The opinion of the court was delivered by
Defendant, Margaret Kelly Michaels, appeals from her jury conviction on 115 counts of sexual offenses involving twenty children who were in the Wee Care Nursery School (Wee Care). The trial court sentenced defendant to an aggregate term of forty-seven years in prison, with fourteen years of parole ineligibility, and $2,875 in fines payable to the Violent Crimes Compensation Board.
She was originally charged on June 6, 1985, in a six-count indictment involving three boys. A second indictment of 174 counts involving twenty boys and girls was presented on July 30, 1985. A third indictment of 55 counts involving fifteen youths was *586filed on November 21,1985. Those three overlapping indictments contained a total of 235 counts.
Pretrial motions spanned the course of the year prior to trial. In July 1986 defendant moved to dismiss the indictments for lack of specificity. The motion was denied. Defendant also moved for permission: (1) to present child-sex-abuser-profile testimony to demonstrate that it was not likely that she had committed the alleged acts; (2) to have the child-victims psychologically examined; and (3) to dismiss the indictments for improper suggestibility by interviewers of the children during the investigation phase. Those motions were denied.
Defendant also moved to admit the results of the State’s polygraph examination of her taken on May 6, 1985 into evidence. She had not been advised that she could stipulate to use of the results and thus no stipulation was entered' into. The State’s polygrapher who administered the examination concluded that defendant had passed; however, the State later asked its chief polygrapher to review the results. After consulting with another polygrapher, he concluded that the results were “inconclusive.” The trial court denied defense counsel’s motion on the ground that an unstipulated polygraph is inadmissible. Defendant has not appealed this ruling. We note the fact of the early administration of the polygraph for historical purposes and for its bearing, if any, on the manner and direction of the State’s investigation into the allegations of child abuse at Wee Care.
The State filed pretrial motions to permit the children to testify by closed-circuit television (CCTV) pursuant to N.J.S.A. 2A.-84A-32.4, and to present expert testimony on the Child Sex Abuse Syndrome (CSAS). The trial court granted the State’s motion to present CSAS testimony and ultimately permitted all of the children to testify on CCTV.
The jury trial began with opening statements on June 22, 1987. Nine months later the case went to the jury for deliberation. At the time the case went to the jury, 131 counts spanning the three indictments remained, charging essentially four types of offenses: *587aggravated sexual assault (N.J.S.A. 2C:14-2a(l)), sexual assault (N.J.S.A. 2C:14-2b), endangering the welfare of children (N.J.S.A. 2C:24-4), and making terroristic threats (N.J.S.A. 2C:12-3). After twelve days of deliberation, which included a replaying of each child’s CCTV testimony at the jury’s request, the jury returned with guilty verdicts on 115 counts of aggravated assault (thirty-eight counts), sexual assault (thirty-one counts), endangering the welfare of children (forty-four counts), and terroristic threats (two counts).
The trial court denied defendant’s motion for bail pending sentencing. On April 18, 1988, we granted defendant’s motion for bail pending sentencing, and at the same time granted the State’s motion for a stay pending the State’s application to the Supreme Court for review. On April 26,1988, the Supreme Court reversed our order granting defendant bail pending sentencing.
The trial judge denied defendant’s motion for a new trial and sentenced defendant on August 2, 1988. The judge denied defendant’s motion for bail pending appeal, and we affirmed his denial in light of our Supreme Court’s eai’lier ruling. Defendant has been incarcerated from April 1988 to the present time.
On appeal, defendant raises the following legal arguments:
POINT I: THE TRIAL COURT’S REFUSAL TO ALLOW DEFENSE EXPERTS TO EXAMINE ANY OF THE CHILD-COMPLAINANTS IN ORDER TO PLACE THE DEFENSE ON EQUAL FOOTING WITH THE PROSECUTION DEPRIVED MS. MICHAELS OF AN ADVERSARIAL TRIAL—THE ESSENCE OF DUE PROCESS.
POINT II: THE USE OF CCTV TESTIMONY WAS IMPROPER BECAUSE THE TRIAL COURT DID NOT FOLLOW THE CCTV STATUTE, AND BECAUSE THAT TESTIMONIAL DEVICE, AS USED IN THIS CASE, EVISCERATED THE PRESUMPTION OF INNOCENCE.
POINT III: THE TRIAL COURT ERRED IN ADMITTING TREACY’S TESTIMONY BECAUSE HER METHODOLOGY IS WHOLLY UNRELIABLE AND SCIENTIFICALLY REPUGNANT.
POINT IV: BY BOLSTERING THE CREDIBILITY OF THE CHILD-WITNESSES, TREACY INVADED THE PROVINCE OF THE JURY.
POINT V: THE TRIAL COURT’S REFUSAL TO PERMIT DEFENSE EXPERT TESTIMONY ON THE LACK OF ANY INDICATIONS OF DEVIANCY *588OR PATHOLOGY VIOLATED MS. MICHAELS’ RIGHT TO PRESENT A DEFENSE TO SUPPORT HER CLAIM OF INNOCENCE.
POINT VI: REVERSAL IS REQUIRED BECAUSE THE QUESTIONING OF WEE CARE CHILDREN WAS SO SUGGESTIVE AND COERCIVE THAT THEY WERE RENDERED INCOMPETENT TO TESTIFY.
POINT VII: THE PROSECUTOR’S INVOCATION OF HITLER AND ITS AD HOMINEM ATTACK ON DEFENSE EXPERTS INEXCUSABLY POISONED THE TRIAL.
POINT VIII: THE TRIAL COURT IMPERMISSIBLY ALLOWED THE JURY TO VIEW THE VIDEOTAPES OF THE CHILDREN’S CCTV TESTIMONY DURING THEIR DELIBERATIONS.
POINT IX: THE ADMISSION OF MASSIVE HEARSAY TESTIMONY SO THOROUGHLY POLLUTED THE TRIAL THAT NO REMEDY OTHER THAN A MISTRIAL WAS APPROPRIATE.
On January 6, 1993, our State Supreme Court handed down its opinion in State v. J.Q., 130 N.J. 554, 617 A.2d 1196 (1993). In the present case, as in J.Q., the opinion of the State’s expert went beyond the limited permissible scope of syndrome testimony. Unquestionably, this erroneously admitted evidence was capable of producing an unjust result and thus requires the reversal of defendant’s convictions. R. 2:10-2; J.Q., supra, 130 N.J. at 574-76, 582, 617 A.2d 1196. We find it necessary nonetheless to expand on this and other issues raised on appeal.
FACTUAL BACKGROUND
In 1984-85, Wee Care, located in St. George’s Episcopal Church in Maplewood, provided services for approximately fifty families with children between the ages of three and six. The nursery school was usually staffed by five teachers and two assistants or aides. The school had use of three levels of the church, although the primary rooms it used were located in the basement.
Located in the basement was a block room or play room. Adjacent to the block room was the work and language room, which doubled as a lunch room. An art room was connected to the block room, separated by a vinyl divider. A large closet was used for art supplies. No bathrooms were in the basement of the school. The church secretary’s office and the pastor’s office were on the first floor above the basement.
*589Mens’ and ladies’ rooms were located on the first-floor level, and the children at Wee Care used that ladies’ room. The gym, a large room with a stage on one end, was on the same level. Behind the stage was a storage room with a doorway that opened into the hallway. The gym also had a doorway which led outside. At the opposite end of the gym was a large kitchen with other small rooms. Another bathroom was located at the end of a hallway. A set of stairs led to the second level.
The church sanctuary was on the second floor. The nursery school director’s office, a room characterized as a living room, and a few other rooms were on one end of this floor. On the other end was a bathroom, Kindergarten room, and the choir room. The choir room, also referred to as the music room and piano room, was adjacent to the Kindergarten room. The hallway door to the Kindergarten room had a glass insert. Pianos were located in the gym, the Kindergarten room, and the choir room. The jury had the opportunity during trial to visit the premises.
At the end of the summer of 1984, Wee Care advertised for teachers in a local paper. Defendant had moved to New Jersey in September 1984 from Pittsburgh to become involved in New York theater work. She was only a few credits short of a bachelor’s degree in theater arts. She moved in with a friend and looked for a temporary job. Defendant answered the Wee Care advertisement. She was interviewed at the end of September for a teacher’s aide position. Defendant, who had no teaching experience, explained that she came from a large family, liked little children, and would like to try teaching. She had musical abilities and could play the piano. Wee Care hired defendant as a teacher’s aide for a one-week probationary period.
After three weeks, Wee Care decided that she would make an excellent teacher for the three-year-old class. As a teacher her day started at 8:30 a.m. and ended at 5:00 p.m. Her teaching schedule was active. Between 8:30 and 9:00, teachers prepared for the upcoming day. Between 9:00 and 9:15 the teachers gathered the children from the playground or the gym to go to *590classes. First period (9:15 to 10:00) consisted of math, language, art or workshop. At about 9:35 or 9:40, there would be a bathroom call by an aide. A mid-morning snack was served between 10:15 and 10:45. The second class period went from 10:45 to 11:45.
The children would wash up for lunch between 11:45 and 11:55. Defendant would set up the afternoon nap mats between 11:55 and 12:20, and performed her lunch room duties from 12:20 to 12:55. Although lunch was split into two periods, everyone had the same nap time, which was between 1:00 and 2:30. Defendant was scheduled to monitor nap time in the block room between 1:00 and 1:45. She usually ate her own lunch between 1:45 and 2:30, after being relieved from her nap duties. Between 2:30 and 3:30 the children were organized for the afternoon snack, which would take place about 3:30. Between 4:00 and 5:00 defendant was usually in charge of her own group of children. Some children began leaving Wee Care as early as 1:00, while others left after nap time intermittently until approximately 6:00.
By letter dated April 15,1985, defendant gave two weeks’ notice that she was leaving Wee Care. She verbally told the school director that she had to leave for personal reasons and was returning home. Defendant also wrote a letter of explanation to the parents of her students. Defendant had in fact taken a job at another day care facility.
Of the nineteen children who testified at trial, only five were actually assigned to defendant’s class. The facts surrounding the alleged sexual abuse came from two sources: the children testified in the judge’s chambers via CCTV, and the children’s parents and grandparents testified regarding what the children had told them after defendant left Wee Care. A good deal of the parental testimony was devoted to the behavioral changes that they had observed in their children. The recollections of the children’s behavior came at the behest of the State’s expert, Eileen Treacy, who provided the parents with a checklist of behavioral changes to consider as they recalled their children’s actions at the time of the *591alleged abuse. Treacy did not become involved in this case until late October 1986. Consequently, the parents were being asked to reconstruct events a substantial time after they had actually observed the events.
A child-specific recitation of the alleged abusive acts would serve no useful purpose at this juncture. No complaints of abuse were made during defendant’s tenure at Wee Care. However, on April 80, 1985, a Wee Care child visited his pediatrician. While the nurse was taking his temperature rectally, the child commented that his teacher did the same thing to him. When the nurse asked the child what teacher, he responded “Kelly”'—the name the children knew defendant by. Although the pediatrician found no evidence of abuse, this comment by the child started the investigation at Wee Care. A defense expert opined at trial that it was patently obvious that the child’s comment was misunderstood and that the child was referring to the rubbing of his back and not the anal penetration.
That child, six-and-a-half years old at trial, started Wee Care when he was almost four. He remembered that Joan was his teacher but that he had Kelly, not Joan, at nap time. He testified that he hated nap time because Kelly had once taken his temperature and he did not want her to. She had put “gasoline” (vaseline) on the thermometer first. Kelly put the thermometer in his “bum,” and she said nothing when he told her not to do it. He stated that she also took the temperatures of two other children, and he saw her pull their pants down. Neither of those two children indicated that their temperatures were taken at school. Their assertions of abuse were far more egregious, as were those of many other children after they had been “interviewed” by the authorities.
However, during the time defendant was teaching at the school, no children had ever complained of experiencing any difficulties with her. Defendant’s co-employees observed no inappropriate behavior by defendant during her employment at Wee Care. Nor did any of the Wee Care employees notice that any of the children *592exhibited any fear or reluctance to be with defendant. Defendant denied ever doing anything improper with the children. She pointed out that she had little contact with many of the children, and that there were always people who arrived unannounced.
After the parent notified the authorities that the child alleged his temperature had been taken by Kelly, the authorities commenced an investigation that initially encompassed only a few children. Eventually, however, on the recommendation of a Division of Youth and Family Services (DYFS) investigator, the interviewing of children was expanded and intensified to include virtually all of the children Kelly could have had contact with.
The accounts of sexual abuse obtained through interviews of the children ranged from relatively minor accounts of touching to virtually incomprehensible heinous and bizarre acts. A common act alleged by both boy and girl students was that Kelly inserted knives, forks, and spoons into their “butts,” penises, or vaginas. One girl stated that Kelly inserted a light bulb in her vagina, and a boy claimed Legos were inserted in his “tushie.” The children told of games where both they and Kelly took off their clothes and, according to varying accounts, laid on each other, licked each other and Kelly, including applying and licking off peanut butter and/or jelly, had “intercourse” with Kelly while she apparently was having her menstrual period, defecated on the floor, ate “pee and poop,” and performed cunnilingus on her.
