HIGBEE, J.A.D.,
dissenting.
I respectfully disagree with the majority on several issues. The trial court excluded almost all of a defense expert’s testimony describing the scientific research on voice identification, as well as the expert’s opinions based on this research, for reasons that in some instances merely implicated the weight of the evidence, and in other instances were grounded in unprecedented and unsupportable precepts. The defense expert’s testimony undermined the testimony of, perhaps, the most credible witness to identify defendant, albeit by voice. Precluding defendant from presenting the evidence arguably in and of itself denied defendant a fair trial; but thei-e was more.
The trial court admitted all but one of the State’s expert’s opinions about street gangs, even though the prejudice to defendant — primarily demonstrating he was a high ranking leader of a dangerous gang — substantially outweighed the purported probative value of explaining other testimony. In addition, the trial court erred during the second trial by replacing a deliberating juror based on a limited and inadequate inquiry into the juror’s reasons for wanting to be excused.
Any one of these errors had the clear capacity to affect the outcome of the trials. Cumulatively, they leave no reasonable doubt defendant was denied fair trials.
It is undisputed that four women were fatally shot in the head during a robbery and left to die in their burning apartment. The State accused defendant, Rolando Terrell, of perpetrating the robbery of two of the victims, the arson of their apartment, and the execution of all four women. The State’s three key lay witnesses were two convicted felons who testified in exchange for sweetheart plea deals, and a young woman who survived the crimes by hiding in a closet. She identified defendant solely by his voice. Because my disagreement with the majority begins with the exclusion of expert testimony tending to undermine the voice identification, I begin with that issue and discuss in turn the two other issues on which I disagree with my colleagues.
I. EXCLUSION OP DEFENSE EXPERT’S TESTIMONY
My disagreement with the majority on this issue involves both the appropriate standard of review and the trial court’s application of legal principles governing the admission of expert testimony. The majority cites State v. Kuropchak, 221 N.J. 368, 385-86, 113 A.3d 1174 (2015), for the proposition that our review of a trial court’s evidential rulings is deferential. Kuropchak involved neither the admission of scientific evidence nor appellate review of a trial court’s decision to bar a defense expert’s scientific opinions in a criminal trial.
In State v. Torres, 183 N.J. 554, 874 A.2d 1084 (2005), the Supreme Court held a trial court’s evidentiary rulings excluding defense expert testimony in criminal trials are reviewed under an abuse of discretion standard, but with less deference than in other settings. The Court explained that when reviewing the admission of scientific evidence, the appellate court must evaluate the reliability of the proffered scientific evidence, noting:
While the trial court is in a better position to shape the record and make credibility determinations, “appellate courts can digest expert testimony as well as review scientific literature, judicial decisions, and other authorities.” The appellate court should carefully review the relevant authorities in determining the correctness of the decision to admit or exclude the disputed testimony. In short, the appellate court need not be as deferential to the trial court’s ruling on the admissibility of expert scientific evidence as it should be with the admissibility of other forms of evidence.
[Torres, supra, 183 N.J. at 567, 874 A.2d 1084 (citations omitted).]
There are other subtle variations in the standard of review of the admission of defense expert testimony in criminal cases. The admissibility of expert testimony is governed by N.J.R.E. 702:
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.
There are three requirements for expert testimony to be admissible: “(1) the ... subject matter [must be] beyond the ken of the average juror; (2) the field ... 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 .., testify].” State v. Townsend, 186 N.J. 473, 491, 897 A.2d 316 (2006) (quoting Torres, supra, 183 N.J. at 667-68, 874 A.2d 1084).
Appellate review of the second prong has moved steadily closer to de novo review. “An appellate court may independently review scientific literature, judicial decisions, and other authorities to determine whether proposed expert testimony is scientifically reliable and has obtained general acceptance so that it may be admitted in our courts.” State v. McGuire, 419 N.J. Super. 88, 123— 24, 16 A.3d 411 (App. Div.) (citing Torres, supra, 183 N.J. at 567, 874 A.2d 1084), certif. denied, 208 N.J. 335, 27 A.3d 948 (2011)). Although appellate courts continue to review a trial court’s rulings on prongs one and three for abuse of discretion, our review of a trial court’s exclusion of defense expert testimony in criminal cases requires consideration of “[N.J.R.E.] 702’s liberal approach favoring admissibility.” State v. Jenewicz, 193 N.J. 440, 456, 940 A.2d 269 (2008). With this in mind, the Supreme Court explained:
That the strength of an individual’s qualifications may be undermined through cross-examination is not a sound basis for precluding an expert from testifying as part of a defendant’s defense, even if it likely will affect the weight that the jury will give the opinion. Rather, a court should simply be satisfied that the expert has a basis in knowledge, skill, education, training, or experience to be able to form an opinion that can aid the jury on a subject that is beyond its ken.
[Id. at 455, 940 A.2d 269.]
Moreover, defendants in criminal cases have “a fundamental constitutional right to a fair trial, which necessarily includes the right to present witnesses and evidence in [their] own defense.” Id. at 451, 940 A.2d 269 (citing Washington v. Texas, 388 U.S. 14, 19, 87 S.Ct. 1920, 1923, 18 L.Ed.2d 1019, 1023 (1967)). This “fundamental right ... is protected not only by the Federal Constitution but also by Article 1, paragraph 1 of the New Jersey Constitution.” Ibid. For this reason, “the substantial liberty interest at stake for [a] defendant” is a factor that must be considered by a trial court in exercising its discretion to admit or deny expert testimony. Id. at 456, 940 A.2d 269. Significantly, “[N.J.R.E.] 702’s liberal approach favoring admissibility ... and the substantial liberty interest at stake for defendant [may] tip the scales in favor of finding error in the trial court’s preclusion of [a defense expert’s] testimony.” Ibid. That should have been the result here.
The trial judge held a pretrial hearing under N.J.R.E. 104 to determine the admissibility of the opinions of defendant’s expert, Dr. Steven Penrod. The testimony elicited at the hearing disclosed the following information. Dr. Penrod has a Bachelor’s degree from Yale, a J.D. from Harvard University, and a Ph.D. in social psychology from Harvard. After completing his education, Dr. Penrod served as an assistant and a full professor of psychology at the University of Wisconsin. He then became a faculty member at the University of Minnesota School Of Law before becoming program director for joint degrees in psychology and law at the University of Nebraska. At the time of the hearing, he held the title of distinguished professor in psychology at the John Jay College of Criminal Justice in New York City where he had worked since 2001.
