct the jury that “reasonable doubt cannot come down to [a] mathematical formula or standard.” The court denied the request.
Defendant now challenges the prosecutor’s invitation for the jury to apply the product rule to the totality of the State’s evidence. Furthermore, defendant alleges that inviting the jury to use the product rule was akin to the performance of an improper demonstration by the prosecutor before the jury.
*216We evaluate the allegations of prosecutorial misconduct in light of the unique responsibilities of a prosecutor. Prosecutors must pursue their duties “with earnestness and vigor,” United States v. Young, 470 U.S. 1, 7, 105 S.Ct. 1038, 1042, 84 L.Ed.2d 1, 7 (1985) (quoting Berger v. United States, 295 U.S. 78, 88, 55 S.Ct. 629, 633, 79 L.Ed. 1314, 1321 (1935)), and make a “forceful presentation of the State’s case.” State v. Ramseur, 106 N.J. 123, 320, 524 A.2d 188 (1987). As advocates, prosecuting attorneys enjoy considerable leeway when making a summation. State v. Perry, 65 N.J. 45, 48, 319 A.2d 474 (1974); Dixon, supra, 125 N.J. at 259, 593 A.2d 266; State v. Michaels, 264 N.J.Super. 579, 641, 625 A.2d 489 (App.Div.1993), aff'd, 136 N.J. 299, 642 A.2d 1372 (1994).
Generally, prosecutorial misconduct does not constitute a ground for reversal unless the conduct is deemed “so egregious that it deprived defendant of a fair trial.” Ramseur, supra, 106 N.J. at 322, 524 A.2d 188. A reviewing court must consider “the tenor of the trial and the degree of responsiveness of both counsel and the court to improprieties when they occurred.” State v. Marshall, 123 N.J. 1, 153, 586 A.2d 85 (1991) (Marshall I). The United States Supreme Court has articulated a similar test: “whether the prosecutor’s comments so infected the trial with unfairness as to make the resulting conviction a denial of due process.” Darden v. Wainwright, 477 U.S. 168, 181, 106 S.Ct. 2464, 2471, 91 L. Ed.2d 144, 157 (1986) (internal quotations omitted).
Here, the prosecutor’s reference to the product rule did not constitute reversible error. The comment was more a rhetorical device than an invitation for the jury to apply a mathematical formula. The prosecutor was responding to defendant’s allegations that the evidence acted to enlarge, rather than limit, the pool of possible suspects. By commenting on the product rule, the State tried to argue that all of the evidence, when considered collectively, pointed to defendant’s guilt.
*217Furthermore, the trial court’s instructions cured any perceived prejudice. On three separate occasions, the court instructed the jury that the comments of counsel were not evidence. Further, the court gave a specific instruction that no testimony couched in mathematical terms, such as percentages or probabilities, could relieve the State of its burden of proof. These instructions ensured that no confusion or prejudice resulted from the prosecutor’s product-rule comment. See Spann, supra, 130 N.J. at 518, 617 A.2d 247 (1993) (stating that appropriate jury instructions can enable juries to cope with complexities of probabilistic evidence).
- B -
Defendant next attacks the prosecutor’s examination of the State’s hair-analysis expert, Theodore Mozer. Specifically, defendant contends that portions of Mozer’s testimony were both inadmissible and beyond the permissible scope of the examination.
Defense counsel objected when Mozer testified that if the characteristics of microscopically-compared hairs match, “we can determine that this hair ... probably came from the person to which we’re comparing the sample to.” Defense counsel contended that the testimony was not made with a reasonable degree of scientific certainty. The State then elicited testimony from Mozer that to a reasonable degree of scientific certainty defendant’s hair compared “microscopically and physically” to the “Negroid” hair recovered from the crime scene.