Kelly allegedly committed fellatio on some of the boys. Kelly was said to have played “Jingle Bells” on the piano during many of those games. The acts were said to have taken place in the music or choir room, the gym, lunch room, nap room, and bathroom. Kelly was said to have “pooped and peed” on or in a piano bench, on the floor, on the lunch table, and made a cake out of poop that the children had to taste. She was also said to have taken off her clothes in the lunch room in the presence of both children and adults. Testing at the Federal Bureau of Investigation laboratories of a wooden spoon and piano benches for inculpatory evidence proved negative.
*593Several of the children claimed to have told their parents of Kelly’s activities while they were happening, and some children claimed that Wee Care personnel were present or had been told of the occurrences. No adults corroborated the children’s contemporaneous complaints. Many of the children asserted that Kelly threatened to harm their parents if they told of the activities.
I.
The first issue we discuss is whether the expert testimony of Eileen Treacy as presented on the State’s case contravened the bounds of permissible latitude pertaining to Child Sexual Abuse Accommodation Syndrome (CSAAS) evidence as enunciated in J.Q., supra, 130 N.J. at 581-84, 617 A.2d 1196. In this case the syndrome testimony was referred to as Child Sexual Abuse Syndrome (CSAS); however, as we explain later, there is no discernible difference for our purposes. In addition, Treacy presented a confounding variable analysis and gave testimony concerning thirty-two child behavioral indicators of abuse, even though a reliable foundation as to their acceptance within the relevant scientific community was not established. See id. at 583-84, 617 A.2d 1196.
Our Supreme Court noted in J.Q. that “we do not intend in any sense to mystify the trial of child-sexual-abuse cases. All that is required is close attention to existing precedent.” Id. at 583, 617 A.2d 1196. The Court accepted the thesis that child-abuse expert testimony may be used in the courtroom as a “rehabilitative tool,” to help explain why many sexually-abused children delay reporting incidents of abuse and why many recant their allegations of abuse and/or deny that it has occurred. Id. at 579, 617 A.2d 1196. In doing so, it recognized two forms of child-abuse behavioral science expert testimony—1) testimony that describes behaviors commonly observed in sexually-abused children and offered on rebuttal to rehabilitate assertions of lack of credibility raised by delayed reports and recantations, and 2) testimony based on clinical observations of a professional as to the “patterns, effects, and dynamics of child sexual abuse” offered to establish that a particular child *594was sexually abused. Id. at 563-64, 617 A.2d 1196. The latter, characterized as CSAAS or CSAS syndrome evidence, goes beyond the proper function of child-abuse expert behavioral science testimony. Id. at 578-79, 617 A.2d 1196. Thus, the Court adopted the conclusion that the proper use of child-abuse expert testimony is as a rehabilitative tool and not as a diagnostic investigative device, as “[tjhe syndrome does not detect sexual abuse.” Id. at 579, 617 A.2d 1196 (quoting John E.B. Myers et al., Expert Testimony in Child Sexual Abuse Litigation, 68 Neb.L.Rev. 1, 67-68 (1989)).
Defense counsel offered no objection to Treacy’s qualifications. She testified as an expert in child psychology and child-sexual-abuse treatment. Although the witness had neither a doctorate degree nor licenses to practice psychology in New York or New Jersey (New York apparently requires a doctorate degree for a license), she had a Masters degree in psychology. She also had done extensive post-graduate work and had clinical experience with child-sexual-abuse victims. See Evid.R. 56(2).
The trial judge, out of the jury’s presence, heard not only from Treacy, but received defense testimony offered to rebut her assertions that a recognizable pattern of behavior existed in child-sexual-abuse cases, conforming to a “theoretical model” that she adopted from the work of Dr. Suzanne Sgroi. Using that model, Treacy explained that there are five phases of behaviors seen in children who have been sexually abused: (1) the engagement phase, usually an affectionate but non-sexual phase, where the offender gains access or seizes an opportunity to get the child involved in a relationship; (2) the sexual interaction phase, wherein sexual activity begins; (3) the secrecy phase, wherein the perpetrator seeks to keep the child quiet by means of bribes, tricks, rewards or threats; (4) the disclosure phase, which causes a crisis for the child and the child’s family; and (5) the suppression phase, wherein the child develops coping mechanisms to deal with the abuse, such as denial, avoidance, minimization, rationalization, rescue fantasy, or disassociation.
*595Treacy explained that thirty-two behavioral indicators were potentially associated with this syndrome. In an abused child, one would expect to see a clustering of between five and fifteen behavioral indicators or symptoms. The number and type of symptoms that a given child exhibits will depend on the child’s prior coping mechanisms and the support systems in the child’s life. Treacy next listed some of the symptoms or behavioral indicators of child sexual abuse: (1) eating disorders; (2) sleep problems; (3) regression; (4) sexual symptomatology; and (5) development of fears.
Treacy admitted that some of the behavioral symptoms could be caused by factors other than sexual abuse which might be operating in the child’s life such as marital discord, death in the family, illness, and the like. Treacy referred to those kinds of factors as “confounding variables.” She acknowledged that the Diagnostic and Statistical Manual of Mental Disorders (DSM3), a reference book used by mental health professionals, does not recognize the child-sexual-abuse syndrome or any other syndromes.
According to Treacy, a finding that child-sexual-abuse syndrome exists in a child is based on four elements: (1) the child’s statement that he or she was abused as well as the child’s accompanying affect or demeanor; (2) the presence of at least four of the five phases described in the “theoretical model;” (3) the clustering of five to fifteen of the behavioral indicators; and (4) an analysis of the confounding variables and their impact on both the extent of the child’s prior sexual knowledge as well as other stress factors in the child’s life which could account for some of the symptomatology-
Treacy asserted that this pattern of behavior is recognized in clinical circles. She explained that Dr. Sgroi’s model, called the child-sexual-abuse syndrome (CSAS), overlaps to a degree with the child-sexual-abuse-accommodation syndrome (CSAAS) developed by Dr. Roland Summit. See J.Q., supra, 130 N.J. at 566-71, 617 A 2d 1196. The difference between the two models appears to be in terminology and sequence of the phases. Dr. Summit *596identifies a secrecy phase and a retraction or recantation phase which Dr. Sgroi calls suppression. Dr. Summit refers to the entrapment of the child which corresponds to Dr. Sgroi’s engagement phase. The word “accommodation” in Dr. Summit’s model refers to the phase of entrapment and accommodation that he found children experience during sexual abuse.
Treaey admitted on cross-examination that nearly all of the items on her checklist of behavioral indicators were common to all types of traumatic stress. She insisted, however, that what is crucial is the clustering of those symptoms, accompanied by the child’s report, four of the five Sgroi phases, and other stress factors. Treaey testified she would not focus solely on the behavioral symptoms to reach a conclusion about a specific child.
The defense offered the testimony of Dr. Jonas Rappeport and Dr. Ralph Underwager to support its position that syndrome testimony should be inadmissible at trial. Dr. Rappeport was board certified in psychiatry and licensed in Maryland and California. He had taught as a clinical professor of psychiatry at the University of Maryland School of Medicine and as an adjunct professor at the University of Maryland School of Law. He was not a child psychiatrist, however, and had never treated children under the age of seven who were sexually abused.
It was his clinical judgment that Treacy’s checklist of behavioral indicators, which supposedly indicated a child was suffering from some type of stress reaction, was misleading. He found only a few areas that might possibly indicate the existence of sexual activity in the child’s life. In his opinion, Dr. Sgroi’s model was not intended for diagnostic purposes. It would not be appropriate for a professional to conclude that a child was sexually abused based upon observations of the five phases in the alleged victim. Dr. Rappeport acknowledged having read about the existence of the child-sexual-abuse syndrome and admitted that, as a syndrome, it has been accepted by the general psychiatric or psychological community.
*597Dr. Underwager also testified for the defense. He served as the director of the Institute for Psychological Therapies. Dr. Underwager had a Ph.D. in clinical psychology, and estimated that he had treated over 500 child-sexual-abuse victims. Dr. Under-wager asserted that none of Treacy’s thirty-two behavioral indicators could be attributed exclusively to the sexual abuse of a child. Turning to Dr. Sgroi’s model, he asserted that models are developed to provide the researcher with suggestions as to what observable behavior will either validate or invalidate the model. A model functions to generate a hypothesis which is then tested. Dr. Sgroi’s model was but one of many which related to the child-sexual-abuse syndrome. He asserted, however, that at the time of trial, none of those models had been tested and none was generally accepted within the scientific community. Each of these models was only related to a broad range of other disorders. Dr. Underwager criticized Treacy’s thirty-two item checklist as having no source other than herself and thus having no validity or reliability.
Treacy was recalled after Dr. Underwager’s testimony. She claimed that the checklist which she relied on was approved by the National Institute of Mental Health to be used in conjunction with a clinical interview. She acknowledged that Dr. Sgroi’s model was not designed for diagnostic purposes, but rather was to be used to create a theoretical and conceptual framework for understanding what a sexually abused child experiences. The model was therefore to be used for clinical treatment and investigation but not diagnosis.
Treacy, presumably in accordance with the prosecutor’s instructions, did not refer to the syndrome per se in her testimony before the jury. She did, however, testify that the behavior which the children exhibited during the alleged period of abuse and after disclosure was “consistent with” that of a child who had been sexually abused. Despite the defense’s objection that the “consistent with” testimony was really vouching for the credibility of the alleged victims, the trial judge admitted the testimony in the *598asserted belief that any purported shortcomings in the validity of the testimony could be presented to the jury through cross-examination of the expert.
Treacy testified in the presence of the jury first in July 1987 and then again in December 1987. Her initial testimony was not child-specific, but rather focused on the stages of intellectual and cognitive development of children at the ages relevant to the children involved in the present case. She discussed the memory capability of young children, their developing sense of good and bad, their ability to distinguish reality from fantasy, and their sexual development patterns—including a child’s defense and coping mechanisms.
In December Treacy testified over a period of five and one-half days. She testified as to her behavioral indicators and the phases of CSAS. She then gave her opinion as to whether the testimony and conduct of the children were consistent with child sexual abuse based on her analysis of the pretrial interviews of the children, the trial testimony of each child, the behavior attributed to the child by the relatives, and the affect and demeanor of the child. She gave an affirmative conclusion with respect to all of the children except one. She was unable to reach a firm conclusion as to this child because of the confounding variables in the child’s life. Even in that case, she did opine that the child’s behavior was consistent with the child’s statement of abuse. However, a qualification was included because the child may have been sexually abused in other circumstances.
It is clear that the jury could have only understood Treacy’s “consistent with” opinion testimony to mean consistent with the child abuse alleged by the State. This is particularly true when' examining her testimony regarding the behavioral indicators and the confounding variables.- Her testimony constituted nothing less than substantive evidence of defendant’s guilt, albeit Treaey’s opinion thereof. This constituted error because syndrome evidence is not probative of sexual abuse. The syndrome assumes the presence of sexual abuse and only seeks to explain the child’s *599reaction to it. State v. Dodson, 452 N.W.2d 610, 612 (Iowa Ct.App.1989). The jury was not instructed that use of the testimony was to be limited to rehabilitation. Indeed, it was not offered merely for rehabilitative purposes. Ibid.; Myers, supra, 68 Neb.L.Rev. at 66-67.
We are convinced that the foundation evidence presented was insufficient to permit the use of the syndrome testimony as substantive evidence of abuse. As we have previously said, child-abuse expert evidence is admissible only for the purpose of rehabilitation—explaining traits often found in children who had been abused. See J.Q., supra, 130 N.J. at 579-80, 617 A.2d 1196. When such testimony is used, the expert would be expected to testify that specific behavior of the victim, while appearing to be inconsistent, may in fact be consistent with sexual abuse. The expert should not be asked to give an opinion about whether a particular child was abused. Ibid. Therefore, care should be taken to avoid giving the jury an impression that the expert believes based on CSAAS or CSAS a particular child has been abused.
Syndrome testimony has properly been judicially circumscribed. Scientific opinion in the field clearly supports the cautiousness of the courts. Four serious problems militate against its use as substantive evidence of guilt. First, a child may perceive that he or she has been abused without actually having been abused as defined by law. Second, although the child may in fact have been abused at some other time, this does not prove that defendant abused the child on the particular date(s) in question. Third, syndromes are not based on hard data but rather on clinical intuition. As such, syndromes lack a firm scientific foundation. Fourth, the symptoms suffered by abused children can result from a variety of traumatic events of which sexual abuse is but one event. Debra Whitcomb, When The Victim Is A Child, National Institute of Justice 116 (2d ed. 1992).