Dr. Penrod described himself as a research psychologist. He testified his primary areas of research and study have focused on jury decision making and eye-witness identification. He has devoted much of his career to the study of law and psychology, specifically of memory encoding, recall, and identification by witnesses particularly in a criminal trial context. As a result, Dr. Penrod has been qualified as an expert and testified about memory formation and eye-witness recognition in various State and Federal courts well over 100 times. He testified he has also been qualified as an expert in voice identification cases approximately ten times in various courts, although voice identification is relevant in criminal trials considerably less often than eye-witness identification. In 2009, the Supreme Court appointed Judge Geoffrey Gaulkin, P.J.A.D., (retired and temporarily assigned on recall), as a Special Master to preside over a hearing on the reliability of eye-witnesses and prepare a report on his findings. Dr. Penrod was chosen as one of the experts to testify and offer his expert opinions at the hearing.
Within the scientific community that concentrates on the study of memory and witness identification, Dr. Penrod has been asked to speak and present his research papers at over 200 conferences involving psychologists from the United States and Europe. He authored or co-authored five books on eye-witness identification and published over 140 articles on jury decision making and witness identification. More than forty percent of his papers have been published in peer reviewed journals.
His research for the last thirty years, and continuing at the time of the hearing, was primarily funded by the National Science Foundation (NSF). He explained that only fifteen percent of proposals submitted to the NSF are approved for grants yearly, and at the time of the hearing, his research was being funded by two separate NSF grants.
Dr. Penrod has done his own independent research and studies on jury decisions, memory, and eye-witness identification; however, he acknowledged he had not conducted his own experiments or studies on voice identification. He studied the research on voice identification, and included the topic in one of the first articles he wrote after graduating from Harvard. Dr. Penrod reviewed research on the subject from the 1930s and a published study from 1944. However, until the 1970s, there had been little research in that specific field. He testified a “new flourishing body of research” was undertaken by psychologists on voice identification starting in the 1970s. He admitted there was still a much smaller body of reseai’ch limited to voice identification than was devoted to eye-witness identification. However, despite the smaller body of research, the investigations focused on the same factors and were “parallel streams of research.”
The doctor explained that recalling and identifying a face or a voice a witness has seen or heard before relies on the ability to encode and then reconstruct the memory. According to Dr. Pen-rod, a memory is not, as is commonly believed by jurors, a photograph or a tape recording sitting in the brain waiting to be retrieved. Rather, memory involves piecemeal construction and reconstruction and is fallible. He testified that psychologists study the same factors affecting reliability of memory reconstruction and eye-witness identification as those studied on ear-witness identification. Dr. Penrod provided a list of more than twenty scientific research papers specifically about ear-witness identification. He particularly referenced the work of the leading researcher on the topic, a Canadian psychologist named Dan Yarmey, who has published over ten research papers specifically on witness voice identification.
Dr. Penrod testified the opinions he gave in his report were based on scientific research on both eye-witness testimony and ear-witness testimony that were generally accepted in the scientific community. He advised the court he would not comment on the witnesses or the specific facts, as he never opines on a particular witness’s credibility. He explained he would simply testify about the science that could be applied by the jury to evaluate the evidence.
Following the hearing, defendant argued Dr. Penrod was a qualified expert in the field of voice recognition and that his testimony would assist the jury in understanding and evaluating the ear-witness testimony.
The State did not call an expert to challenge the reliability of the science underlying Dr. Penrod’s testimony; nor did it produce any evidence disputing the general acceptance by the scientific community of Dr. Penrod’s opinions on face and voice identification. The State moved to exclude the testimony of Dr. Penrod, arguing he was not qualified as an expert to give testimony on voice recognition; that the testimony he proffered was not outside the ken of the average juror; that he did not give an opinion about the facts of the case; and the jury should determine credibility without assistance.
The trial judge issued a written opinion and order excluding almost all of Dr. Penrod’s testimony. First, the judge acknowledged that although the evaluation of credibility of any witness was solely within the province of the jury, expert testimony in cases involving witness identification was admissible to demonstrate to the jury a witness may genuinely believe their identification is accurate even when it is incorrect. The trial judge stated:
In eases where expert testimony has been admitted to elucidate witness statements, it was therefore not to address whether the witness was giving truthful testimony, but rather to give the jury a “context in which to more realistically and fairly .., appraise and consider the witness’ perceptual accuracy.” The trend is toward admitting such testimony after ensuring that it meets the rules for the admissibility of expert testimony.
[ (Citations omitted).]
The trial judge proceeded to rule specifically on the admissibility of each separate topic presented in Dr. Penrod’s report based on the criteria set forth in State v. Kelly, 97 N.J. 178, 208, 478 A.2d 364 (1984), stating:
(1) the intended testimony must concern a subject matter that is beyond the ken of the average juror; (2) the field testified to must be at a state of the art such that an expert’s testimony could be sufficiently reliable; and (8) the witness must have sufficient expertise to offer the intended testimony.
In his analysis of the admissibility of evidence, the trial judge did not exclude any of the testimony based on the second prong of the three prong test. In fact, the prosecutor did not request exclusion of the testimony under prong two. Instead, the prosecutor and the trial judge — as does the majority — focused on an out-of-context quotation by the expert that he relied on “experimental” research. The prosecutor and trial court relied on this quotation when discussing the expert’s qualifications to testify, thus seemingly implying the science was in doubt, but making no finding that the subject of voice identification was not sufficiently accepted or reliable for admission. In fact, Dr. Penrod’s testimony included a description of studies based on results from scientifically controlled experiments and other types of studies that gathered data from field observations or statistical databases. Dr. Penrod at one point differentiated other types of research from “experimental” research where experiments were actually conducted. At no time did he use the word “experimental” in the context of “unproven.”
The following is a review of the rulings made by the trial judge on each of the reliability factors affecting ear-witness testimony as explained by Dr. Penrod.
A. STRESS
The testimony proffered by the expert on stress was based on studies showing that when an individual is under “flight or fight” stress, which is when they believe they are in immediate danger, their perception and ability to construct an accurate memory of a face or a voice is diminished. Their subsequent recall of that face or voice, and thus their identification of a defendant, is less reliable than if they were not under stress.
The judge barred Dr. Penrod from testifying on the effect of stress on the reliability of ear-witness testimony, thus significantly limiting the defendant’s ability to challenge the survivor’s voice identification of defendant. The judge found the stress testimony was not within the ken of the average juror. Specifically, he found “information concerning stress and perception appears to be of potential help to the trier of fact in understanding the identification testimony of [the survivor].” The judge made a general finding that the field of study demonstrating stress diminishes the capacity for perception is “at such a state of the art that it can be reasonably relied upon.” Indeed, Dr. Penrod testified at the hearing there is general acceptance within the scientific community of the studies affecting both eye-witnesses and ear-witnesses. The judge stated: “cases evaluating the admissibility of expert testimony considering ‘ear-witness’ identifications have drawn correlations to, or seen it as a subset of, expert testimony concerning eye witness testimony.”