Defendant argues that certain questions to which his counsel successfully objected invited answers that exceeded the scope of the extent of the witness’s testimony, as represented by the prosecutor. Consequently, defendant maintains that the prosecution improperly created the impression that the expert had never made a mistake and that the hair came from defendant. We disagree. The trial court promptly sustained defense objections to the questions. Our reading of the record leads us to *218conclude that the prosecutor’s questions do not warrant reversal of defendant’s conviction.
- C -
Defendant’s third allegation of prosecutorial misconduct is that the State improperly injected into the case the specter of sexual assault. Specifically, defendant complains about the introduction of a schematic diagram and testimony regarding items sent to the State Police Laboratory.
Before trial, the State prepared a schematic diagram of Schnaps’s apartment to explain where it found certain items of evidence. Of the approximately sixty individual items of evidence retrieved from the apartment, eight were featured in the diagram. At the bottom of the diagram was a legend describing the eight items, indicating their location in the apartment.
Defendant challenges the inclusion in the diagram of the location of the victim’s “panties” and a pair of her shorts. Defense counsel objected to those items being highlighted in the diagram, arguing that they had no evidential value and that reference to them was potentially prejudicial.
The trial court overruled the objection, noting that the shorts, which were made of terry cloth, were designed for outerwear. Consequently, reference to them was devoid of potential prejudice. The court decided that the panties were relevant to the State’s allegation that Schnaps’s body had been washed to remove traces of blood. Defense counsel accepted the court’s offer of a limiting instruction. Subsequently, a State investigator testified to the long list of evidence removed from Schnaps’s apartment, including the panties. The trial court then instructed the jury:
Ladies and gentlemen, before [the prosecutor] goes further, you heard the investigator indicate to you what he retained from the apartment and I know he referred to panties and perhaps other undergarments. I have allowed -that testimony but please understand that there is no charge against Mr. Harvey of any type of sexual assault. Please understand that.
*219Thereafter, the investigator described the items sent to the State Police Laboratory on June 20,1985. Two of the items were “a pubic hair control taken from the victim,” and “oral, vaginal, and anal swabs taken from the victim at the time of autopsy.” Defense counsel did not object to the description. Hence, we treat the reference to them as a matter of plain error.
The investigator then began to read from the list of items that were later sent to the police laboratory including “pubic hair sample from the sus... ” At a sidebar conference, defense counsel objected to mention of defendant’s pubic hairs. The court agreed and instructed the prosecutor that the investigator could not refer to the submission of defendant’s pubic hair.
Defendant now maintains that introduction of the schematic diagram featuring the location of the victim’s panties, the mention that vaginal and anal swabs were sent to the police laboratory, and the reference to pubic hair, combined to infect the proceedings with prejudicial inferences of sexual assault. The contention is without merit.
Through special instructions, the trial court ensured that the jury understood that the case did not include any allegations of sexual assault. The court did not abuse its discretion in admitting the schematic diagram. That evidence was relevant to the prosecution’s theory that defendant had wiped blood from the victim’s body with a damp towel. Nor do we find that the State’s investigator raised the specter of sexual assault. The court properly prevented the investigator from testifying that a sample of defendant’s pubic hair was sent to the police laboratory. Likewise, we do not find plain error in the reference to the vaginal and anal swabs sent for analysis to the police laboratory. See State v. Harper, 128 N.J.Super. 270, 277, 319 A.2d 771 (App.Div.) (“Trial errors which were induced, encouraged, or acquiesced in or consented to by defense counsel ordinarily are not a basis for reversal on appeal”), certif. denied, 65 N.J. 574, 325 A.2d 708 (1974). The trial court’s instruction that defendant was not accused of sexual assault was sufficient to have dispelled any possible taint which *220could have resulted from that reference. Moreover, unlike the dissent, we do not view these independent trial events to be violative of our directive in Harvey I, supra, 121 N.J. at 407, 581 A.2d 483, that references at retrial should not include the information that the hair found at the crime scene was a pubic hair. Post at 313, 699 A.2d at 693. The State never characterized the hair found in Schnaps’s apartment as a pubic hair.