“Unlike some other syndromes, such as battered women’s syndrome or rape victim’s syndrome, the child sexual abuse accommo*600dation syndrome was not created as a diagnostic tool, and children who display signs of the syndrome may not have been abused.” Jill Birnbaum, Expert Testimony in Custody and Visitation Cases Involving Child Sexual Abuse, National Center on Women and Family Law, Inc. 699 (Nov. 1990). Another writer has instructed that psychologists should not be asked to tell a court whether or not a particular child has been abused. Psychologists are not capable of making that determination, nor can they make the determination of whether a particular offender is guilty or innocent. Marian D. Hall, The Role of Psychologists as Experts in Cases Involving Allegations of Child Sexual Abuse, 23 Family Law Quarterly 451, 453 (1989). Hall further explains the problem facing the scientific community as follows:
It would appear that the prospect of designing checklists, inventories, and rating-scales to provide objective measures of abusive behavior, its antecedents, correlates, and consequences, holds promise of yielding information that may be useful both in individual and epidemiological data gathering. Designing and validating such measures, however, depends on theory and research that is currently the focus of much study and controversy. Nowhere is that more evident than in the scientific and legal arguments about whether behaviors exist that are unique to sexually abused children and whether such behaviors fall into patterns that suggest a typical “profile” or “syndrome” for the child sexual abuse victim.
Children’s reactions to sexual abuse vary dramatically and, to date, the methodological problems.involved in compiling results of the scores of diverse studies have led only to lists of very general symptoms, which occur to some extent in all children, and are especially prevalent in children who suffer from various forms of emotional trauma, separation, or loss of security. [7d at 462-63.]
The position of our New Jersey courts on the use of syndrome testimony is echoed by our neighboring New York courts. In People v. Knupp, 179 A.D.2d 1030, 579 N.Y.S.2d 801 (1992), the State’s witness first testified about the symptoms associated with the child-sexual-abuse syndrome. The expert then stated that defendant’s three daughters suffered from those symptoms. “Obviously, the prosecution was attempting to raise the inference that, because the children exhibited those symptoms, they had been sexually abused.” Id., 579 N.Y.S.2d at 802. The court found this to be improper. Ibid.
*601The Supreme Court of Pennsylvania in Commonwealth v. Dunkle, 529 Pa. 168, 602 A.2d 830 (1992), because of the flaws inherent in syndrome evidence, decided that it should not be admissible for any purpose. The State’s witness had been permitted to explain why a victim would delay reporting an offense, why a victim would be unable to recall exact dates and times of the alleged offense, and why a victim would omit details of an incident when he or she first told his or her story. The expert did not relate any of her testimony to the child victims in the case. Id., 602 A.2d at 831.
The Pennsylvania Supreme Court first addressed the admissibility of the child-sexual-abuse-syndrome testimony. The court surveyed the literature explaining the nature of the syndrome and noted the uniformity of opinion that “there is no one classical or typical personality profile for abused children.” Id. at 832 (footnote omitted). The court noted that the child-sexual-abuse syndrome construct is an attempt to create a diagnostic or behavioral profile about sexually abused children. However, it concluded that “[t]he existence of such a syndrome as either a generally accepted diagnostic tool or as relevant evidence is not supportable.” Ibid. The court observed that the behavior patterns of sexually abused children are not necessarily unique to sexually abused children. Rather, the behavior patterns may be similar or even identical to reactions of children who are not abused, whose parents divorce, or who are psychologically abused. Id. at 832-33. The court made further observations about the case before it, which apply equally to the manner of the State’s prosecution of the case against defendant.
The damage created by this testimony was also compounded by the testimony about those who knew the child in question. There was testimony admitted about the behaviors exhibited by the child after the alleged incident. As such, the prosecutor’s introduction of the testimony by those who observed the child served to confirm certain behavior patterns that the expert suggested were exhibited by abused children. Permitting an expert to testify about an unsupportable behavioral profile and then introducing testimony to show that the witness acted in conformance with such a profile is an erroneous method of obtaining a conviction. For this reason, we hold that the expert should not have been permitted to testify about behavior patterns generally exhibited by abused children and that the error requires reversal. [ Id. at 836.]
*602The Pennsylvania court then parted company with our New Jersey courts by rejecting the admissibility of testimony explaining why a sexually abused child would delay reporting the incident to family members, why such a child would omit details of the incident, and why the child would have an inability to recall dates or times of the incident. Id. at 837-38.
The court asserted that it is commonly understood why sexually abused children do not always come forward immediately after the abuse—they are afraid, embarrassed, threatened by the perpetrator, and possibly not knowledgeable enough to tell someone what happened. Id. at 836. The same reasons explain why an abused child may sometimes omit traumatic details of the incident. The court found no need for expert testimony because the understanding of these factors is well within the common knowledge of jurors. Id. at 837-38. As for a child’s ability to recollect dates or events, a jury’s function is to weigh the child’s credibility, and expert testimony would simply infringe upon its ability to do so. Id. at 838. The Pennsylvania Supreme Court thereafter reaffirmed this holding in Commonwealth v. Sees, 529 Pa. 450, 605 A.2d 307, 308-09 (1992).
The Supreme Court of Kentucky in Hellstrom v. Commonwealth, 825 S.W.2d 612 (Ky.1992), has also carefully limited the admissibility of CSAS evidence. The State presented testimony from the director of the child abuse center at the University of Kentucky Medical Center. He possessed a Masters degree in clinical social work. The expert testified about the victim’s symptoms (bad dreams, anxiety, anger, distrust of men, stomach complaints, nervous symptoms) and explained that delayed disclosure is common in a child who has been victimized by a member of her family. When asked whether he found these reasons for the delayed disclosure to exist in this victim’s case, he responded affirmatively. Id. at 613.
The Kentucky court concluded it did not matter that the witness never labeled any of the symptoms as belonging to CSAS. Citing a prior decision which reversed a conviction after a social worker *603testified that the victim’s behavior was consistent with abuse, the court noted that the prosecution’s witness
listed the symptoms but refrained from classifying them directly as the “child sexual abuse syndrome.” Avoiding the term “syndrome” does not transform inadmissible hearsay into reliable scientific evidence. Neither the syndrome nor the symptoms that comprise the syndrome have recognized reliability in diagnosing child sexual abuse as a scientific entity. [Id. at 614.]
Here, although Treacy did not explicitly state that she believed the children were abused, she said that the children’s behavioral indicators (bed wetting, excessive masturbation, unfounded fears, etc.) were “consistent with” child abuse. She defined consistent with as having a “high degree of correlation,” “over point six [.6]” in numerical terms of probability. She testified that the children exhibited several of the phases of Sgroi’s model (some exhibited “four out of five” phases). Treacy’s testimony was not focused on explaining why these children would delay reporting the abusive incidents, or why they would recant and then reassert claims that defendant abused them. There was very little testimony in this rehabilitative vein.
Clearly, her purpose in testifying was to tell the jury that the laundry lists of “abnormal” behaviors testified to by the children’s parents or care givers was “consistent with” the behavioral patterns of sexually abused children. Treacy’s use of her confounding variables to filter out any other cause for the children's behavioral symptoms was intended to support the conclusion the jury was to imply—the children were sexually abused. The State’s case was structured so that Treacy did not have to directly say that these children were sexually abused or that she believed that they were abused. Ruling out any other possible cause for the children’s unusual behavior, as reconstructed by the parents at her direction, was sufficient.
The parents testified at length and in minute detail about their children’s behavioral changes. The jurors were exposed for weeks to understandably concerned and anxious parents describing the regression of their children, the terrors these children experienced, and the personality changes which they had now come to *604relate to their children’s exposure to defendant. In at least one instance, the destruction of the parents’ marriage was attributed to the allegation of child abuse.
Treacy, in the precise manner depicted by the Pennsylvania Supreme Court, provided a seemingly scientific analysis demonstrating what sexually abused children go through in their relationship with the abuser and the type of behavioral, transformations which occur after the children experience sexual abuse. Treacy had asked these parents to consider a checklist of thirty-two behaviors which she provided for them. The parents indicated on the checklist the many behaviors their children exhibited. These behaviors were the same ones she described for the jury as being manifested by sexually abused children. Treacy opined that she would expect to find five to fifteen of the thirty-two behaviors in a child who had been traumatized by sexual abuse. Under the prosecutor’s direction, “the expert” then referred specifically to each child and asserted that the child had demonstrated a certain number of the phases of Dr. Sgroi’s model and that the behavioral indicators were “consistent with” those of an abused child.
In so doing, Treacy recited the confounding variables, if any, that she found present in the child’s life and asserted that the variables could not have been the cause of the child’s behavioral symptoms. Treacy made this finding for all but one of the children. Treacy thereby validated the children’s reports of sexual abuse to the jury by demonstrating an alleged scientific process of determining whether the children were actually sexually abused. She was permitted to lead the jury to believe that the process was rooted in science and thus was a reliable means of determining sexual abuse.
Dr. Sgroi, who formulated the model upon which Treacy based her conclusions, points out that numerous variables are involved in reaching the end of the validation process and that the process lacks scientific certainty.
Validation of child sexual abuse depends upon recognizing behavioral indicators, the capacity to perform investigative interviewing, the ability to do credibility *605assessment, recognizing physical indicators, and the capacity to perform comprehensive medical examinations. Behavioral indicators of child sexual abuse may be helpful but are rarely conclusive. [Suzanne M. Sgroi, Handbook Of Clinical Intervention In Child Sexual Abuse 78 (1982).]
Contrary to the belief of the trial judge, no amount of cross-examination could have undone the harm caused by Treacy’s purported validations. Defendant’s convictions were obtained by use of expert testimony which should have been excluded. The impact of the error was so overwhelming as to require reversal.
II.
Defendant contends that the trial court erred in denying her motion to have the children examined by her own psychiatrist or psychologist. On March 26, 1986, defendant moved for psychiatric examinations of the children based on our holding in State v. R.W., 200 N.J.Super. 560, 568-69, 491 A.2d 1304 (App.Div.1985), that one accused of sexual assault of a child should, because of “fundamental fairness,” be allowed to develop evidence with respect to the child’s testimonial capacity. No affidavit or brief supported this motion as our decision did not require a showing of special circumstances to warrant the examination. On September 23, 1986, the Supreme Court modified this ruling in State v. R.W., 104 N.J. 14, 514 A.2d 1287 (1986). The State therefore sought to quash the defendant’s motion in accordance with the Supreme Court’s decision. In October 1986, the trial court denied defendant’s motion.
On April 1, 1987, defendant filed a new motion, asserting the right to examine the children under the “substantial need” test articulated in R.W. On April 24, defendant filed an additional motion asserting that because the State’s expert, Eileen Treacy, had examined the children, defendant deserved an equivalent right to do so. Defendant’s brief attacking the constitutionality of the CCTV statute also asserted a need for a psychiatric examination of the children. On May 28, 1987, the court held that defendant was not entitled to examine the children either to rebut the State’s *606expert psychological testimony or to respond to the State’s motion to have the children testify via CCTV.
Defendant argues that without the opportunity to have the children examined by her own expert, she was unable to counter the testimony that the children were traumatized. Defendant points out that the State’s expert testimony was used to: (1) demonstrate that the children were abused; (2) justify suggestive and coercive investigative questioning; (3) explain the failure to report the abuse; and (4) justify the use of CCTV. Defendant also maintains that the State capitalized on its exclusive access to the children by emphasizing that Treacy had interviewed the children and assessed their credibility.
The State responds that most of the psychologists testifying at trial were “fact” witnesses who described their therapy sessions with various children. It insists that it was defendant who elicited on cross-examination the testimony about which she now complains—e.g., Treaey’s statements concerning her personal observations of the children and Dr. Susan Esquilin’s perceptions of their degree of fear and trauma. The State also maintains that defendant was afforded extensive discovery concerning the children which provided her with sufficient information to challenge the State’s expert psychologists if there were a basis to do so.
Treacy became involved in the case in October 1986, after the children had already been seen by the prosecutor, DYFS investigators, police investigators, and the children’s therapists. She interviewed all but four of the prospective child-witnesses. The defense was supplied with the reports Treacy prepared from the interviews and, in many cases, audiotapes of the interview. During argument on defendant’s motion to compel an examination of the children, the State represented that it would not elicit before the jury that Treacy had seen the children, and would not have any expert testify that he or she diagnosed the children as having CSAS.
Dr. Esquilin, as an expert witness, also testified in detail as to her group meetings in the spring and summer of 1985 with Wee *607Care students and parents. These meetings occurred during the early stages of the investigation. She and several other psychologists testified concerning their individual therapy sessions with eleven of the twenty children alleged to be victims. The State represented pretrial that “the therapists will be used in this case merely for the purposes of hearsay and will not be giving an opinion [on whether the children had been sexually abused].” The State adhered to this agreement. The defense was given the therapists’ reports containing their descriptions of the statements the children had made in treatment and the therapists’ personal observations. The therapists also testified as to the children’s ability to testify in open court. They prepared separate reports on this issue which were given to the defense.
In R.W., our Supreme Court considered a defense request for a psychiatric examination of a three and one-half-year-old child to explore whether the child had the competency to be a witness under Evid.R. 17 and/or “would in all psychiatric or psychological probability have sensory and/or mental defects that would be relevant in assessment of that witness’ credibility.” R.W., supra, 104 N.J. at 17, 514 A.2d 1287. The Court held that the age of a child alone would not warrant deviating from the requirement that there must be an extraordinary reason to order a psychiatric examination to determine witness competency. Id. at 22, 514 A.2d 1287. Because defendant had presented no evidence of a “mental or emotional disorder, aberrational behavior, inappropriate conduct, bizarre attitudes or other kinds of peculiarity or abnormality” on the part of the child, the Court ruled that defendant had not demonstrated a substantial need and justification for a psychiatric examination. Id. at 27, 514 A.2d 1287.