However, the judge barred the testimony on stress’s effect on voice identifications holding “it appears that Dr. Penrod is therefore unqualified to testify concerning the effects of stress on voice recognition, and that such testimony would present a risk of misleading the jury.” The judge reasoned Dr. Penrod was unqualified because he never conducted his own studies on voice identification. The State in their brief does not cite any precedential published case decided by any New Jersey court that one must have conducted their own experiments or studies to be qualified as an expert witness. To the contrary, N.J.R.E. 702 states “a witness qualified as an expert by knowledge, skill, experience, training, or education may testify thereto in the form of an opinion or otherwise.” (Emphasis added). The rule does not require that an expert personally conduct experiments.
The trial court cites to one case from 1992 where a federal judge barred Dr. Penrod’s testimony on stress and eye-witness testimony. In United States v. Nguyen, 793 F.Supp. 497, 515-16 (D.N.J. 1992), the federal district court judge found one of the reasons for barring his testimony was Dr. Penrod had not conducted his own experiments. Nguyen, which the State relied upon and which the trial judge discussed at length in his opinion, presents two legally unsupported and untenable assertions which fly in the face of legal precedent in this State and elsewhere.
The first is the assertion that not having done original research or experiments on a particular scientific topic is a basis for finding an expert is not qualified to give an opinion in that field. The second is the assertion that a subject is not outside the ken of the average juror if the jurors could read the scientific studies themselves, and understand them.
Both of those misguided assertions were repeated by the trial judge here as reasons to bar Dr. Penrod’s testimony. Nguyen, however, does not reflect New Jersey controlling law in 1992 or in 2011, or the law in any other state. The trial judge here cited several other federal court decisions from the 1990⅛ that admitted eye-witness expert testimony, including a 1991 decision by the Third Circuit that admitted the testimony of Dr. Penrod. See U.S. v. Stevens, 935 F.2d 1380, 1397 (3d Cir. 1991)
The trial judge, again borrowing from the legally unsound reasoning of Nguyen, further held: “It is likely if the jury were provided documentation of the study, they would be able to come to the same conclusions ....” This reasoning is incompatible with our Rules of Evidence.
In 1991, the Supreme Court adopted the federal rule of evidence on the admission of learned treatises in Jacober v. St. Peter’s Medical Ctr., 128 N.J. 475, 495, 608 A.2d 304 (1992). Before that decision, published textbooks, research papers or articles describing the current scientific knowledge in a field were not admissible except when used on cross-examination of a witness who acknowledged them as authoritative. In Jacober, the court adopted the federal evidence rule that allowed the admission into evidence of learned treatises. What prevents the jury from considering as evidence any article published by anyone anywhere is the requirement that the treatise be vouched for by an expert on the stand.
As Justice Stein wrote: “Recently, we noted that in determining reliability ‘[t]he focus should be on what the experts in fact rely on, not on whether the court thinks they should so rely.’ ” Jacober, supra, 128 N.J. at 495-96, 608 A.2d 304 (quoting Ryan v. KDI Sylvan Pools, Inc., 121 N.J. 276, 289, 579 A.2d 1241 (1990)). Thus, a jury can consider a learned treatise only when an expert can testify to its reliability, explain it, and be cross-examined about it. Justice Stein explained this principle as follows: “[t]he rule’s emphasis on expert guidance, as well as its prohibition on the receipt of learned treatises as exhibits, limits the risk that factfin-ders will misunderstand or misapply learned-treatise statements and discourages the use of learned treatises as substitutes for expert testimony.” Id. at 491, 608 A.2d 304.
Following the Jacober decision, N.J.R.E. 803(c)(18) was adopted and learned treatises were made an exception to the hearsay rule under the condition they be introduced by an expert on the stand and not be given directly to the jury. The Rule states:
To the extent called to the attention of an expert witness upon cross-examination or relied upon by the expert in direct examination, statements contained in published treatises, periodicals, or pamphlets on a subject of history, medicine, or other science or art, established as a reliable authority by testimony or by judicial notice, If admitted, the statements may not be received as exhibits but may be read into evidence or, if graphics, shown to the jury.
[N.J.R.E. 808(c)(18) (emphasis added).]
The trial judge’s statement that if the jurors were given the documentation they could come to the same conclusion as the expert as a reason to bar expert testimony is incomprehensible in light of N.J.R.E. 803(c)(18)’s explicit preclusion of statements in learned treatises being received as exhibits. The question is, does the average juror know the information and the scientific acceptance of the information? The trial judge made the inconsistent finding that a distinguished expert in the field of witness identification did not know enough to explain the research, but the jury would somehow know and understand it if given the studies. Jurors not only cannot be given the studies, but are instructed to not use the internet or do their own research as they cannot consider studies that are not in evidence through the testimony of an expert.
A finding that Dr. Penrod was not qualified in the face of the evidence that was presented on his qualifications was as unsound as the court’s statement the jurors themselves could i’ead and understand the studies. The ruling constituted a clear abuse of discretion.
B. WITNESS CONFIDENCE
The survivor testified she heard defendant speak only once before the date of the crime. She could not identify who he was by his voice when interviewed by police immediately after the crime. The next day she spoke to her boyfriend and she testified, while talking to him, she realized the voice she had heard was defendant’s voice. She did not report this to the police until they called her to come into the station two days after the crime. At that point, the co-defendant had already identified defendant as the murderer to his friend, his brother, and the police. The police advised the survivor they had a suspect. The police did not provide her with a voice identification test similar to a line up or photo identification test. The witness identified defendant’s voice only by her stated recollection. At trial she testified she was 100 percent certain about her identification of defendant.
Dr. Penrod would have testified that studies generally accepted by the scientific community demonstrate there was only a very modest correlation between the level of confidence asserted by a witness and the accuracy of their identification. Yet, research conducted by Dr. Penrod and others showed that jurors are systematically influenced by the level of confidence the witness displays in their identification and do not understand how little that means in terms of accuracy of the identification.
The doctor further opined that the level of confidence of a witness is influenced by subsequent events unbeknownst to the witness. He explained even a response as slight as the researcher saying “good” when the identification was made would substantially increase the confidence of the witness. Other studies demonstrated that when the witness became aware the person they identified had been identified by someone else, their confidence in the identification was substantially increased.