- X -
Defendant argues that the police illegally seized the Seiko LaSalle watch found in the trunk of his car. He asserts further that the police violated his right to silence four times on October 28, 1985, by “unrelenting[ly] coercive police methodology.” Those violations of his right to silence, he alleges, coerced him into admitting that he committed various burglaries and a sexual assault in West Windsor Township. He argues that the “coerced” admissions “forced” him to consent to a police search of his car and Jamesburg apartment on October 29. Defendant accordingly argues that the trial court should have suppressed the watch.
- A-
At the pretrial suppression hearing, the trial court concluded that defendant had consented to the search and that the watch could be admitted into evidence at trial. The court acknowledged that October 28, the date of defendant’s arrest, was an “exhausting” day for defendant. Nevertheless, the court concluded that the police did not employ any coercive tactics and that defendant voluntarily sought out Detective Swanhart to confess to the West Windsor crimes. Likewise, the court held that defendant voluntarily signed the consent-to-search form. The voluntariness of defendant’s consent, the court noted, was confirmed by the fact that he did not believe that any incriminating evidence was in his car.
*221- B -
In Harvey I, supra, 121 N.J. at 417, 581 A.2d 483, defendant sought to suppress only his October 30 confession to the Schnaps murder. Accordingly, we declined to address defendant’s allegations that his right to silence was violated on October 28. Ibid. We now conclude that defendant’s concerns are without merit.
In general, the police must “scrupulously honor” a suspect’s right to silence. Michigan v. Mosley, 423 U.S. 96, 104, 96 S.Ct. 321, 326, 46 L.Ed.2d 313, 321 (1975); Miranda v. Arizona, 384 U.S. 436, 473-74, 86 S.Ct. 1602, 1627-28, 16 L.Ed.2d 694, 723 (1966); State v. Johnson, 120 N.J. 263, 282, 576 A.2d 834 (1990); Hartley, supra, 103 N.J. at 260-61, 511 A.2d 80. “Scrupulously honoring” a defendant’s right to silence requires a cessation of questioning once the defendant asserts his Fifth Amendment right. Johnson, supra, 120 N.J. at 282, 576 A.2d 834 (“Where the invocation of the right to remain silent is followed by no interruption in questioning, and where the interrogation continues as if nothing had happened, the right is not scrupulously honored.”); State v. Bey, 112 N.J. 45, 68-70, 548 A.2d 846 (1988) (Bey I) (holding defendant’s rights not scrupulously honored when police officers ignored his attempt to remain silent and continued questioning); Hartley, supra, 103 N.J. at 287, 511 A.2d 80 (“[a]uthorities must cease interrogation of suspect on his request”....). If police are unsure whether a defendant is asserting his right to silence, they must either stop the interrogation completely or “ask only questions narrowly directed to determining whether defendant was willing to continue.” Johnson, supra, 120 N.J. at 284, 576 A.2d 834; see also State v. Wright, 97 N.J. 113, 120 n. 4, 477 A.2d 1265 (1984).
Once a defendant invokes his or her right to silence, interrogation can resume only if the police administer a fresh set of Miranda warnings. Hartley, supra, 103 N.J. at 267, 511 A.2d 80; accord State v. Adams, 127 N.J. 438, 445, 605 A.2d 1097 (1992); State v. Fuller, 118 N.J. 75, 83, 570 A.2d 429 (1990); State *222v. Mujahid, 252 N.J.Super. 100, 109, 599 A.2d 536 (App.Div.1991), certif. denied, 127 N.J. 561, 606 A.2d 372 (1992). That rule, however, does not apply if the defendant initiates a dialogue about the crime. Fuller, supra, 118 N.J. at 85, 570 A.2d 429.