Our Supreme Court emphasized that a trial court “must balance the possible emotional trauma, embarrassment, and intimidation to the complainant, particularly an extremely young child, against the likelihood that the examination will produce material, as distinguished from speculative, evidence.” Id. at 28, 514 A.2d 1287 (footnote omitted). The Court cited Roland C. Summit, *608M.D., The Child Sexual Abuse Accommodation Syndrome, 7 Child Abuse and Neglect 177 (1983), for the proposition that repeated questioning of an abused child about a traumatic event will double the trauma. R.W., supra, 104 N.J. at 28 n. 3, 514 A.2d 1287.
The Court, however, appeared to recognize that there may be cases “in which an unusual condition bearing upon witness reliability, such as the influence of one parent over the child-witness,” may “constitute a basis for subjecting the child to a psychiatric examination.” Id. at 25, 514 A.2d 1287. The Court noted that it may have been appropriate to give defendant an opportunity to present evidence supporting his request for a psychiatric examination. Id. at 31, 514 A.2d 1287.
In State v. D.R.H., 127 N.J. 249, 604 A.2d 89 (1992), a case involving a request for a physical examination of a child-sexual-abuse victim, the Supreme Court recognized that a defendant in a criminal ease is entitled to broad discovery under R. 3:13-3, and that the judiciary has the inherent power to order discovery in a criminal prosecution “when justice so requires.” Id. at 256, 604 A.2d 89 (quoting In re W.C., 85 N.J. 218, 221, 426 A.2d 50 (1981)). However, it emphasized, as it had in R.W., that a defendant is not entitled to forage for evidence without a reasonable basis, and referred, as it had in R.W., to the emotional trauma and mental distress a forced physical examination could inflict. The Court concluded that a physical examination could constitute an invasion of privacy and present a danger of serious harm in the form of emotional trauma. Id., 127 N.J. at 258-59, 604 A.2d 89. Therefore, a defendant charged with sexually abusing a child must demonstrate a “substantial need” for a physical examination. Id. at 260, 604 A.2d 89. It ruled that a court’s inquiry must focus on the likelihood that such an examination will produce competent and probative evidence. Id. at 260-61, 604 A.2d 89.
The Supreme Court expressly rejected the theory that a defendant has an automatic entitlement to conduct a physical examina*609tion simply because the State had the child submit to an examination. Id. at 262, 604 A.2d 89. It held that:
[CJourts may order the physical examination of a child sex-abuse victim only when satisfied that the defendant has made a sufficient showing that such an examination can produce competent evidence that has substantial probative worth, and if admitted and believed by the trier of fact, that evidence could refute or neutralize incriminating evidence or impugn the credibility of prosecution witnesses. Further, the court must be satisfied that the defendant’s need clearly outweighs the possible harmful consequences to the alleged victim. [Id. at 260-61, 604 A.2d 89.]
D.R.H. quoted State v. Drab, 546 So.2d 54, 56 (Fla.Dist.Ct.App.), review denied, 553 So.2d 1164 (Fla.1989) for the holding that a compelled examination is not justified merely because of a defendant’s concern that a jury may afford greater weight to testimony from the expert who has actually conducted an examination versus one who has not. D.R.H., supra, 127 N.J. at 264, 604 A.2d 89.
Defendant maintained before the trial court that psychiatric examinations of the children were necessary to determine if they were really traumatized. Defendant offered no particularized reasons as to any conditions or characteristics of the majority of the children which would necessitate a psychiatric examination. There were specific conditions relating to four children. Two received treatment for a condition related to their parents’ divorce, and two were treated for seizure disorders. These conditions, however, were not shown to relate to their competency to testify, and defendant does not refer to them on appeal.
Defendant’s general assertion that a defense psychiatric examination of the children was necessary to determine whether they were traumatized is patently insufficient to allow such relief. Without more, a court could not evaluate the probative worth of the evidence to be obtained from an examination. Although the jury learned of Treacy’s and Dr. Esquilin’s access to the children, the State examination did not of itself trigger defendant’s right to an examination. Under D.R.H. a defense examination is not mandated because of the possibility that a jury might accord more weight to the opinion of an expert who has conducted an examination. Id. at 264, 604 A.2d 89.
*610We need not speculate as to whether relief might have been in order if, at an evidentiary hearing, defendant interlinked the suggestibility of the questioning of the children by the authorities to the ability of the children to testify. In the event of a retrial, the issue of whether the examinations are necessary to a proper defense will take on a far different perspective because of the passage of time and the outside influences on the now teenaged children caused by family dialogue and publicity surrounding the many years of litigation.
Defendant also contends that her psychiatric expert was hindered in her ability to perform a “confounding variables” analysis because she could not interview the children. We cannot conclude that the defense was unduly hampered in this case as it had access to the grand jury transcripts of parents and other witnesses which revealed stress factors in the children’s lives. The experts could have prepared relevant questions for defense counsel to ask of the various witnesses, including the children and parents, or the defense could have acquired any necessary facts through independent investigation.
Similarly, defense counsel had the opportunity to cross-examine the State’s experts who testified that, based on their personal observations, the children were traumatized. In addition, their testimony was available for review and rebuttal by defense experts. It was not essential that the defense have access to the children to challenge the experts’ conclusions. Defendant countered the State’s experts’ opinions that the described behaviors evidenced trauma by inquiring as to what stress factors other than sexual abuse could have caused the behavior, and by challenging, through defense experts or cross-examination, the premise that the behaviors were unusual. In any event, we do not expect this problem to recur under the restricted use of such testimony mandated by the holding in J.Q., supra.
Regarding the defense need to examine the children in order to counter the State’s CSAS testimony, we conclude that, in view of the extensive discovery provided to defendant and the availability *611of grand jury testimony, the inability to psychiatrically or psychologically examine the children did not impede the defense’s ability to challenge the validity of CSAS. The defense cross-examined the experts as to whether various behaviors could be indicative of problems other than sexual abuse, and it cross-examined the children’s parents as to other life events which may have precipitated the alleged behaviors their children were exhibiting. We also expect that in future cases a transcript or videotape of the judge’s interview with the children may be made available to the experts to aid in forming an opinion as to the need and level of protection required for the witness.
III.
Defendant, in contending that the trial court improperly allowed the Wee Care children to testify by CCTV, argues that: (1) the testimony by psychologists and parents concerning the inability of the children to confront defendant in open court was insufficient as a matter of law because the trial court itself never interviewed the children; (2) parental testimony can never, standing alone, support use of the procedure; and (3) a psychologist’s testimony with respect to four children was invalid because of the psychologist’s blanket statement that the CCTV procedure should be used for any preschool age child who is an alleged victim of sexual abuse. Further, defendant urges that the use of CCTV in this case undermined the presumption of innocence because it tacitly reinforced the State’s expert testimony that the children were traumatized and fearful of defendant, thereby implying that defendant had abused them.
N.J.S.A. 2A:84A-32.4a provides that, in cases of sexual assault and child abuse, a court may order the taking of testimony of a child witness by closed circuit television. The court must conduct an in camera hearing prior to entering such an order. It must find that the witness is sixteen years of age or younger and that there is a “substantial likelihood that the witness would suffer severe emotional or mental distress if required to testify in open *612court.” N.J.S.A. 2A:84A~32.4a, b. The order must be specific as to whether the witness will testify outside the presence of spectators, the defendant, the jury, or all of them, and “shall be based on specific findings relating to the impact of the presence of each.” N.J.S.A. 2A:84A-32.4b.
The trial court held in camera hearings in which the State presented testimony from the parents and/or treating psychologists of the children. These witnesses gave their opinions as to whether the children would suffer severe mental or emotional distress if required to testify in front of defendant and/or a jury and spectators. Defense counsel cross-examined the witnesses, but presented no witnesses of its own after having failed in its motion to have the children examined by its own psychologist.
All of the parents and psychologists expressed the belief that the children would suffer emotional distress if forced to testify in front of defendant; however, some of the witnesses believed that various children might be able to testify in open court if defendant was not present. Defense counsel represented that defendant would voluntarily absent herself from the courtroom if necessary for a child witness to testify in open court. See State v. Crandall, 120 N.J. 649, 658-59, 577 A.2d 483 (1990).
At the conclusion of the pretrial proceeding, the judge found that all of the children would suffer severe emotional distress if forced to testify in defendant’s presence, and that most of them would be traumatized by testifying in front of a jury and spectators. For those children who might be able to testify in open court, the judge stated that he would wait until the day of their testimony to determine whether they were actually able to do so.
In reaching his determinations, the judge did not interview the children and made no independent observations with respect to each child. The judge merely repeated the testimony of the parent or therapist who had testified concerning that prospective witness. He stated that he had read the transcripts of the children’s grand jury testimony, and asserted that the children had trouble communicating even without defendant’s presence.
*613The United States Supreme Court in Maryland v. Craig, 497 U.S. 836, 110 S.Ct. 3157, 111 L.Ed.2d 666 (1990), stated that, in order to use CCTV, a trial court must: (1) hear evidence and determine that the procedure is necessary to protect the welfare of the particular child witness who seeks to testify; (2) find that the child witness would be traumatized, specifically by the presence of the defendant, not merely by the courtroom; and (3) find that the distress experienced would be more than de minimis— more than mere nervousness, excitement or some reluctance to testify. 497 U.S. at 856, 110 S.Ct. at 3169, 111 L.Ed.2d at 685.
Crandall, supra, upheld the constitutionality of N.J.S.A. 2A:84A-32.4 against a challenge that it violated a defendant’s Sixth Amendment right to confrontation. Our Supreme Court in Crandall directed that a trial court examine and evaluate a child witness before determining the child’s ability to testify in defendant’s presence. The Court found that the trial judge had satisfied all the Craig criteria and that N.J.S.A. 2A:84A-32.4 was facially constitutional because it required both case-specific findings that the closed-circuit procedure is necessary and a finding that severe emotional or mental distress would result from requiring the child to testify in open court. Crandall, supra, 120 N.J. at 655-57, 577 A.2d 483.
Our Supreme Court delineated the procedure a trial court should follow in considering a motion under N.J.S.A. 2A:84A-32.4.
We conclude that trial courts should conduct a thorough face-to-face interview with the child and make detailed findings concerning the child’s objective manifestations of fear. We decline to hold that expert testimony is required to show that a child will suffer severe emotional or mental distress from testifying in open court. If, after reviewing the evidence at the hearing, a court is unable to make a determination on its own, it may then appoint an expert to evaluate the child. [120 N.J. at 663-64, 577 A.2d 483.]
There is, however, no Sixth Amendment constitutional requirement that a court personally observe a prospective child witness at a pretrial hearing in order to allow the child to testify via CCTV. Craig, supra, 497 U.S. at 858-61, 110 S.Ct. at 3170-71, 111 L.Ed.2d at 687-88. In addition, we do not interpret Crandall *614to require us to invalidate an otherwise sound judgment, reached before its ruling permitting CCTV testimony, merely because of the lack of a face-to-face interview with the witness. Nonetheless, the trial court’s failure in this case to interview the children or to make specific findings with respect to each child leads us to the conclusion that N.J.S.A. 2A:84A-32.4 was improperly applied.
We have had the opportunity to review the videotapes of the children’s testimony in their entirety as presented to the jury from the judge’s chambers. Our personal observations cast serious doubt, with one or two possible exceptions, upon the trial judge’s broad conclusion that the children would suffer serious emotional distress if compelled to testify in open court. The children apparently had previously testified before twenty-three grand jurors and legal staff without any protective measures, in a room they described as comparable to the trial judge’s courtroom. They had also visited the courthouse and courtroom. More importantly the children manifested virtually no reticence or emotion when speaking of the defendant or the alleged acts of abuse. Further, the children were subjected to repeated, intense investigative interviews without untoward consequences being reported. Having been visited by the prosecutor’s staff the night before their testimony in almost every instance, their testimony appeared well prepared, rote and detached.
Further, under N.J.S.A. 2A:84A-32.4b, a trial court must make specific, particularized findings concerning the child’s objective manifestations of fear about testifying in the particular environment being considered. This was not done, and, in many instances, we do hot believe that the record supports a finding of “a substantial likelihood that the witness would suffer severe emotional or mental distress if required to testify in open court.” The testimony of the experts was in most instances general and lacking in detail.
We recognize that there are disadvantages associated with the use of CCTV testimony. See Hochheiser v. Superior Court, 161 Cal.App.3d 777, 208 Cal.Rptr. 273, 278-79 (1984); *615James J. Armstrong, Comment, The Criminal Videotape Trial: Serious Constitutional Questions, 55 Or.L.Rev. 567 (1976). However, we can conceive of circumstances where the use of videotape testimony, although not fully appropriate, might not require reversal. Unfortunately, we are unable to even consider whether the use of the CCTV procedure here was harmless. This is so because the in-chamber proceedings were conducted in a manner so far removed from proper standards of impartially presenting the testimony of the children witnesses that the defendant was denied a fair and impartial trial.