The trial judge ruled he would allow testimony regarding witness confidence being only modestly correlated with the accuracy of an identification, but would not allow Dr. Penrod to testify that confidence was increased or decreased by feedback. The judge again discussed the decision in Nguyen pointing out that the federal judge had barred Dr. Penrod from giving any testimony about confidence and accuracy.
The trial judge explained: “Dr. Penrod’s [proffered] testimony that subsequent events may effect a witness’ confidence in her testimony is a matter of common sense, and is not outside the knowledge of the average juror.” The evidence provides no support for this finding.
To the contrary Dr. Penrod testified:
Well, there have been a number of surveys of the general public and what they believe about the influence of a variety of factors that have been the subject of research investigation, and whether they have the same understanding of these effects as research psychologists.
The common pattern of findings is that some portion of the general public shares a belief that matches the research findings, but another substantial poi’tion of the general public doesn’t share that belief.
The average juror does not necessarily have any knowledge about the impact of even slight feedback. The judge’s decision to bar this testimony was arbitrary. He did not examine the totality of Dr. Penrod’s testimony and abused his discretion.
0. LISTENERS’ VIEW OF FACE
Dr. Peni’od testified that there is an inherent level of unreliability in witness voice identification which is increased when a witness looks at the person while hearing them voice. He described studies where individuals listened to a voice and then had to identify the voice. In the other arm of the studies, individuals saw a person’s face as they spoke and then had to identify the voice. The results demonstrated that when a person hears just a voice there is a better chance they will be able to subsequently identify it than if they see the person and hear the voice simultaneously. The reason proffered by Dr. Penrod is that people focus more on faces than on voices, and therefore do not create as accurate a memory of the voice.
This is information the jury should have been allowed to hear because it could have assisted them in their evaluation of the identification of the defendant’s voice by the survivor. The one time she heard defendant speak before the crime was committed, she was simultaneously looking at him. She next heard his voice on the night of the crime. It was from her memory of the prior occasion that she was able to identify his voice. According to the research, the reliability of her memory could be affected if the memory of his voice was developed while viewing defendant.
The judge again used the language and reasoning of the district court judge in Nguyen to bar Dr. Penrod’s testimony, finding he was not qualified to testify as an expert because he relied upon the studies of other researchers, did not provide his own analysis, and provided only a “minimal recitation of the facts and processes underlying the research.” The judge also noted, “if the jury were provided documentation of the study, they would be able to come to the same conclusions.” The indisputable error in that statement has already been discussed. Such a misunderstanding and misstatement clearly constitutes an abuse of discretion. See Moraes v. Wesler, 439 N.J. Super. 375, 378, 109 A.3d 218 (App. Div. 2015) (noting abuse of discretion arises when, among other things, a decision impermissibly departs from established policies, rests on an impermissible basis, or is based upon consideration of irrelevant or inappropriate factors).
D. UNCONSCIOUS TRANSFERENCE
Dr. Penrod testified that sometimes a witness is influenced to identify the wrong perpetrator when they have some limited prior exposure to that person. He described several studies including one where researchers staged thefts both in classroom and in street settings. Witnesses were asked to identify the thief. Twenty-five percent of witnesses wrongly identified an innocent bystander, unconsciously transferring a face they had some memory of to the perpetrator. The jury should have been permitted to consider unconscious transference in evaluating the reliability of the witness identification.
The studies described by Dr. Penrod on this topic did not include specific studies of voice identification versus eyewitness identification. However, the judge did not find his opinions on voice identification were unscientific. Rather, he again barred the testimony finding it could be misleading because Dr. Penrod did not do his own studies or “apply his own analysis or expertise to it.” He also found the jury “would be able to come to the same conclusions” if they were given the studies to read. Barring this testimony on that basis, which has no foundation in the law, was an abuse of discretion.
E. DURATION OF EXPOSURE TO VOICE
Dr. Penrod testified at the hearing that studies show the longer a witness is exposed to a voice, the more accurate the witness identification of the voice. The trial court barred expert testimony on both factors.
As to the proffer that the longer the exposure the more accurate the identification, the trial judge barred the testimony and found this was common sense and within the ken of the average juror. The judge was correct on that limited finding. The expert, however, was actually proffering this testimony to explain the common understanding that duration of exposure is equated to accuracy is only true “if all other [factors] are equal.” Dr. Penrod referenced a study conducted on military personnel who were interrogated for forty minutes face-to-face in both high stress and low stress situations. The effect of very high stress was of such significance that even though the interrogation was of long duration the accuracy rate of subsequent identifications was only twenty-seven percent.
Dr. Penrod also used the concept of duration of time to explain that although studies show that longer time exposure improved accuracy, if the exposure was broken up in several episodes, the observer would have better recall. For example, if one hears a voice for a period of time, then there is a break in time, and then exposure begins again, the witness will usually have more accurate recall than if the exposure to the voice was for the same length of time but was continuous. Thus, Dr. Penrod had more to offer to the jury than the isolated concept that the longer one is exposed to a voice, the more accurate the identification. For this reason, the judge’s finding, which did not address the entirety of what was being proffered and took Dr. Penrod’s testimony out-of-context, was made arbitrarily and was an abuse of discretion.
F. TIME ESTIMATIONS
The doctor’s second opinion about time of exposure concerned the overestimation by test subjects of the length of time of non-routine brief events. A witness’s description of how long a period of time an event lasted, and thus how long they were exposed to a persons’ voice or face, according to Dr. Penrod, is almost always overestimated by a significant amount when the event is brief and not routine. Researchers found witness time estimates are significantly more accurate for longer events than shorter events. The trial judge barred this testimony because it was within the ken of the average juror, but offered no support for this holding. Some jurors may believe it to be true and others may not, but few, if any, jurors will have read the scientific literature and know what studies have demonstrated. Nor could they use them in the jury room without expert testimony.
The judge, in barring this testimony and several of the factors that follow, also held that the accuracy of the witness’s testimony could be tested based on the witness’s demeanor and through cross-examination. This finding by the court demonstrates a lack of comprehension of the entire purpose of allowing expert testimony on witness identification. Normally jurors can evaluate credibility because deciding whether someone is telling the truth is a skill learned through experience. The problem with witness identification is it can be mistaken for reasons explained by Dr. Penrod, even in the absence of a motive to lie that can be revealed on cross-examination. The witness may be an honest person with good intentions who sincerely but mistakenly believes they are telling the truth. The witness’s demeanor will be that of a truth teller if they believe they are being truthful.
Finally, the judge again found the expert just recited the findings of others as opposed to doing original research, and that the jury could read and understand the studies themselves. These reasons are unsupported by the law and demonstrate an abuse of discretion.