- C -
Police interrogated defendant three different times on October 28 before he confessed to the West Windsor crimes. When the police brought defendant to the station at 8:30 a.m., they informed him of his Miranda rights, and he signed a waiver. Defendant ate lunch at around 3:00 p.m. The first interrogation began at 3:37 p.m.
At this first interrogation session, the police again read defendant his rights, and he signed another rights form. The interrogation lasted a little over an hour, at which point defendant started to cry and asked for a half-hour “to think.” The interrogation ceased, and the police returned defendant to his cell.
By requesting “time to think,” defendant was not invoking his right to silence. See State v. Bey, 112 N.J. 123, 138-40, 548 A.2d 887 (1988) (Bey II) (holding defendant’s request to “lay down and think about what happened” did not constitute invocation of right to silence as “[n]ot every break in questioning compels renewed administration of the Miranda warnings”). Although the police granted defendant the requested respite, they were not required to formally re-Mirandize him before resuming their interrogation.
Questioning resumed at 4:50 p.m. with the police “reminding” defendant of his rights. That session lasted until defendant began to cry and requested to speak with his mother-in-law. That request was tantamount to an invocation of defendant’s right to silence. In Harvey I, supra, 121 N.J. at 418-22, 581 A.2d 483, we held that defendant invoked his right to silence when, on October 30, he requested to speak with his father. Id. at 419, 581 A.2d 483.
*223Defendant was emotionally upset by the end of the 4:50 p.m. session and wanted the advice of a trusted family member. The police properly ended the interrogation and allowed Harvey to meet with Pearl Thomas.
When the next interrogation began at 7:30 p.m., the police did not readminister Miranda rights. That omission violated the bright-line rule of Hartley. As defendant had invoked his right to silence at the end of the 4:50 p.m. session, the police failed to honor scrupulously his right to silence. Hence, we concluded that “[a]ny statements made prior to the new warnings must be suppressed.” Harvey I, supra, 121 N.J. at 420, 581 A.2d 483.
Defendant, however, made no incriminating statements during the 7:30 p.m. session. When defendant again became upset, the police stopped the questioning and returned him to his cell at around 8:00 p.m. The investigators’ failure to honor scrupulously defendant’s rights, therefore, did not prejudice defendant.
At 8:15 p.m., Harvey initiated efforts to speak with Detective Swanhart alone. As we concluded in Harvey I, supra, 121 N.J. at 418, 581 A.2d 483, there was “no evidence of police coercion or misconduct” on October 28. Furthermore, defendant was not subjected to “extended interrogations designed to wear down [his] will.” Ibid. Investigators provided defendant with lunch, dinner, cigarettes, and chewing tobacco. The police also accommodated defendant’s wish to have his mother-in-law brought to the jail to speak with him.
Defendant’s voluntary reinitiation of dialogue with Detective Swanhart relieved the police of the duty to readminister Miranda warnings. Fuller, supra, 118 N.J. at 84-85, 570 A.2d 429. Nonetheless, Swanhart administered those warnings before recording defendant’s confession.
Neither was defendant subjected to any police coercion on October 29. The record is devoid of evidence of coercion during the ear tour of West Windsor or afterward.
*224The consent-to-search form was based on information provided by defendant in his voluntary confession. Defendant knew what the police wanted to search, what they were searching for, and that he had the right to refuse consent. Nonetheless, defendant voluntarily and knowingly signed the consent-to-search form. Because of defendant’s consent, the police did not need a warrant. Schneckloth v. Bustamonte, 412 U.S. 218, 219, 93 S.Ct. 2041, 2043-44, 36 L.Ed.2d 854, 858 (1973).
The detectives thus lawfully discovered the Seiko-LaSalle watch. After searching defendant’s car, they tied the bag containing the watch, put the bag in the trunk, and notified the Plainsboro police department. Later, the Middlesex County Prosecutor’s Office obtained a valid search warrant and seized the watch.