The trial judge, in his zeal to make the children feel at ease so that their testimony might be obtained, failed to recognize that he could be perceived as crossing the line between an impartial judge and the prosecution. The judge, in the televised-view of the jury, played ball with the children, held them on his lap and knee at times, whispered in their ears and had them do the same, and encouraged and complimented them. See State v. R.W., 200 N.J.Super. 560, 570-71, 491 A.2d 1304 (App.Div.1985), aff'd and modified, 104 N.J. 14, 514 A.2d 1287 (1986). The judge also unduly interfered with defense counsels’ cross-examination of the children and often took charge of the questioning, which in many instances was overly suggestive. For all appearances, the State’s witnesses became the judge’s witnesses. The atmosphere became such, after this manner of presentation of testimony from nineteen children, that a jury considering a verdict in favor of the defendant might feel that it was personally offending the judge. The required atmosphere of the bench’s impartiality was lost in this trial.
We do not question the integrity of the trial judge, who otherwise did a commendable job in a very long and difficult case. He clearly felt he was doing the right thing by assisting in the search for “the truth.” This case demonstrates the critical need for the judge to be impartial in an adversary proceeding, and it highlights how difficult a task it is in many instances. We as judges learn by doing and many times by our own mistakes. CCTV in child abuse *616trials was new to the courts, having been adopted by the Legislature only a relatively short time before the present trial. We caution that where testimony is taken in the judge’s chambers, special care must be taken to ensure that there is no suggestion of favoritism or prejudice to either side.
We reject as clearly without merit defendant’s arguments that (1) a parent’s testimony can never meet the standards of the statute, and (2) the testimony with respect to some of the children was improper because their therapist opined that all children in their age group would experience severe emotional distress from testifying in open court. R. 2:11—3(e)(2). Any generalized “presumption of innocence” argument must fail as Craig implicitly authorized CCTV testimony despite the possibility that some jurors might view the child’s inability or unwillingness to confront a defendant as indicative of the truth of the charges. See Crandall, supra, 120 N.J. at 659-60, 577 A.2d 483. Here, however, the improper manner of conducting the CCTV in chambers clearly prejudiced the defendant.
IV.
Defendant contends that the trial court erred in refusing to permit defendant to present to the jury a character defense buttressed by the testimony of Dr. Jonas Rappeport and other experts that defendant did not fit the psychological profile of one who would commit the crimes charged. Defendant maintains that expert testimony of her psychological condition was relevant to the issue of guilt and was based upon accepted scientific techniques. The trial judge heard testimony from Dr. Rappeport and two State experts. The court also had before it reports of psychological examinations made of defendant by two other defense experts. The theme of the three defense experts’ testimony was essentially identical: defendant exhibited no psychological personality characteristics normally associated with an individual who would commit the sexually abusive acts with which she was charged.
*617Dr. Rappeport, in his testimony, asserted that he had identified certain criteria or pathologies which enabled him, in his experience, to evaluate defendant and conclude that she did not have those characteristics. He explained that there was “a general body of knowledge upon which people operate.” While this body of knowledge is “not sacrosanct,” it does illustrate “certain things that one looks for.” Dr. Rappeport measured defendant’s responses, signs, and symptoms against these things. The State offered expert testimony to rebut Dr. Rappeport’s assertions.
The court was satisfied that the more up-to-date and knowledgeable information was provided by the State’s experts who denied the predictability of such deviant conduct. The court concluded that there are no particular mental characteristics peculiar to child sexual abusers, male or female. In any event, psychiatrists cannot so determine by examination. The judge stated: “That type of evidence, I’m satisfied, is worthless. The Court finds that none of the tests have been met either by recognized experts’ testimony, by scientific and legal writings that are responsible nor by judicial opinions.”-
Defendant appeals from the trial court’s ruling and argues that State v. Cavallo, 88 N.J. 508, 443 A.2d 1020 (1982), in which the Court ruled that the defense profile testimony should not be allowed, is distinguishable and should not be followed. Defendant urges us to follow the lead of California in People v. Stoll, 49 Cal.3d 1136, 265 Cal.Rptr. 111, 783 P.2d 698 (1989).
In Cavallo our Supreme Court held that expert opinion character testimony is admissible under Evid.R. 47, which generally allows character testimony in the form of opinion, reputation, or evidence of conviction of a crime which tends to prove a trait. Cavallo, supra, 88 N.J. at 515, 443 A.2d 1020. Expert opinion character testimony is, however, subject to Evid.R. 56(2) which sets forth the general requirements for expert opinion testimony: it must be based primarily on facts, data or other expert opinion established by evidence at trial and must be within the scope of *618the special knowledge, skill, expertise or training possessed by the expert witness.
Expert opinion character testimony is relevant for the same reason that all character testimony is relevant, that is, favorable evidence of a defendant’s character tends to raise doubts as to whether the defendant committed the offense. Cavallo, supra, 88 N.J. at 515-16, 443 A.2d 1020. When this expert opinion character testimony was measured against the scientific reliability requirement of Evid.R. 56(2), defendant’s case for admissibility failed in Cavallo as it does here because of the lack of scientific reliability of such testimony. Id. at 520-26, 443 A.2d 1020.
The Court in Cavallo stated that such evidence is admissible if it has “sufficient scientific basis to produce uniform and reasonably reliable results and will contribute materially to the ascertainment of the truth.” Id. at 517, 443 A.2d 1020 (quoting State v. Cary, 49 N.J. 343, 352, 230 A.2d 384 (1967)). The issue is whether there exists special knowledge, skill, experience or training by which the expert can determine, with reasonable reliability, whether an individual had the mental makeup that rendered the person unlikely to have committed the act charged on a specific occasion.
The Court reiterated the three recognized methods of proof by which scientific evidence can establish its general acceptance and accompanying reliability: (1) expert testimony, (2) scientific and legal writings, and (3) judicial opinions. Id., 88 N.J. at 521, 443 A.2d 1020. The Supreme Court noted that defendants offered no expert testimony as to the general acceptance by psychiatrists or any other medical community of the scientific premise upon which their expert based his analysis. Ibid. Nor did defendants cite any articles of which a court could take judicial notice of scientific acceptance of these principles. Id. at 522, 443 A.2d 1020. The Court was not persuaded by its review of judicial precedents that medical or legal communities generally accept the proposition that psychiatrists possess such knowledge or capabilities. Id. at 523-25, 443 A.2d 1020. Therefore, the Court concluded that the defendants in Cavallo failed to establish that the proffered expert *619opinion character evidence was based on reasonably reliable scientific premises. Id. at 526, 443 A.2d 1020.
We find the differences between this case and Cavallo to be inconsequential. The testimony in both instances was inadmissible because it was scientifically unreliable. The scientific and medical community have not recognized that psychiatrists were able to reliably determine the characteristics of sexual offenders, be they rapists as in Cavallo, or in this case, child abusers. Further, it was not demonstrated that such particular characteristics were even readily identifiable. The trial judge properly assessed the proffered expert opinion character testimony and correctly concluded that it was not based on reasonably reliable scientific premises. Ibid.
Stoll, supra, upon which defendant relies, based its holding on People v. Jones, 42 Cal.2d 219, 266 P.2d 38 (1954), which the Cavallo Court rejected as inapposite. In Stoll, as in Jones, the California court relied upon sections of the California Penal Code, which it interpreted as allowing an accused to present expert opinion testimony to indicate an accused’s nondisposition to commit a charged sex offense. This expert opinion testimony included an “introduction of defense expert opinion of ‘good character’ to show noncommission of charged crimes.” Stoll, supra, 265 Cal.Rptr. at 121, 783 P.2d at 708 (citation omitted). From this the court concluded that “the Legislature ... implicitly endorsed ‘lack of deviance’ as a relevant character trait in a lewd and lascivious conduct case, even though the ‘sexual psychopathy’ provisions cited in Jones” had been repealed in prior years. Ibid.
Clearly, this holding is distinguishable from our law in that no similar statutory provision exists in this State which relieves the court of the responsibility to determine the scientific reliability of such testimony. We are not alone in our rejection of such testimony. See James M. Peters and William D. Murphy, Profiling Child Sexual Abusers, Legal Considerations, 19 Crim. Justice and Behavior 38, 39 (1992) (“With the notable exception of courts in California, ... virtually every appellate court that has ruled on *620the admissibility of expert testimony regarding the psychological profile of child molesters has rejected it.”).
V.
We have combined two of the Points raised in defendant’s brief on appeal because of what we consider to be an overlapping of the underlying issue, and the need for a hearing on that question upon remand in the event of a retrial. In Point VI of her brief defendant argues, “Reversal is required because the questioning of Wee Care children was so suggestive and coercive that they were rendered incompetent to testify.” In Point IX it is argued that, “The admission of massive hearsay testimony so thoroughly polluted the trial that no remedy other than a mistrial was appropriate.” Although these contentions appear to be widely separated, they are closely linked by the underlying question of the trustworthiness of the children’s statements and testimony, allegedly thrown into doubt by improper interviews and techniques employed in the early stages of the investigation and the ensuing “26 months of investigatory questioning of Wee Care children [which] was suggestive and coercive.”
Defendant urges that the interviews of the children by investigators and the prosecutor for more than two years prior to trial were so suggestive and coercive that the court should have granted her motion to dismiss the indictments. She insists that “virtually every child-interviewing safeguard to preserve trustworthiness and reliability was violated.” As a result, the allegations were stripped of their trustworthiness, and the competency of the children as complainants was destroyed.
During the pendency of this appeal we granted defendant’s motion to supplement the record with the Manual of the National Center for Prosecution of Child Abuse. We also have been supplied with transcripts of investigatory interviews of the children and audio tapes of various interviews by investigators in June and July of 1985, when the children were three, four or five years of age. Not all of these children were named as victims in *621an indictment. Several of these interview tapes were played for the jury, and the jury had the transcripts available to them as the tapes were played. It does not appear the transcripts were formally introduced into evidence; however, the court agreed with defense counsel’s statement that the audiotapes themselves had been “incorporated into the record.”
Defendant points out and the State concedes that many of the interviews reveal extremely leading and/or suggestive questions. Certain questions planted sexual information in the children’s minds and supplied the children with knowledge and vocabulary which might be considered inappropriate for children of their age group. Children were encouraged to help the police “bust[ ] this case wide open.” Peer pressure and even threats of disclosing to the other children that the child being questioned was uncooperative were used. A child was told that she needed to talk to help her friends and that the investigator had already spoken to five other children who revealed what happened. In some cases, certain children were told in detail what another child had disclosed. Sexualized discourse was encouraged and applauded.
It is clear that all of these young children were convinced, either by their own experiences, other children, investigators, parents, or some combination thereof, that defendant was “bad” because she had done “bad things” to children at Wee Care. The trial testimony of most of the children began with this type of introduction. Children were told they could keep Kelly locked in jail by cooperating; therefore, they and their families would be safe. Anatomical dolls were used in the interviews, and in some cases the children did not disclose anything until they were either presented with the dolls, shown various eating utensils, or encouraged to demonstrate how Kelly might have hurt a little girl or boy. The records of the interviews show that these methods caused certain children to use their imagination and stray from reality, even to the dismay of the investigator at times. In several instances, the children were tired and/or resistant to participating *622in the interviews, but the investigators continued to press for cooperation.
The Manual of the National Center for Prosecution of Child Abuse contains guidelines which encourage interviewers to remain open and neutral and not to display reactions which would be interpreted as reinforcing certain responses. It cautions that leading questions, should be avoided, .and that a reluctant child should never be forced to talk or continue an interview. It indicates that anatomical dolls may not be appropriate in all instances, and that careful attention should be paid to the methods employed when they are used.
There is an enormous amount of literature on children’s memory, suggestibility, ability to distinguish fact from fantasy, and jurors’ perceptions of children’s credibility. See Josephine A. Bulkley, The Impact of New Child Witness Research on Sexual Abuse Prosecutions, in Perspectives on Children’s Testimony 208, 213 (Stephen J. Ceci et al. eds., 1989). Bulkley comments that the research is so overwhelming that even researchers cannot keep up with it. Id. at 215. Moreover, the views and conclusions of the researchers and writers vary greatly.
There is, however, consistent concern about the interview process and the possibility of distorting recollections by suggestive or leading questions. For example, two researchers (who acknowledge that many of their recommendations are slanted in favor of prosecution of sex abuse cases), emphasize that an interviewer should be highly trained and unbiased, because the least accurate reports in sex abuse cases are obtained from child witnesses when the interviewer harbors preconceived notions about what happened. Gail S. Goodman and Vicki S. Helgeson, Child Sexual Assault: Children’s Memory and the Law, 40 U. of Miami L.Rev. 181, 195, 207-08 (1985) (citing H.R. Dent, The Effects of Interviewing Strategies on the Results of Interviews with Child Witnesses, in Reconstructing the Past 279 (A. Trankell ed., 1982)).