G. CHANGES IN SPEAKER’S TONE
Dr. Penrod testified another factor that affects voice recognition is whether the speaker’s tone of voice is the same. He described a Canadian study where individuals were played a recording of a speaker talking in a normal non-emotional voice and subsequently were asked to identify that voice from other voice recordings, including the original speaker speaking in very emotional tones. Here, the survivor first heard defendant’s voice in a conversation while sitting in a truck, then two weeks later at the crime scene where the speaker was shouting and threatening in an angry voice. The effect a change of tone can have on the reliability of an identification could be helpful to the jury in evaluating reliability.
The trial judge barred it, however, because he found it was within the ken of the average juror that hearing a voice in a different tone would make it more difficult to identify. What the judge ignored is that the study described by the expert demonstrated statistically that subjects in the study who heard a voice in a different tone were no more statistically likely to identify the right voice than would have occurred by chance, or in other words, just guessing. While a juror might assume through common sense that a change in tone might affect the accuracy of a voice identification, the experiments and the research demonstrate that tone of voice is very significant in terms of recollection. The trial judge did not look at the totality of the information Dr. Penrod proffered. The barring of testimony applying this limited examination of the proffered testimony was an abuse of discretion.
H. COMPETING VOICES
Dr. Penrod’s proffered testimony on competing voices was that the reliability of voice identification is diminished when there are other voices and sounds heard at the same time. He offered no additional information. The trial judge barred the testimony because he found this is common sense. There was no error in this ruling.
I. RETENTION INTERVAL
The trial judge found Dr. Penrod’s testimony on the effect of the duration of time between hearing the voice and the identification of the voice was not beyond the ken of the average juror, and he barred it as falling within the realm of common sense. The trial judge oversimplified Dr. Penrod’s testimony by describing it as simply an opinion that the longer the time period between the identification and the crime, the less accurate the identification.
In fact, there are two time periods in this case possibly affecting the reliability of the identification. First, the survivor heard defendant’s voice two weeks before she heard it during the crime. Second, there was a different period of time following the crime before she was able to identify the voice as defendant’s. The studies Dr. Penrod described in his testimony relate to how the duration of time between first hearing a voice, and then hearing it a second time, impacts the accuracy of the identification. Although the average juror would have a common sense understanding that the longer the time gap, the less accurate the identification will be, Dr. Penrod proffered more scientific information that is not common sense and does fall outside the ken of the average juror. Dr. Penrod testified:
[M]ost people do recognize that, with the passage of time, we lose information. What people typically will not understand, and what the research shows — and, indeed this is research dating back to the 1880s with regard to memory — is that the loss of memory is most rapid in the first few hours and then days following an event.
Dr. Penrod testified that he, along with other colleagues, have done a meta-analysis of the research on loss of face memory, which demonstrated that the most rapid loss of memory occurred in the first ten hours. Specifically, in voice recognition studies, he testified other researchers found between a nineteen and sixty-nine percent accuracy in identifying a voice after two weeks. This is not information within the ken of the average juror, and it was an abuse of discretion to bar the testimony on this basis.
J. VOICE VERSUS FACE RECOGNITION
The trial judge barred Dr. Penrod’s testimony that voice identification is substantially less accurate than face identification, i.e., eye-witness versus ear-witness. The judge again found this was common sense and not beyond the ken of the average juror. The fact that it is easier to identify a face than a voice may not be common knowledge. Dr. Penrod’s testimony went beyond the general knowledge the average juror might have as to the level of accuracy of voice identification. The research demonstrates, according to Dr. Penrod, “dramatically worse” results for voice recognition. He discussed the specific differences and study results in his pretrial testimony. It was an abuse of discretion to bar all of his testimony because of a finding the average juror may have some information based on his or her own experience, without recognizing the average juror is unlikely to know the accepted scientific research that puts the juror’s general understanding in the context of the magnitude of difficulties in voice recognition.
K. WITNESS IDENTIFICATION GENERALLY
Dr. Penrod proffered testimony about a body of research that shows witness identification by both eye-witnesses and ear-witnesses is much more unreliable than the average juror would expect. The three Kelly criteria were met by this proffer, but the trial judge barred the testimony finding inexplicably this could be explored on cross-examination. I disagree.
There is no lay witness that can be cross-examined about the research that has found the reliability of witness identifications to be problematic. The judge’s determination this was within the ken of the average juror is not supported by the studies relied on by the expert. According to Dr. Penrod, the research demonstrates that jurors have mixed understandings about the reliability of witness identification and the factors that affect it. The trial judge barred this area of testimony because he found that “while courts have allowed testimony concerning special factors affecting identification accuracy, it does not appear that courts have admitted testimony concerning the general reliability or unreliability of eyewitness identifications.” It is understandable that this lack of prior precedent would cause a trial judge to pause and consider the reason for the same. However, evidence cannot be excluded because it has not been found admissible in prior cases. As we explained in State v. Burr, 392 N.J. Super, 538, 557, 921 A.2d 1135 (App. Div. 2007), affd as modified, 195 N.J. 119, 948 A.2d 627 (2008):
Our court rales allow the admission of “all relevant evidonce” that is not otherwise excluded by law. Relevant evidence is defined as any evidence that has “a tendency in reason to prove or disprove any fact of consequence to the determination of the action.” In determining whether proffered evidence is relevant, the trial court should inquire as to whether a “logical connection” exists between the evidence and a fact in issue. Stated another way, if the evidence renders a desired inference more probable or logical, then the evidence should be admitted. The test for relevancy is a broad one that generally favors admissibility.
[ (Citations omitted).]
Considering this standard for admissibility of relevant evidence, it was an abuse of discretion to preclude Dr. Penrod’s testimony about studies on the general lack of reliability of witness identification. Dr. Penrod could be cross-examined by the State about any flaws in the studies themselves, his reasoning, and the applicability of eye-witness studies to voice identification.
L. RECONSIDERATION BY THE COURT
After the judge ruled on the admissibility of Dr. Penrod’s testimony, the Special Master’s Study ordered by the Supreme Court was completed and released. Defendant made a motion for reconsideration on the admission of Dr. Penrod’s testimony. The trial judge denied the motion, and stood by his prior decisions, except for one.
The judge addressed in some detail two specific areas discussed in the report. The Special Master found that jurors tend to underestimate the importance of memory decay. The trial judge acknowledged this contradicted his prior finding that this area of testimony was within the ken of the average juroi\ Nonetheless, he ruled it was barred as being common sense, based on one statement, taken out-of-context, where Dr. Penrod said the general principle was common sense. The judge overlooked the balance of the expert’s testimony.