- XI -
Defendant contends that the prosecution failed to produce sufficient supporting evidence of aggravating factor c(4)(f) to justify submission of that factor to the penalty-phase jury. Aggravating factor c(4)(f) applies to murders “committed for the purpose of escaping detection, apprehension, trial, punishment or confinement for another offense committed by the defendant or another.” N.J.S.A. 2C:11-3c(4)(f). Defendant avers that the killing itself cannot support a submission of the c(4)(f) factor. Rather, defendant argues that the State must produce proof that the defendant killed with the specific intent of escaping apprehension. Defendant also argues that the c(4)(f) factor impermissibly duplicates aggravating factor N.J.S.A. 2C:11-3c(4)(g) which pertains to the commission of a murder in the course of a felony (the c(4)(g) factor). His point is that all felons want to avoid capture. After careful consideration, we believe that the trial court properly submitted the c(4)(f) factor to the penalty-phase jury.
- A-
“The key to finding factor c(4)(f) is that the defendant intended to eliminate a potential witness to his crimes.” Martini *225I, supra, 131 N.J. at 281, 619 A.2d 1208. The c(4)(f) aggravating factor can apply in the context of a contemporaneous felony murder. Merely because a killing occurred during the course of a felony, however, does not, in itself, support submission of the c(4)(f) factor. Hightower I, supra, 120 N.J. at 422, 577 A.2d 99. Rather, the State has the burden of producing sufficient evidence from which a reasonable jury could conclude that at least one of the purposes motivating the killing was defendant’s intent to avoid apprehension for the commission of the underlying felony. Ibid.; State v. Loftin, 146 N.J. 295, 377, 680 A.2d 677 (1996). Because direct evidence of a defendant’s intent to avoid apprehension is rarely available, the State may establish a defendant’s motive through circumstantial evidence. Martini I, supra, 131 N.J. at 282, 619 A.2d 1208. Avoiding apprehension, however, need not be the killer’s sole motive. Ibid. The requirement that the State adduce evidence in support of the intent to avoid apprehension ensures that the c(4)(f) factor does not merely duplicate the c(4)(g) factor pertaining to the commission of murder in the course of a felony.
- B -
In the instant case, the prosecution presented sufficient circumstantial evidence to support the jury’s conclusion that one of defendant’s motives in killing Irene Schnaps was to eliminate her as a witness to his burglary of her apartment and to avoid apprehension and punishment for his crime. Defendant entered Sehnaps’s apartment by forcing open a glass patio door. Apparently Schnaps awoke and discovered defendant in her bedroom, which was the only room showing signs of disturbance. The jury reasonably could have inferred that defendant decided to kill Schnaps to prevent her from alerting her neighbors, calling the police, and later identifying defendant as the person who intended to rob her. See Loftin, supra, 146 N.J. at 377-78, 680 A.2d 677 (finding that State presented sufficient circumstantial evidence to support c(4)(f) factor even when defendant wore mask because *226victim “might have been able to identify his assailant’s voice, height, weight, and overall build”). Indeed, when questioned by police, neighbors in the apartment complex denied hearing any noises from Schnaps’s apartment. The State adduced sufficient evidence to enable a jury to conclude that at least one of defendant’s motives was to silence Schnaps as a potential witness to avoid apprehension for his crime.
- C -
In arguing to the penalty-phase jury that the State had met its burden in proving the existence of each of the aggravating factors beyond a reasonable doubt, the assistant prosecutor made the following statement concerning the c(4)(f) factor:
Aggravating factor number three is was this murder additionally committed in an attempt to avoid prosecution, apprehension? Yes, it was. Items were removed not only to be stolen and kept by Mr. Harvey, but items were removed from this apartment to detect or to prevent detection. I suggest to you bedding. The body of Irene Schnappes [sic] was washed clean in an attempt to prevent the police from locating the individual that committed this offense, Mr. Harvey.