Goodman and Helgeson also report that the most accurate information is obtained at the initial interview, and both they and *623more defense-oriented commentators recommend that the initial interview be videotaped. See id. at 195, 198-99; David C. Raskin & John C. Yuille, Problems in Evaluating Interviews of Children in Sexual Abzise Cases, in Perspectives on Children’s Testimony 184, 195-96 (Stephen J. Ceci et al. eds., 1989) [hereinafter Raskin and Yuille], Unfortunately, the initial interviews of the children in this case were not recorded. We do not have any reason to believe, however, that they were any less offensive than the subsequent interviews. See also Margaret A. Berger, The Deconstitutionalization of the Confrontation Clause: A Proposal for a Prosecutorial Restraint Model, 76 Minn.L.Rev. 557, 608 (1992) (suggesting that when interview is conducted of “vulnerable” witness—i.e., a child—prosecution should always provide tape or transcript of every interview to determine suggestibility).
In a recent article focusing on the decision in Idaho v. Wright, 497 U.S. 805, 110 S.Ct. 3139, 111 L.Ed.2d 638 (1990), the author suggests that the child’s initial statements concerning the alleged abuse are the most reliable. At this point the child has not been subjected to coaching or suffered from memory loss. Sharon Kennedy, Idaho v. Wright: The Confrontation Clause Limits on the Admissibility of Hearsay Evidence in Child Abuse Cases, 59 U.M.K.C.L.Rev. 1093, 1109-10 (1992) [hereinafter Kennedy].
When a child has been subjected to numerous interviews, there is a greater likelihood that his statement may be tainted by suggestions made during the interview. In addition, the child may suffer memory loss of certain details during the time lapse between the initial report and subsequent interviews. [Id. at 1110 (footnotes omitted).]
Defendant also insists that anatomical dolls were improperly used in the interviews. Some commentators voice concern about the use of anatomically-eorrect dolls. See Raskin & Yuille, supra, at 193-94 (dolls should be employed as aids of last resort because there are no standards for correct usage and almost no base data on how non-abused children interact with the dolls). Raskin & Yuille report two conflicting studies—one found clear differences between how abused and non-abused children interact with the dolls while another found that non-abused children played with the *624dolls in a manner indicative of abuse. Id. at 193. Another writer suggests, based on existing research, that the use of anatomically-correct dolls “can lead interviewers to believe that abuse has occurred” when it has not. Diana Younts, Evaluating And Admitting Expert Opinion Testimony In Child Sexual Abuse Prosecutions, 41 Duke L.J. 691 (1991) [hereinafter Younts],
John E.B. Myers, a well-respected expert in the field of child abuse, states that by three years of age, children are able to remember “salient events quite well.” 1 John E.B. Myers, Evidence in Child Abuse and Neglect Cases 76 (1992). Myers notes that children are more likely to remember events that they participated in, as opposed to events they merely observed. Id. at 78 (footnote omitted). He notes, however, that there are conflicting reports as to whether stressful events increase or decrease the ability to remember. Id. at 78-79 (footnotes omitted).
Myers differentiates between free recall and recognition. Free recall is much more difficult because younger children “spontaneously recall less information than older children and adults.” Id. at 79 (footnote omitted). Myers notes that children do lie, but no more so than adults. He states:
Situational factors influence the moral decisions of adults and children. A child’s decision to stray from the truth or engage in other improper behavior may be influenced by peer pressure, pressure from, influential adults, the likelihood of detection, motive to fabricate, desire to engage in the prohibited activity, and attempts to escape blame or punishment. [Id. at 95 (emphasis added) (footnote omitted).]
Myers also discusses the ramifications which may result from children’s tendencies to fantasize more than adults.
Although children can distinguish real from pretend, issues of personal knowledge may arise when children are subjected to highly suggestive questions during interviews. The possibility arises that a child’s memory of actual events is distorted or replaced by “events" suggested by the interviewer. Alternatively, an interviewer might encourage a child to fantasize events, and then encowivge the child to adopt these fantasies as reality. In such scenarios, the question is whether the child has sufficient personal knowledge to testify, or whether improper interview methods robbed the child of personal knowledge. [Id. at 113-14 (emphasis added) (footnote omitted).]
*625Myers wrote another article in which he discussed the suggestibility of children. He relied heavily on research conducted by the two psychologists, Elizabeth Loftus and Graham Davies. John E.B. Myers, Child Witness: Techniques for Direct Examination, Cross-Examination, and Impeachment, 18 Pac.L.J. 801, 888 (1987). Loftus and Davies suggested that children have a more difficult time retrieving information from long-term memory.
Subjects were especially prone to suggestion when considerable time, say several days, had elapsed between the initial event and the introduction of misinformation. Apparently, when the initial memory is weakened over time, it becomes especially vulnerable to the introduction of new inputs. [Ibid, (quoting Elizabeth Loftus and Graham Davies, Distortions in the Memory of Children, 40 J.Soc. Issues 51, 56-57 (1984)).]
Loftus and Davies further concluded that if the information is not “encoded well” from the outset, or if a delay weakens the child’s memory, “the fragments of the event that remain in the child’s memory may not be sufficient to serve as a barrier against suggestion, especially from authoritative others.” Id. at 889 (citing Loftus and Davies, supra).
Thus, Myers concluded:
As memory fades over time, its accuracy may decline while suggestibility increases. Fading memory is susceptible to new input which can come from suggestive questioning. Inaccurate information can be incorporated into memory, actually supplanting accurate data. Yet, when the child takes the stand, he or she testifies believing in the accuracy of what is said. [Ibid. ]
Significantly, in the present case, a good deal of time elapsed between the alleged events and the children’s statements regarding the events. Therefore, according to Myers, the factors which can influence suggestibility include: 1) whether the interviewer believes in the presumption of guilt; 2) whether the questions asked were leading or nonleading; and 3) whether the interviewer was a “trusted authority figure.” Ibid.
Myers also notes that child-sexual-abuse cases often involve coached testimony.
In abuse litigation, children often endure a series of interviews with police officers, child protective service workers, therapists, investigators, and attorneys. The possibility of coaching exists during each interview. While the law enforcement and legal professionals who work with abused children attempt to guard against *626improper influence, the desire to win, which drives the adversary system, sometimes tempts conscientious individuals over the line. [Id. at 892 (footnote omitted).]
Mary Ann King and John C. Yuille suggest that if an interviewer gives signals as to what the interviewer is searching for in an answer, children are very responsive to such signals. Mary Ann King and John C. Yuille, Suggestibility and the Child Witness, in Children’s Eyewitness Memory 29 (Stephen J. Ceei et al. eds., 1987).
Relative to adults, children are more suggestible because they find themselves in more situations in which they are unfamiliar. As a consequence, children are likely to pay attention to anyone (especially an adult) who they believe knows how to behave in that situation____ [Y]ounger children can be expected to be particularly sensitive to contextual cues in a verbal situation where the child is supposed to listen and respond to questions and instructions from an interviewer. If children are interviewed concerning events they have understood, that is, an event they could assimilate to their level of social and cognitive competence, and if they are interviewed in a manner that is consistently meaningful and not contradicted by nonverbal cues, then they should be no more suggestible than adults____ But if a child lacks competence concerning the event, and/or the interviewer is not adept at anticipating potential confusions, then problems will arise. Thus, the child will be susceptible to leading questions and may make an erroneous choice ... or he or she may invent answers to questions. [Id. at 30-31.]
King and Yuille, in discussing the problems surrounding interviewing techniques, also state that the use of anatomically-correct dolls during an interview is in itself a form of suggestive questioning. Id. at 31.
Experiments conducted on children who varied in age from three to twelve years suggested that younger children, after receiving misleading information, provided less accurate details about the original event than did older children. Stephen J. Ceci et al., Age Differences in Suggestibility, in Children’s Eyewitness Memory 82 (Stephen J. Ceci et al. eds., 1987). Ceei and his colleagues observed that younger children may be more likely to conform their answers in the hope of pleasing an adult interviewer. “Clearly, the youngest children demonstrated a sensitivity to the age of the manipulator and a desire to conform to their perceptions of an adult authority figure’s expectations.” Ibid. Ceci and his colleagues concluded that if erroneous information is
*627suggested to the young child, this erroneous information may “resurface in the form of the child’s reconstruction of the events” if the child is given a choice between the original information and the misleading information. Id. at 90.
In a recent article specifically examining children’s suggestibility, the author extensively examined current social science literature and came to the following conclusions:
Careful review of the social science literature indicates that children are susceptible to suggestive interviewing techniques and that such techniques can render children’s accounts of abuse unreliable. A number of studies have shown that children will lie when they have a motivation to lie, that they are susceptible to accommodating their reports of events to fit what they perceive the adult questioner to believe, and that inappropriate post-event, questioning can actually change a child’s cognitive memory of an event. Even the studies that concluded that children are resistant to suggestion found a small percentage of children who were not. [Younts, supra, 41 Duke L.J. at 692 (footnotes omitted).]
Several factors can influence children to provide misleading information: (1) adults may misinterpret what a child states; (2) the possibility of abuse may lead to hysteria; or (3) an adult may have malicious motives. Id. at 697. In addition,
As there is more media coverage of sexual abuse, parents and the professional community are more likely to suspect sexual abuse as a cause for symptom formation, even when sexual abuse has not occurred. [Id. (quoting Elissa P. Benedek & Diana H. Schetky, Problems in Validating Allegations of Sexual Abuse, Part 2: Clinical Evaluation, 26 J.Am.Acad. Child & Adolescent Psychiatry 916, 917 (1987)).]
Younts believes that Stephen Ceci’s 1990 study provides the most impartial results of the nine studies Younts examined. Ceci found that “even small material and psychological rewards” prompted children to lie about events. Id. at 722. According to Younts, when a study provides motivational factors for lying, children will often lie. Ibid.
Younts reviewed a study conducted in 1989 by Clarke-Stewart in which children interpreted an event which could either be sexually abusive or innocent.
The researchers found that when the adult interviewers contradicted what actually happened to the children, two-thirds of the children changed their stories to conform to the suggestions of the interviewer. Most of the remaining children merged their account of the factual events with the interviewer’s suggestions. At *628the end of the interrogation, only one child answered all the questions accurately. When the second interviewer contradicted the first interviewer, the children again changed their stories. The children’s reports to their parents were consistent with the interviewers’ suggestions. [Id. at 723 (footnotes omitted).]
Another study by Ceci revealed similar results. Ceci found that children are very susceptible to modifying their story based upon an adult’s post-event suggestions. However, children are even susceptible to suggestions by older children. Id. at 724. The suggestiveness can be incorporated even when the child retains memories of the original event. Based upon these two studies, Younts suggests that “courts should pay particular attention to whether the abuse investigator had a preconceived notion of what happened to the child and then sought the child’s confirmation.” Id. at 725.
Younts expresses the opinion that post-event suggestion poses a significant problem especially with children six-years old or younger. Id. at 726. When interviews include suggestive and leading questions, children may eventually incorporate the suggested responses into memory. Vitally important is the fact that the children’s credibility will not be disturbed because the children actually believe what they are saying. Id. at.727.
Younts provides some suggestions for interviews:
Because the research shows that a significant percentage of children are responsive to adult preconceived ideas of what has happened to the children, it is crucial to be aware of the common pitfalls of investigative interviews and to recognize that interviewers often give children cues as to what they expect the children to relate through means other than leading questions. A major goal for investigators should be to insure that a child’s information does not become distorted or falsified by confounding influences. Furthermore, “[i]t must be acknowledged that all interviews will be influenced to some extent by interviewer behaviors. All investigatory interviewers must be vigilant during the interviewing process to minimize the impact of behavioral factors.”
Factors of the interview itself that can falsify and confound a child’s report include the lack of investigatory independence, pursuit of an agenda, leading, and coercion. Investigatory independence requires the interviewer to adopt an “objective stance of not allying him/herself with any particular individual involved in the investigation of the allegation.” The interviewer should also maintain independence by not giving the child verbal and behavioral cues that can contaminate the interview. An example of the latter would be stroking the child’s head as a *629reinforcement for an answer that confirms the interviewer’s assumptions about the allegation. [Id. at 729-30 (footnotes omitted).]
Younts notes that one of the worst things an interviewer can do to adversely affect reliability is to follow a preconceived agenda. Id. at 730.
Finally, Younts notes other factors which may alter the child’s statements:
Other influences that may alter the child’s statements include factors outside the interview. These include communications between parents and children, media coverage, formal educational experiences regarding child safety and abuse prevention, and other discussions that concern the abuse in the presence of the child. The investigator must evaluate this contamination’s impact on how the disclosure was made and the content of the allegations. [Id. at 730-31 (footnote omitted).]
The record of available interviews does not disclose that any of the children related their testimony of the alleged abuse by “free recall.” Although some of the parents testified as to statements made by the children, no evidence of the circumstances surrounding the statements was offered. Besides the improper interview techniques, which included all of the factors the experts indicate are to be avoided, many of the children played with other members of the day care center and underwent therapy together. In view of disclosures to that effect from some of the children, it is clear that exchanges of information took place between the children. Although one can only speculate as to the extent the children discussed the alleged events with one another, it seems more likely than not that considerable information was exchanged.