The second area the judge addressed was defendant’s argument that because the witness was told by the police they had a suspect before she identified defendant, this might have tainted the identification. The judge applied the same flawed criteria he relied upon to bar other testimony, namely, he would only consider allowing the testimony into evidence if Dr. Penrod had conducted his own study on this factor. Defense counsel did not contend Dr. Penrod performed such a study.
In view of the trial judge’s rulings — which eviscerated the expert’s opinion — defendant did not call the expert as a trial witness to testify about two factors: (1) the effect of stress on eyewitness testimony, not including the effect of stress on voice identification or the crossover between the studies; and (2) studies showing a witness’s confidence level and the accuracy of his or her identification are not correlated, not including the studies showing how outside factors can affect the witness’ confidence level. The majority faults defendant for not presenting Ms expert’s eviscerated opimon, stating: “On appeal, defendant claims ‘the limited nature of testimony that he would be permitted under the Court’s ruling’ neutralized its effectiveness. We disagree.” Our agreement or disagreement with defendant’s decision is irrelevant. The question is whether defendant was deprived of “a fundamental constitutional right to a fair trial, wMch necessarily includes the right to present witnesses and evidence in Ms own defense.” Jenewicz, supra, 193 N.J. at 451, 940 A.2d 269. “[N.J.R.E.] 702’s liberal approach favoring admissibility ... and the substantial liberty interest at stake for defendant [may] tip the scales in favor of finding error in the trial court’s preclusion of [a defense expert’s] testimony.” Id. at 456, 940 A.2d 269. Here, the trial judge not only overlooked these considerations, he barred defendant from presenting evidence in Ms own defense based on unsupported assumptions and a misunderstanding of the rules of evidence.
The law in the area of witness identification is still developing. The Supreme Court’s decision in State v. Henderson, 208 N.J. 208, 27 A.3d 872 (2011), is not retroactive, but the evidence that was wrongfully barred in this ease was admissible under our ease law at the time of the trial. The tidal court’s rulings barring testimony of Dr. Penrod should be reversed for the reasons and with the exceptions set forth above.
II. ADMISSION OF TESTIMONY BY STATE’S GANG EXPERT
The trial judge allowed the State to call Lieutenant Earl Grave as an expert on gangs. Lieutenant Graves works for the Essex County Prosecutor’s Office. Defendant argues that while Lieutenant Graves’ testimony may have been otherwise admissible, it should have been barred under N.J.R.E. 404(b), which states:
Except as otherwise provided by Rule 608(b) evidence of other crimes, wrongs, or aets is not admissible to prove the disposition of a person in order to show that such person acted in conformity therewith, Such evidence may be admitted for other purposes, such as proof of motive, opportunity, intent, preparation, plan, knowledge, identity or absence of mistake or accident when such matters are relevant to a material issue in dispute.
It is accepted that even when the jurors are instructed that prior crimes or bad acts are being admitted into evidence for reasons other than demonstrating defendant’s propensity to engage in criminal activity, such evidence is uniquely prejudicial and inflammatory. The Supreme Court recognized this as we acknowledged in State v Hernandez, 334 N.J. Super. 264, 269-70, 768 A.2d 1139 (2000), affd as modified. 170 N.J. 106, 784 A.2d 1226 (2001):
Because of the “widespread agreement that other-crimes evidence has a uniquo tendency to turn a jury against the defendant ...,” State v. Stevens, 115 N.J. 289, 302, 558 A.2d 883 (1989), Idle compromise between the antagonistic interests that the Rule seeks to effect can be achieved only by the most delicate balancing. As Stevens, supra, at 303, 558 A.2d 833, explains, “ffltis this inflammatory characteristic of other-crimes evidence that mandates a careful and pragmatic evaluation by trial courts, based on the specific context in which the evidence is offered, to determine whether the probative worth of the evidence outweighs its potential for undue prejudice." The tension between undue prejudice to the defendant and probative value to the State to prove a fact legitimately in issue induced the Supreme Court in State v. Cofield, 127 N.J. 328, 338, 665 A.2d 230 (1992), to articulate further the conditions of admissibility of other-crimes evidence, the Court defining those conditions as follows:
1. The evidence of the other crime must be admissible as relevant to a material issue;
2. It must be similar in kind and reasonably close in lime to the offense charged;
3. The evidence of the other crime must be clear and convincing; and
4. The probative value of the evidence must not be outweighed by its apparent prejudice.
[ (Alteration in original).]
Defendant argued under the first and last Cofield prongs the admission of the prosecution’s gang expert’s testimony in both trials was error. Lieutenant Graves’ testimony was of little or no probative value, and had such a strong potential for prejudice it should have been excluded.
The State argues the testimony by the gang expert was admissible as relevant to several material issues in dispute.
A. “GIRLFRIEND” AS GUN
Defendant was intercepted on a wiretap telling a man he was coming by the next day to pick up his “girlfriend.” The man testified for the State that he had a gun belonging to defendant and the gun was the “girlfriend.” He further testified he returned the gun to defendant the day before the murder. One reason the State gave for calling the gang expert was to explain that on the street, “girlfriend” means gun, This could have been explained without a reference to gangs. Instead, Lieutenant Graves was asked:
Prosecutor: In your experience, have blood gang members, or gang members in general, used coded words to, I guess, hide their activities?
Expert: Absolutely, code words,
Prosecutor: How about the weapons? Do they use any code words to hide the meaning of a handgun or any weapon?
Expert: Yes.
Prosecutor: What code words or phrases do they use?
Expert: Girlfriend or wifey.
Prosecutor: And what is the purpose of using these coded words?
Expert: To, uh, conceal it from law enforcement, or just hide it.
These references to illegal gang activity were unnecessary and had limited probative value. The jury heard the tape. The same testimony offered by the expert about slang used for guns came in from the man who had the gun. No one on the jury had any reason to think defendant had left his actual girlfriend with the man for days and wanted to pick her up.
B. USE OF JEEP
Lieutenant Graves testified that a lower ranking gang member had to lend his personal property to a higher ranking member of the gang:
Prosecutor: What about personal property within the gang? How is that shared or used?
Expert: It’s shared with other members, and, of course, the higher you are up the food chain, the hierarchy, the more power you would have and influence and use of whatever you want, just as — I’m a captain. If I want something, you know, I’ll use something one of my — one of my lieutenants or detectives have, I’ll just tell them to give it to me, order ’em.