Although defense counsel did not object, the argument was improper. Actions taken by a killer subsequent to the murder are irrelevant to the c(4)(f) determination. They do not inform the inquiry concerning the defendant’s motives for killing. See Hightower I, supra, 120 N.J. at 422, 577 A.2d 99; State v. Monturi, 195 N.J.Super. 317, 326-27, 478 A.2d 1266 (1984).
The trial court remedied the prosecutor’s error by properly instructing the jury that “[a]ny evidence of actions taken by the defendant to conceal the murder itself cannot be used to prove this aggravating factor.” Accordingly, the prosecutor’s remarks in summation did not constitute plain error.
- XII -
Defendant’s argument that the c(4)(f) factor violates the Eighth Amendment is without merit. As discussed above, c(4)(f) is not duplicative of the c(4)(g) factor.
*227- XIII -
Defendant argues that the trial court committed reversible error by instructing the jury that it could consider all of the guilt-phase evidence during its penalty-phase deliberations. He avers that the trial court should have instructed the jury on which evidence was relevant to the aggravating-factor inquiry and limited the evidence accordingly. Defendant contends that unguided jury consideration of the guilt-phase evidence impermissibly increased the likelihood that the jury would return a capital sentence. We disagree.
- A-
Instead of relating the guilt-phase evidence to the aggravating factors, the court told the jury that it was to consider all of the evidence at both the guilt and penalty phases. Defense counsel neither objected to the instruction nor requested that the court specify how the jury should relate the guilt-phase evidence to the penalty phase.
We acknowledge that the trial court should have instructed the jury “concerning the evidence that it may use in its penalty deliberations and the purposes for which that evidence may be used.” State v. Erazo, 126 N.J. 112, 133, 594 A.2d 232 (1991). To the same effect is the Bench Manual For Capital Cases, at 231-32 (November 1, 1996) ‘Manual ’ which states:
[W]here there is evidence introduced during the guilt phase which may have the capacity to prejudice penalty phase proceedings, (e.g., photographs, other-crimes evidence, movies), the court should provide instructions to the jury which delineate to what extent they may consider this evidence, and for what purposes, during the penalty phase.
The Model Charge advises, moreover, that when the “State is relying on facts established during a guilt phase verdict to prove an aggravating factor,” Manual at J-6 n. 8, the trial judge should give the following instruction:
However, the guilt and sentencing phases are considered as separate proceedings. The State contends that certain facts established by your verdict in the guilt phase
*228***
also prove the following aggravating factors
***
I am instructing you that it is your duty to deliberate again on these facts to determine whether they prove the aggravating faetor(s) the State alleges. You have the right to reach a different conclusion about whether these facts prove an aggravating factor than the conclusion you reached as to whether they proved guilt.
[Manual at J-6.]
Here, however, the court’s failure to designate specifically the evidence that the jury should consider did not constitute plain error. The court instructed the jury not to consider aggravating factors other than those alleged by the State and that it should consider evidence adduced at both phases of the trial. Defendant did not object or seek a more specific charge. Our review of the record reveals that the error was not of such a nature as to have been clearly capable of producing an unjust result. R. 2:10-2.
- B -
In summation, the prosecutor directed the jury to the evidence that supported the State’s allegation of aggravating factors. In support of the c(4)(c) factor — that the killing involved aggravated assault to the victim — the prosecutor invited the jury to:
Recall the testimony of Dr. Shuster to the nature and the extent of the wounds. Recall the testimony of Captain Rizzo as to where and how that blood was found in that bedroom, blood found underneath a hospital chair, ... on a towel where there was a fan, her body found on the other side of that bedroom laying back down, face up with blood splashed on the end table.
Dr. Shuster ... testified that these wounds to the head bleeds [sic] veiy profusely, that she would have become unconscious simply because there was such a loss of blood in the head area. Look at those wounds again, those photographs that I’m sure you looked at before. Look at those wounds.