In State v. Wright, 116 Idaho 382, 775 P.2d 1224 (1989), the case ultimately affirmed by the United States Supreme Court, the court noted some experts’ opinions that “development of accurate recall skills” does not occur until the child is five years of age. Id., 775 P.2d at 1227. Importantly, the Idaho court looked to several sources and concluded:
The risk with young children is that they may be unable to distinguish between a memory of something which actually happened from a memory of something they imagine happening.... If an interview technique leads a child to imagine an event, the child’s memory of that imagined event will be indistinguishable from memories of events which the child actually experienced.... Once this tainting of memory has occmred, the problem, is irrem.ediable. That memory is, from then *630 on, as real to the child as any other. [7d at 1228 (emphasis added) (citations omitted).]
This observation clearly and concisely spells out our concern that, in the event of a retrial, the suggestibility issue be adequately addressed in the present case.
Defendant’s assertions that the indictments should be dismissed are rooted in the contention that the children’s grand jury testimony was fatally defective. Defendant further contends that any subsequent trial testimony was also tainted because the children’s capacity to independently recall events at Wee Care had been destroyed and replaced with allegations suggested by interviewers or by discussions among the children themselves. We do not have the benefit of the trial court’s reasoning on the motion because the court reporter’s notes were misplaced and never found. However, the parties agree that the judge analogized defendant’s claim to one seeking to prevent an in-court identification on the grounds that it would be irremediably tainted by the suggestiveness of a pretrial identification. We find the analogy appropriate. See State v. Hurd, 86 N.J. 525, 547-48, 432 A.2d 86 (1981).
In United States v. Stevens, 935 F.2d 1380, 1388-92 (3d Cir. 1991), the court directed that an inquiry be made into whether an in-court identification was tainted by the nature of pretrial identification procedures where those pretrial identification procedures were so impermissibly suggestive that the prior identification lacked reliability. Stevens derives from a line of cases in which federal courts have held that an out-of-court identification will be admitted at trial if, despite its suggestiveness, other factors support its reliability. See Manson v. Brathwaite, 432 U.S. 98, 114, 97 S.Ct. 2243, 2253, 53 L.Ed.2d 140, 154 (1977); Stovall v. Denno, 388 U.S. 293, 302, 87 S.Ct. 1967, 1972, 18 L.Ed.2d 1199, 1206 (1967); accord State v. Clausell, 121 N.J. 298, 325-26, 580 A.2d 221 (1990); State v. Madison, 109 N.J. 223, 536 A.2d 254 (1988).
We have applied this “taint” analysis in the context of a motion to dismiss an indictment. State v. Peterkin, 226 N.J.Super. 25, *631543 A.2d 466 (App.Div.), certif. denied, 114 N.J. 295, 554 A.2d 850 (1988). In Peterkin, we considered whether indictments should be dismissed because of police failure to preserve the photographic arrays from which an investigator had identified the individuals from whom he had made drug purchases. Id. at 31, 39, 543 A.2d 466. We held that the “prosecutions should be aborted only upon a finding that defendants’ rights to a fair trial have been irretrievably lost,” a question which depended in part on whether the police officer could make an in-court identification of the defendants independent of the taint caused by the failure of the police to preserve the arrays. Id. at 39, 543 A.2d 466 (emphasis added). We ordered a remand to explore this point. Ibid.
In Hurd, supra, our New Jersey Supreme Court stated:
In cases like the present one, where the use of hypnosis results in a subsequent out-of-court or in-court identification of the defendant, the defendant may also challenge on constitutional grounds the admissibility of testimony concerning the identification. The Supreme Court has held that an in-court identification following a pretrial identification procedure must be excluded if the pretrial proceeding was “so unnecessarily suggestive and conducive to irreparable mistaken identification” as to amount to a violation of due process. Stovall v. Denno, 388 U.S. 293, 302, 87 S.Ct. 1967, 1972, 18 L.Ed.2d 1199 (1967). The Court has elaborated on this rule by stating that “reliability is the linchpin in determining the admissibility of identification testimony.” Manson v. Brathwaite, 432 U.S. 98, 114, 97 S.Ct. 2243, 2253, 53 L.Ed.2d 140 (1977). In determining whether introduction of identification testimony would violate due process, the corrupting effect of the suggestive identification is weighed against the factors set out in Neil v. Biggers, supra, 409 U.S. [188] at 199, 93 S.Ct. [375] at 382 [34 L.Ed.2d 401] [(1972)], indicating the accuracy of ihe identification. Manson v. Brathwaite, supra, 432 U.S. at 114, 97 S.Ct. at 2253. (Hurd, supra, 86 N.J. at 547-48, 432 A.2d 86.]
Because the State has “the burden of demonstrating reliability by clear and convincing proof, any other showing of suggestiveness amounting to a constitutional violation should be the burden of the defendant.” Id. at 548, 432 A.2d 86; State v. Matthews, 233 N.J.Super. 291, 295, 558 A.2d 1329 (App.Div.1989).
We have found no direct precedent for applying a “taint” analysis to child-witness testimony in sex abuse cases. However, we believe that courts must provide a remedy where the record demonstrates that an accuser’s testimony is founded upon *632unreliable perceptions, or memory caused by improper investigative procedures if it results in a defendant’s right to a fair trial being irretrievably lost. A factual hearing would be required for this purpose. Hurd, supra, 86 N.J. at 543-47, 432 A.2d 86; accord State v. L.K., 244 N.J.Super. 261, 271, 582 A.2d 297 (App.Div.1990) (principles of Hurd applied to defendant in criminal cases). Further, the testimony of a child-victim might be stricken without dismissal of the charge in an appropriate case if there is other independent untainted evidence of guilt.
One unpublished Hawaii opinion apparently has held that two children were disqualified from testifying in a sexual abuse case, based on an expert’s opinion that “each child had been subjected to layers and layers of interviews, questions, examinations, etc., which were fraught with textbook examples of poor interview techniques.” State v. McKellar, No. 85-0553 (Haw.Cir.Ct. Jan. 15, 1983), cited in Josephine A. Bulkley, The Impact of New Child Witness Research on Sexual Abuse Prosecutions, in Perspectives on Children’s Testimony 208, 218 (Stephen J. Ceei et al. eds., 1989).
Turning to the somewhat related issue of the admissibility of the children’s hearsay statements, we note that the defendant did not request and the court did not conduct separate Evid.R. 8(1) hearings in order to predetermine whether the children’s statements were probably trustworthy before allowing the parents to testify. A hearing is required under Evid.R. 63(33)(b) to determine whether “on the basis of the time, content, and circumstances of the statement there is a probability that the statement is trustworthy.” Evid.R. 63(33). This hearsay exception, adopted and effective on June 13,1989, would be applicable in the event of a retrial, thereby rendering, subject to subsection (c), qualifying pretrial hearsay statements concerning abuse admissible as substantive evidence. R.S. v. Knighton, 125 N.J. 79, 97, 592 A.2d 1157 (1991); State v. Bethune, 121 N.J. 137, 146, 578 A.2d 364 (1990).
*633 Evid.R. 63(33) permits admission of the children’s pretrial statements about sexual abuse if:
(a) the proponent of the statement makes known to the adverse party his intention to offer the statement and the particulars of the statement at such time as to provide him with a fair opportunity to prepare to meet it;
(b) the court finds, in a hearing conducted pursuant to Rule 8(1), that on the basis of the time, content, and circumstances of the statement there is a probability that the statement is trustworthy; and
(c) either (i) the child testifies at the proceeding, or (ii) the child is unavailable as a witness and there is offered admissible evidence corroborating the act of sexual abuse.... [Evid.R 63(33).]
Some insight into the issue of trustworthiness is found in the United States Supreme Court’s decision in Idaho v. Wright, supra. The case dealt with whether a child’s out-of-court statements, in response to leading questions, had sufficient indicia of reliability and trustworthiness to justify admission against a defendant over Confrontation Clause objections. Wright, supra, 497 U.S. at 808, 110 S.Ct. at 3143, 111 L.Ed.2d at 648. The Court held that the statements of a two and one-half year-old child, deemed incompetent to testify, could not be admitted because: (1) they were elicited in a suggestive manner by the child’s physician; (2) there was no reason to deem them particularly trustworthy; and (3) they were not otherwise admissible under a “firmly rooted” hearsay exception. 497 U.S. at 827, 110 S.Ct. at 3152, 111 L.Ed.2d at 659-60. It held that the inquiry should be whether the child was particularly likely to be telling the truth under the totality of the circumstances in which the statement was made. 497 U.S. at 820, 110 S.Ct. at 3149, 111 L.Ed.2d at 655-56.
The Court commented that spontaneity in and of itself is insufficient to justify admission. “If there is evidence of prior interrogation, prompting, or manipulation by adults, spontaneity may be an inaccurate indicator of trustworthiness.” 497 U.S. at 827, 110 S.Ct. at 3152, 111 L.Ed.2d at 659-60 (quoting State v. Robinson, 153 Ariz. 191, 735 P.2d 801, 811 (Ariz.1987)). The Court adopted a “totality of the circumstances” standard for evaluating the admissibility of out-of-court statements, and specifically rejected the Idaho court’s focus on the lack of procedural *634precautions surrounding the interview—ie., that the physician did not tape the interview, asked leading questions, and questioned the child with a preconceived idea that sexual abuse had occurred. 497 U.S. at 817, 110 S.Ct. at 3148, 111 L.Ed.2d at 654. It reasoned that out-of-court statements made by children regarding sexual abuse arise in a wide variety of circumstances. The Constitution does not impose a fixed set of procedural requirements to permit admission of the statements at trial. Ibid.
Furthermore, it is particularly noteworthy that the Supreme Court rejected the Idaho court’s holding that the existence of corroborating evidence of a challenged statement should bear on its admissibility. The Supreme Court deemed this “bootstrapping” and inconsistent with the requirement that hearsay evidence admitted under the Confrontation Clause be so trustworthy that cross-examination would be of marginal utility. 497 U.S. at 822, 110 S.Ct. at 3150, 111 L.Ed.2d at 657. We perceive that corroborating evidence or prior disclosures might be proper at a hearing on suggestibility to help assess the reliability of a child’s proposed in-court testimony, even though the Supreme Court has held that other evidence cannot support the reliability of hearsay statements which are not subject to effective cross-examination.
New Jersey courts have recognized that the nature of questioning in child-sex-abuse cases may raise doubts about the veracity of the allegations. Our Supreme Court in Bethune considered the degree of questioning which may precede a statement sought to be admitted as a fresh complaint. It commented that a trial court must determine the line between questioning that merely precedes a complaint and coercive questioning, which disqualifies a statement for admission on fresh complaint grounds. Bethune, supra, 121 N.J. at 145, 578 A.2d 364. Among the factors to be considered in arriving at this determination are
the age of the child, the child’s relationship with the interviewer, the circumstances under which the interrogation takes place, whether the child initiated the discussion, the type of questions asked, whether they were leading, and their specificity regarding the alleged abuser and the acts alleged. [Ibid. ]
*635 Bethune implies that the degree of probing permissible in the Evid.R. 63(33) context may be substantially greater than that allowed under the fresh complaint exception. Id. at 146, 578 A2d 364. We point out, however, that reliability and trustworthiness are constitutionally mandated by the Confrontation Clause and may not be statutorily abrogated. See Ohio v. Roberts, 448 U.S. 56, 66, 100 S.Ct. 2531, 2539, 65 L.Ed.2d 597, 608 (1980); The General Assembly of State of N.J. v. Byrne, 90 N.J. 376, 391, 448 A.2d 438 (1982). We again caution that not only must subsection (b) be satisfied, but (c) as well. In that respect we offer no opinion as to when a child’s in-court testimony is so tainted as to be inadmissible that it would render the child “unavailable” within the meaning of Evid.R. 63(33)(c). This question may depend on whether the out-of-court statement preceded the taint or not. See State v. D.R., 109 N.J. 348, 369-70, 537 A.2d 667 (1988).
We do not rule on the trial court’s handling of these matters. We have set forth the foregoing in the expectation that it may provide future guidance.
VI.
Even though it is not required in view of our present disposition of this appeal, we choose, because of the seriousness of the allegations, to consider defendant’s assertion that the State’s summation amounted to prosecutorial misconduct so egregious as to deny her a fair trial. Defendant cites four instances of misconduct during summation: (1) the prosecutor poisoned the trial by equating defendant with Adolf Hitler; (2) there was an ad hominem attack upon the character of defense witness Dr. Ralph Underwager; (3) the prosecutor advised the jury of her personal belief that defendant was guilty; and (4) the word “guilty” was impermissibly printed on a board during part of the summation.
The prosecutor is entitled to be forceful and graphic in arguing to the jury as long as the summation fairly comments on the evidence. State v. Rose, 112 N.J. 454, 518, 548 A.2d 1058 (1988); State v. DiPaglia, 64 N.J. 288, 305, 315 A.2d 385 (1974). *636The prosecutor is limited to reasonable inferences supported by the evidence, but entitled to make a vigorous and forceful presentation of the State’s case. State v. Zola, 112 N.J. 384, 426, 548 A.2d 1022 (1988), cert. denied, 489 U.S. 1022, 109 S.Ct. 1146, 103 L.Ed.2d 205 (1989); State v. Bucanis, 26 N.J. 45, 56, 138 A.2d 739, cert. denied, 357 U.S. 910, 78 S.Ct. 1157, 2 L.Ed.2d 1160 (1958).