Because the red Jeep that was identified as being driven to and from the crime by defendant was not owned by defendant, the State asserts the expert testimony was needed to show why defendant had use of the Jeep, even though the expert did not mention the Jeep directly. However, there was testimony by other witnesses that defendant was seen driving the Jeep on other occasions, and that it was owned by a man who let him use it. There was no need to show this was because of gang ties. Additionally, there was testimony that the police had stopped defendant for traffic violations while in the Jeep. There was no dispute defendant frequently drove the Jeep. The gang expert testimony had little or no probative value because defendant’s use of the Jeep was available and given by other witnesses and not even directly addressed by the expert. Under these circumstances, the references by the gang expert did little more than remind the jury defendant was a high ranking gang member.
C. OPPORTUNITY
The victim, who defendant was convicted of robbing in the first trial and murdering in the second, was the girlfriend of a man in jail. The boyfriend and victim were drug dealers. The co-defendant testified defendant told him he was going to rob the victim and it would be easy. The State established through testimony that defendant had been to the victims’ home before the murder. The State claims the gang expert was called to explain that defendant had the opportunity to gain entry into the residence. The State asserts that since defendant was a higher ranking gang member than the victim’s boyfriend, the victim, who was not a gang member, had to admit defendant into her home. The gang expert never gave this proffered testimony and it would have been unnecessaiy as there was other testimony defendant was allowed in the home and knew the victim’s boyfriend was in jail.
D. TATOOS
Lastly, the State asserts the gang expert’s testimony about gang tattoos had probative value. The co-defendant testified defendant showed him a tattoo on his neck after the crime, telling the co-defendant that he did not need any “codefendants.” The co-defendant testified he knew defendant was a gang member after seeing the tattoo. He also testified seeing the tattoo made him nervous because he knew the defendant’s gang was vicious, and he feared for his family’s safety. It was unnecessary, therefore, for the gang expert to explain that gang members had tattoos to identify them as part of a gang. Defendant presented no evidence he was not part of a gang or the tattoo was not a gang symbol.
The gang expert testimony simply repeated facts already established by the lay witnesses. The probative value was limited or non-existent. The prejudice, however, was significant because the gang expert testimony was not limited by the judge to the testimony the State claims was relevant.
E. PREJUDICIAL EFFECT
The error in the admission of expert testimony is that it included prejudicial testimony of defendant’s involvement with the “Bloods” and specifically included Lieutenant Graves’ history and habits of the Bloods. This history included references to the Bloods’ feud with the Crips, which had no relevance to the crime. This testimony was followed by the expert’s identification of defendant as a top leader of the Bloods gang, or as he described it, an original gangster or founder of the gang. These gang names are well known, and the public perception is that people who belong to these groups are bad people with a propensity to commit crimes. Even if any of the jury had not heard of the Bloods before the trial, they knew about their involvement with law enforcement by the trial’s end; a fact relevant to no material issue in dispute. Defendant’s gang membership would have come before the jury, but not with the same impact as when a member of law enforcement testifies about the gang, its history, its hierarchy, and law enforcement’s prior focus on and encounters with gang members.
The probative value of the references to defendant’s gang ties were substantially outweighed by its undue prejudice. See N.J.R.E. 403 (“evidence may be excluded if its probative value is substantially outweighed by the risk of [ ] undue prejudice”). Gang references are admissible only if N.J.R.E. 404(b) is satisfied. “Other crimes evidence is considered highly prejudicial.” State v, Vallejo, 198 N.J. 122, 133, 965 A.2d 1181 (2009). “The prejudice of other-crime evidence is its tendency to demonstrate a criminal predisposition; therefore, it poses a distinct risk that it will distract a jury from an independent consideration of the evidence that bears directly on guilt itself.” State v. G.S., 145 N.J. 460, 468, 678 A.2d 1092 (1996). An individual may not be convicted merely for belonging to an organization that advocates crime. Scales v. United States, 367 U.S. 203, 220-21, 81 S.Ct. 1469, 1481-82, 6 L.Ed.2d 782, 796, reh’g denied. 366 U.S. 978, 81 S.Ct. 1912, 6 L.Ed.2d 1267 (1961).
III. REPLACEMENT OF A DELIBERATING JUROR
After deliberations commenced in the retrial, two jurors requested to be excused. Defendant contends the judge’s inquiry and conclusory findings were flawed, and replacement of one juror, over defendant’s objection, rather than declaring a mistrial, was error.
A courts substitution of an alternate juror is limited by Rule 1:8 — 2(d)(1), which provides in relevant part:
Following the drawing of the names of jurors to determine the issues, the comí may in its discretion order that the alternate jurors not be discharged, in which event the alternate jurors shall be sequestered apart from the other jurors and shall be subject to the same orders and instructions of the court, with respect to sequestration and other matters, as the other jurors, If the alternate jurors are not discharged and if at any time after submission of the case to the jury, a juror dies or is discharged by the court because of illness or other inability to continue, the comí may direct the clerk to draw the name of an alternate juror to take the place of the juror who is deceased or discharged. When such a substitution of an alternate juror is made, the court shall instruct the jury to recommence deliberations and shall give the jury such other supplemental instructions as may be appropriate.
We review a trial court’s decision to substitute an alternate juror for an abuse of discretion. State v. Musa, 222 N.J. 554, 564-65, 120 A.3d 214 (2015). We are guided by certain principles in applying this standard; Rule 1:8 — 2(d)(1) “delicately balances two important goals: judicial economy and the right to a fair jury trial.” State v. Jenkins, 182 N.J. 112, 124, 861 A.2d 827 (2004). The Supreme Court has explained that juror substitution should only be invoked as a last resort because it “poses a clear potential for prejudicing the integrity of the deliberation process,” State v. Hightower, 146 N.J. 239, 254, 680 A.2d 649 (1996). With this in mind, “[t]he court must be prepared to declare a mistrial if a substitution would imperil the integrity of the jury’s process,” State v. Ross, 218 N.J. 130, 147, 93 A.3d 739 (2014). In making its decision, “the trial court must determine the cause of the juror’s concern and assess the impact of the juror’s departure on the deliberative process.” Ibid. Additionally, the court “must ascertain whether a reconstituted jury will be in a position to conduct open-minded and fair deliberations.” Ibid.
Here, the trial judge was confronted with the difficult task of learning the source of the juror’s distress without asking a question that might elicit information about the jury’s deliberations. To avoid such disclosure, “[jjudges must caution a juror at the outset of the colloquy that she must not reveal the way in which any juror plans to vote, or the vote tally on a verdict.” Jenkins, supra, 182 N.J. at 134, 861 A.2d 827.