This is not an attempt to kill with one blow. This is an attempt to commit, a successful attempt to commit severe bodily pain by Mr. Harvey, and he was successful.
That’s aggravating factor number one.
With regard to the c(4)(g) factor — that the murder was committed during the course of a felony — the prosecutor told the jury to *229consider the evidence indicating that defendant was in the midst of committing burglary and robbery when he killed Irene Sehnaps.
Although the prosecutor’s comments regarding which facts were relevant to prove the c(4)(f) (avoiding apprehension) factor were improper, the trial court properly informed the jury that “[a]ny evidence of actions taken by the defendant to conceal the murder itself cannot be used to prove this aggravating factor.” See supra Part XI.C.
Furthermore, the trial court correctly defined the elements of each of the charged aggravating factors and informed the jury it was not bound by any conclusions that it reached in the guilt phase.
The court’s instructions, when viewed against the background of the prosecutor’s summation, sufficiently informed the jury of the evidence concerning the aggravating factors.
The photographs of the crime scene and of the victim’s body were properly submitted to the penalty-phase jury as relevant to the c(4)(c) inquiry. See Moore, supra, 122 N.J. at 469, 585 A.2d 864 (holding that trial court had discretion to admit relevant photographs in penalty phase); Bey II, supra, 112 N.J. at 182, 548 A.2d 887 (same). Both the prosecutor and defense counsel referred to the photographs in summation only in the context of the c(4)(c) aggravating factor. Although the trial court properly should have instructed the jury to use these photographs only when considering the c(4)(c) factor, the omission was not prejudicial. The jury, moreover, did not unanimously return a finding of factor c(4)(e). We conclude that the photographs did not adversely impact the jury’s consideration of the c(4)(f) and c(4)(g) aggravating factors.
- C -
Defendant also challenges the admission into evidence of two photographs of the victim’s apartment that showed family photographs. He contends that the photographs constituted im*230permissible victim-impact evidence. We disagree. The State introduced the photographs to familiarize the jury with the crime scene and to demonstrate that no struggle had occurred in the victim’s living room. The prosecution never commented upon the photographs in any “manner that serve[d] only to highlight the victim’s virtues in order to inflame the jury.” State v. Williams, 113 N.J. 393, 452, 550 A.2d 1172 (1988) (Williams II).
-XIV-
Defendant contends that the prosecutor committed reversible error in his penalty-phase summation by inviting the jury to consider whether a non-capital sentence constituted “sufficient punishment.” We find that argument unpersuasive.
- A-
As part of his penalty-phase case, defendant presented Dr. Richard Moran, a professor of sociology at Mount Holyoke College, as an expert on the correlation between age and criminal behavior. Dr. Moran testified that as people get older, they generally are less likely to commit a violent crime. Based on that theory, Dr. Moran testified:
I believe that from Mr. Harvey’s birth date [1950] and the date that he would be first eligible for parole, and I stress here could be eligible for parole after thirty years, that would bring us up to the year 2014, and he would be 64 years old.
***
If he does get paroled when he’s first eligible at the age of 64, he will be in the lowest possible category [of individuals likely to commit crime]. And if you look at the fact that he committed the crime in his thirties ... people under 40 commit eighty-nine percent of the crime. He’ll move to the 60-plus category where he’ll be in the lowest possible risk category.
Accordingly, Dr. Moran concluded that “there would be a very tiny minute chance that [Harvey] would ever offend again.”
On cross-examination, the prosecutor questioned Dr. Moran about his assumption that defendant would first be eligible for parole in 2014. The prosecutor asked: “And that is twenty years *231from today or twenty years from 1994. Is that correct?” When Dr. Moran agreed, the prosecutor ceased cross-examination. •
In anticipation of Dr. Moran’s testimony, the State asked that the court instruct the jury that defendant’s thirty years -without parole would start from the date of his original incarceration, October 31, 1985, not from the end of the trial in December 1994. Defense counsel did not object.