The Supreme Court has “restrietively interpreted the phrase ‘inability to continue’ in Rule 1:8 — 2(d)(1) to protect a defendant’s right to a fair jury trial, forbidding juror substitution when a deliberating juror’s removal is in any way related to the deliberative process,” Jenkins, supra, 182 N.J. at 124, 861 A.2d 827. “A deliberating juror may not be discharged and replaced with an alternate unless the record ‘adequately establishes] that the juror suffers from an inability to function that is personal and unrelated to the juror’s interaction with the other jury members.’ ” Id. at 124-25, 861 A.2d 827 (quoting Hightower, supra, 146 N.J. at 254, 680 A.2d 649).
In Jenkins, the Supreme Court strongly emphasized the importance of not removing a juror because of inter-juror conflict:
A juror cannot be removed merely because she is taking a position at odds with other juror’s views. A juror has the unassailable right to see the evidence in her own way and to reach her own conclusions, regardless of how overwhelming the evidence or how illogical her view may appear to other jurors, “If a court suspects that the problems with the juror are due to interactions with other jurors, the court should instruct the jury to resume deliberations.”
Id. at 125, 861 A.2d 827 (citations omitted).]
We review what the judge did in this case in light of the controlling law. Here, jurors two and six asked to be excused via a note sent to the court. The judge interviewed juror two who was described as clearly upset and emotional. The judge described the juror’s voice as cracking.
The judge first told juror two she could not discuss the deliberations. This was appropriate, but the interview itself was extremely short consisting of only three questions to which the juror gave conflicting answers. Most importantly, the juror was never asked to explain her reason for being upset. The judge was required to ascertain why the juror was upset and asking to be removed from the jury. He never achieved this goal and made very little effort to get the information he needed. His finding she was upset and emotional is undoubtedly true. However, if her reasons for being upset were personal, she could be removed and replaced, but if she was upset because of how deliberations were going or because of interactions with other jurors during deliberations, she either had to continue or the court was obligated to declare a mistrial. That decision could not be made without information about the cause of her distress.
The colloquy between the court and the juror follows:
THE COURT: Do you feel that there is emotionally an inability for you to proceed and perform your duties as a deliberating juror?
THE JUROR: Yes.
THE COURT: Do you feel that these emotions that you have, again, would impact upon your ability to perform your function in this case?
THE JUROR: No. I know it’s not balanced in what I’m saying, but there’s [sic] reasons why I can’t speak without giving away—
THE COURT: I don’t want you to talk about that. But emotionally, you feel you can’t continue?
THE JUROR: Correct.
THE COURT: I’m going to leave it at that for now. Thank you,
It is difficult to reconcile that limited exchange with the majority’s conclusion that “the trial judge sought the explanation for juror two’s request to be excused.” Ante at 273,173 A.3d 194. The judge appears to have avoided eliciting an explanation that could require a mistrial, at the cost of wrongfully replacing a juror. The juror never said she could not be fair to both sides. In her limited response to question two, the juror denied that was the situation, but was prevented from giving an explanation. She agreed in response to question one and three that she did not feel she could continue with the deliberations but gave no explanation as to why. The majority opinion states:
He [defendant] additionally infers from juror two’s comments she was at odds with other jurors, a circumstance not justifying excusal. See [ [Jenkins, [supra,[ 182 N.J. [at] 124-25 [861 A.2d 827] [ ] (holding excusing a juror cannot be based on juror interaction with other jurors). We cannot agree the juror’s comments revealed she faced hostility from fellow jurors ....
[Ante at 272-73, 173 A.3d 194.]
We do not know, based on this record, if the juror was upset because she was at odds with other jurors since the juror was never given the opportunity to reveal the reason for her distress. Moreover, her assertion that “there’s [sic] reasons why I can’t speak without giving away — ” does reasonably imply that she needs to discuss her emotional state in the context of something relating to jury deliberations, which was the only thing she was instructed not to mention. The trial judge discharged juror two on a record that did not adequately establish her inability to function.
The trial court’s inquiry was insufficient to determine the cause of the juror’s unwillingness to continue deliberations. The juror started a statement that gave the trial court reason to suspect the juror’s problem was due to interactions with other jurors, but he stopped her before she could complete her answer. Even if there is some other way to interpret the juror’s statement, the trial record does not adequately establish the juror suffered from an inability to function personal to her and unrelated to her interaction with other jury members.
Although it is clear the trial court’s effort was designed to preserve the verdict and “avoid the deplorable waste of time, effort, money, and judicial resources inherent in a mistrial,” Hightower, supra, 146 N.J. at 254, 680 A.2d 649, the cause of juror two’s emotional distress could have been attributable to either personal problems or her interaction with other jurors. The record does not adequately establish the cause. For that reason, it is necessary to conclude the trial court misapplied its discretion by replacing juror number two.
IV.
Our Supreme Court has explained the doctrine of cumulative error: “the rule is that where any one of several errors assigned would not in itself be sufficient to warrant a reversal, yet if all of them taken together justify the conclusion that defendant was not accorded a fair trial, it becomes the duty of this court to reverse.” State v. Orecchio, 16 N.J. 125, 134, 106 A.2d 541 (1954) (quoting State v. Dolliver, 150 Minn. 155, 184 N.W. 848, 849 (1921)). Here, the trial court excluded relevant and potentially exculpatory testimony, admitted mostly irrelevant testimony about gangs, and improperly excused a deliberating juror. In my view, each of these errors deprived defendant of a fair trial. Their cumulative effect clearly did so. For the reasons set forth above, I would reverse the convictions in both trials and remand for a new trial on all remaining charges.
. The prosecutor, on one hand, argued for exclusion because Dr. Penrods testimony would not address the facts of the case, while arguing on the other hand his testimony would invade the province of the jury to determine credibility. The judge did not bar testimony based on either of these arguments.
The record does not reflect whether the survivor was looking at defendants face when he spoke because this area of testimony became irrelevant when the experts testimony was barred. It most certainly would have been probed during her examination if the experts testimony was not already barred. Then, depending on her testimony, Dr. Penrod could have been cross-examined on the reliability of the research. If the testimony established the survivor did not see defendants face as he spoke, the State could have moved to bar this area of testimony. Significantly, this was not part of the trial courts reasoning for barring the testimony.
The judge pointed out another psychologist had testified on this factor before the Special Master and not Dr. Penrod.
Dr. Penrod’s testimony before the Special Master and his published work were quoted at length in the decision, which focused on eye-witness identification.
The judge could have simply sent the jury home for the day to allow the juror to calm down or could have asked whether her reasons were related to other juror’s interactions with her, after telling her to limit her initial response to yes or no.