In summation, defense counsel referred to Dr. Moran’s testimony regarding the age of the defendant at the time of the offense as a mitigating factor, N.J.S.A. 2C:11-3c(5)(c). Defense counsel argued to the jury:
What this says is that his age when he committed this offense was such that the earliest he could ever possibly get out would be sixty-four. And you heard testimony that would indicate that people in their sixties are very unlikely, only one percent of all violent crimes are committed by people of that age. You heard that testimony.
The prosecutor responded:
Is Mr. Harvey’s age a mitigating factor for you to consider? Sure you can consider it. You heard the testimony, the penalty, if it is not death, it is thirty years to life, minimum thirty years. What does that mean? Very simple, thirty years. What does that mean in this case? The year 2011, twenty years from now Mr. Harvey will be eligible for parole, twenty years from today. Is that sufficient punishment do you feel? You may consider that.
Defense counsel did not object.
- B -
Defendant now argues for the first time that the assistant prosecutor’s comment justifies reversal. According to defendant, this argument was tantamount to asking the jury to consider a non-statutory aggravating factor during their death-penalty deliberation. Furthermore, defendant argues that the prosecutor made this comment knowing that a non-capital sentence, combined with the periods of incarceration that defendant was already serving for unrelated crimes, would not enable defendant to be eligible for parole in 2014. We reject the argument.
*232- C -
Defendant’s penalty-phase case consisted of an attempt to convince the jury that a sentence of life with a thirty-year parole bar was a more appropriate sentence than death. The prosecutor asked the jury to consider whether such a sentence was justified. Rather than arguing a non-statutory aggravating factor to the jury, the prosecutor simply posed a rhetorical question designed to prompt the jury to consider the weight of the “age” factor, N.J.S.A. 2C:11-3c(5)(c).
The allegation that the prosecutor misled the jury by referring to the year 2014 as a potential parole date, is likewise meritless. Although defendant could not be eligible for parole in 2014, Dr. Moran included that date in his analysis. That conclusion was designed to keep from the jury defendant’s prior convictions for unrelated violent crimes. Because the prosecutor could not discuss defendant’s true prospects of parole, he confined himself to commenting on the evidence as presented by the defense. Defendant cannot have it both ways. He may not knowingly present an inaccurate date for parole eligibility to the jury, bar the prosecution from rebutting it with proof of prior sentences, and then claim error.
-XV-
Defendant argues that his entire trial was so infected with error that even if the alleged individual errors do not constitute reversible error, in the aggregate, the errors denied him a fair trial. State v. Orecchio, 16 N.J. 125, 129, 106 A.2d 541 (1954). We disagree. Defendant’s allegations of trial error, both singly and in aggregation, are without merit. We are satisfied that defendant’s trial was fair and that he suffered no undue prejudice.
- XVI -
Defendant argues that the death-penalty statute violates the prohibition against cruel and unusual punishment contained in *233the Eighth Amendment of the Federal Constitution. As this Court originally held in Ramseur, supra, 106 N.J. at 166-97, 524 A.2d 188, and recently restated in Harris, supra, 141 N.J. at 574, 662 A.2d 333, DiFrisco II, supra, 137 N.J. at 508, 645 A.2d 734, and Martini I, supra, 131 N.J. at 221-22, 619 A.2d 1208, we reject that argument. The New Jersey death-penalty statute is not violative of the protections offered by the Eighth Amendment.
- XVII -
Pursuant to N.J.S.A. 2C:11-3e, defendant requests that we determine whether his “sentence is disproportionate to the penalty imposed in similar cases, considering both the crime and the defendant.” We will review the proportionality of defendant’s sentence pursuant to a briefing and argument schedule to be established by the Clerk of the Court after consultation with counsel.
In conclusion, defendant’s conviction for the murder of Irene Schnaps and his sentence of death are affirmed.