Reversal on the ground of prosecutorial misconduct in the summation should occur only when the improper comments so infected the trial as to render 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); State v. Pennington, 119 N.J. 547, 565, 575 A.2d 816 (1990) (misconduct must be so egregious that it deprives defendant of fair trial). When the prosecutor exceeds the bounds of fair comment on the evidence, a reviewing court does not automatically reverse, but instead must determine whether the misconduct was prejudicial and denied the defendant a fair trial.
The determination of whether prosecutorial misconduct denied defendant the right to a fair trial must take into account the tenor of the trial and the degree of responsiveness of both counsel and the court to improprieties when they occurred____ “In determining whether prosecutorial misconduct is prejudicial and denied defendant a fair trial, we consider whether defense counsel made a timely and proper objection, whether the remark was withdrawn promptly, and whether the court ordered the remarks stricken from the record and instructed the jury to disregard them.” [State v. Marshall, 123 N.J. 1, 153, 586 A.2d 85 (1991) (citations omitted).]
Where the prosecutor’s conduct has been so egregious as to prejudice a fair trial, we “have not hesitated to find reversible error due to prosecutorial misconduct.” State v. Pratt, 226 N.J.Super. 307, 323, 544 A.2d 392 (App.Div.1988); see State v. Johnson, 65 N.J. 388, 391-92, 323 A.2d 450 (1974); State v. Spano, 64 N.J. 566, 567, 319 A.2d 217 (1974); State v. West, 29 N.J. 327, 338, 149 A.2d 217 (1959).
Defendant’s closing spanned three days while the State’s summation covered six full days. This is the setting in which the State’s one reference to Hitler must be evaluated. The exact words employed by the prosecutor were:
*637[Defense counsel] stated that the State will seek to excuse its conduct in a manner similar to the way the Nazi’s [sic ] sought to excuse theirs at Neuremberg (phonetic) [sic ]. We were only doing our job. Everyone of us from the State truly resents being equated with the Nazis. There are no words to describe and to express the distaste we feel towards him for that comment.
This is a long, grueling, painful, intense, sad investigation for everybody involved. And with each revelation, horror. Witch hunt? Nonsense. Nazis? No one is making excuses. This investigation was conducted fairly and this investigation, unrelentingly leads to the conviction within ourselves that this defendant perpetrated these acts upon these children.
Defense says, impossible. Improbable. Impossible, couldn’t happen. April 12 of 1912, approximately, a ship sailed out from South Hampton [sic], England. There are very few lifeboats in that ship because that ship couldn’t sink. It was an unsinkable ship. It could not be sunk, no matter what. And on April 15th of 1912, April 14th, that ship, on its maiden voyage, the first time it ever sailed, that unsinkable ship, hit an iceberg. And on April 15th, that unsinkable ship, on its maiden voyage, sunk to the bottom of the sea killing a thousand people. And in 1938, one man, one man was able to tell hundreds of other grown-ups, adults, that many—that Jews and Gypsies and Blacks and Luthwanians [sic ], and Czechoslovakians were no good. One man was able to tell hundreds of adults that these people should be able to—should be exterminated and one man was able to get hundreds and hundreds of adults to put 21,000,000 people—12,000,000 people, twenty-one—some million people over, over six million into concentration camps and kill them and the rest of the world said impossible. It could not happen. Impossible until they saw the graves and we know what happened because we’ve seen the pictures.
If one man can do this with an entire world, what can one woman do in a little school with not adults, not adults but little, tiny, three, four and five-year old children?
Defense counsel objected to the attempt “to equate my client with Adolf Hitler.” The State responded that these were fair comments given the defendant’s opening. The prosecutor stated before the jury that, “The State is no way saying Miss Michaels is Hitler, in no way do we say that. The analogy is pointed out to show what one person can do.” The comment was allowed with the provision that the court would charge the jury “appropriately.”
Defense counsel in his opening had accused the State of “a witch hunt type mentality,” and asserted “[wjitch hunts, ladies and gentlemen, are built and predicated upon accusations so heinous as the ones that my client is being accused of, and ones that are irrevocable and she becomes branded just simply because she is *638accused.” He characterized the ease as resting upon “certain individuals who come forward and say, well, gosh, folks, we were just doing our job, but that sounds so strikingly similar and we hope that what took place at the [Nuremberg] trials ... in the Second World War will put that position to rest.”
Defendant’s opening theme was reiterated frequently in closing. The defense concluded that the “system” had victimized the children and “[w]e can’t let that happen.” Defendant’s summation reminded the jury that, as promised in its opening, the defense case included the “witch hunt issue.” Defendant’s summation asserted that the investigators got these children to say what the investigators wanted to hear; the State’s ease rested upon mere words of children who had those words planted in their heads by the State’s investigative team. Defendant’s opening characterized the prosecution’s single-minded pursuit of defendant as equivalent to those who offered, as a defense for Nazi genocide, the excuse that they were just doing their jobs.
The prosecution, in the presence of the jury, maintained it cited Hitler only to demonstrate the ability of a person to commit atrocities which the jury might find unimaginable. It openly disavowed any intent to equate defendant with Hitler. This noninflammatory reference to Hitler, in light of the totality of the circumstances, a six-day summation, and the State’s declaration that it did not seek to equate defendant with Hitler, cannot be reasonably viewed as conduct sufficient to warrant a new trial.
Defendant also objects to the State’s summation comments regarding defense expert, Dr. Underwager. We conclude that the State’s comments, when examined in the context they were made, were not unfairly prejudicial.
Defendant mistakenly asserts that the State called Dr. Under-wager a “witch doctor” because he expressed an opinion about whether these children were abused without ever personally examining the children. Rather, the State’s “witch doctor” characterization was aimed only at Dr. Underwager’s attempt to estab*639lish, by his “time and motion study,” how long it would have taken defendant to accomplish the abuse of which she was accused. The State was attacking him for offering time estimates even though he had never visited Wee Care or timed persons engaged in similar activities such as walking from one class to another, dressing or undressing children, or coordinating the activities of a group of young children.
The “witch doctor” reference was drawn from the expert’s own manuscript. He was asked on cross-examination:
Q: Doctor, are you saying that you used subjective experience in terms of how long it takes someone to get undressed?
A: Sure.
Q: And, Doctor, the scientist who relies on subjective experience to make an opinion, has regressed to the level of a witch doctor, isn’t it so?
A: That’s a totally different thing.
Q: Doctor, isn’t that what you wrote in your book, the opinion, “A doctor who bases his opinion on unchecked subjective experience has regressed to the level of a witch doctor.”?
A: That’s unchecked.
Q: Yes.
A: This is not unchecked.
The State’s further questioning demonstrated the prosecutor’s assertion that Dr. Underwager fit his own definition of a witch doctor.
Thus, the prosecutor’s comments during summation merely recalled the cross-examination of Dr. Underwager about whether he fit his own definition of a witch doctor.
Dr. Underwager has written in his manual, in his unpublished manual, that the scientist who bases an opinion upon unchecked subjective experience has regressed to the level of the witch doctor.
By his own definition, Dr. Underwager is a witch doctor because a scientist who based an opinion, as he did here, upon unchecked subjective experience is regressed to the level of the witch doctor.
And you know what? Dr. Underwager is no witch doctor because to call him so would be an insult- to witch doctors.
The use of the witch doctor phrase was incidental to the jury’s evaluation of Dr. Underwager’s credibility. The jury was urged to reject the doctor’s testimony because it was based upon faulty *640methods, even according to his own written statements about proper scientific technique. Additional references to the doctor’s lack of sincerity and good faith were not of such a nature as to require reversal. See State v. Koedatich, 112 N.J. 225, 323, 548 A.2d 939 (1988), cert. denied, 488 U.S. 1017, 109 S.Ct. 813, 102 L.Ed.2d 803 (1989).
Defendant also claims that the prosecutor’s assertion of the prosecution team’s belief in defendant’s guilt had the potential to affect the jury’s verdict. A prosecutor’s statement of personal belief in the guilt of a defendant is improper. Id. at 322, 548 A.2d 939; State v. Ramseur, 106 N.J. 123, 321, 524 A.2d 188 (1987). The expression of personal opinion places the prosecutor’s own credibility and the prestige of the office against that of defense witnesses and violates “the well-established rule that a prosecutor may not declare his personal belief in a defendant’s guilt.” Ibid. Our Supreme Court in Marshall, supra, 123 N.J. at 154, 586 A.2d 85, cited with approval the American Bar Association guideline which provides: “It is unprofessional conduct for the prosecutor to express his or her personal belief or opinion as to the truth or falsity of any testimony or evidence or the guilt of the defendant.” ASA Standards for Criminal Justice § 3-5.8(b) (2d ed. 1980).
Despite the expression of the prosecutorial belief in defendant’s guilt, the comment here would not require a new trial. See State v. Thornton, 38 N.J. 380, 398-99, 185 A.2d 9 (1962), cert. denied, 374 U.S. 816, 83 S.Ct. 1710, 10 L.Ed.2d 1039 (1963). In the context of the State’s lengthy summation only one instance of improper personal opinion has been identified on appeal. The absence of objection to that expression of opinion indicates that, in the context of the trial, defense counsel, as does this court, found that the comment had little potential for prejudice. State v. Wilson, 57 N.J. 39, 51, 269 A.2d 153 (1970); State v. Darrian, 255 N.J.Super. 435, 457-58, 605 A.2d 716 (App.Div.), certif. denied, 130 N.J. 13, 611 A.2d 651 (1992); see Ramseur, supra, 106 N.J. at 323, 524 A.2d 188.
*641Defendant next contends that the State’s display of the word “guilty” on a board during summation was prejudicial misconduct. The prosecutor referred to the jury’s task of sorting through nine months of trial evidence as “difficult to assimilate, to collate, to grasp all the elements of evidence in this case as overwhelming as that evidence was.” It was “[s]ort of like a puzzle. The pieces of a puzzle.” The prosecutors, therefore, illustrated the solution by assembling magnetic letters on the board to spell “guilty,” just as the evidence would spell guilty when properly assembled by the jury. Defense counsel objected and stated: “I would appreciate his removing the word guilty because, I think, it implies more than he’s entitled to have implied, especially over a prolonged period of time.” The court allowed the word to remain on the board.
A prosecutor may comment on the evidence in a vigorous and forceful presentation during summation and draw any reasonable inferences supported by the proofs. State v. Dixon, 125 N.J. 223, 259, 593 A.2d 266 (1991). It is for the jury to accept or reject those inferences and conclusions drawn therefrom. State v. Carter, 91 N.J. 86, 125, 449 A.2d 1280 (1982). It was not improper to use a puzzle analogy to argue that defendant was guilty. There is no basis on which to conclude that placing the word “guilty” on a board had either an immediate impact upon the jurors, or a “subliminal” influence as suggested by defendant. The State was free to contend that the evidence proved defendant guilty as charged. Defendant’s contention is clearly without merit.
VII.
Defendant asserts that it was error to grant the jury’s request to review the videotaped trial testimony of the children during jury deliberations. Despite our earlier determination and observations regarding the use of CCTV and the content of the tapes themselves, we choose to make further observations on the issue as there are presently no New Jersey cases on point.
*642Defendant argued during trial that it would be grossly unfair to replay the CCTV testimony since any request for repetition of defendant’s testimony would be satisfied by the much less dramatic form of reading her testimony from a transcript. Therefore, defendant contended it was essential to a fair trial, since her testimony could not be replayed on videotape, that all the testimony should be read back from a transcript. The trial court issued an order allowing replay of the CCTV testimony upon the jury’s request. Defendant sought relief from this court prior to summations. We denied leave to appeal.
Soon after retiring to deliberate, the jury requested a review of the CCTV testimony of-each child. The jury wanted to deliberate after reviewing the videotape and other evidence as to each child before proceeding to view the videotape and evidence concerning the next child. Defendant again argued that only a reading of the children’s testimony would be fair and that replay of the CCTV tapes would prejudice her right to a fair trial. The court again ruled that the jury’s request for the taped testimony was the same as a request for a reading of testimony which had only been recorded stenographically. Accordingly, the tape of each child’s testimony was played in its entirety in the courtroom with all jurors present; after each replay, the jury retired to deliberate before viewing the next tape. The jury never asked for a reading of defendant’s testimony. Defendant, in her motion for a new trial, again argued that the video playback of the children’s testimony during deliberations was unfair and prejudicial.
The statute authorizing children to give CCTV testimony requires that a transcript be made and included as part of the appellate record. The statute provides:
[A] stenographic recording of that [CCTV] testimony shall also be required. A typewritten transcript of that testimony shall be included in the record on appeal. The closed circuit testimony itself shall not constitute part of the record on appeal except on motion for good cause shown. [N.J.S.A. 2A:84A-32.4e.]
The Legislature clearly contemplated that a videotape would be made when CCTV testimony was given, but it provided that the videotaped testimony could only be part of the record upon motion *643and for good cause shown. We do not read this to preclude replaying the videotape at the jury’s request. The matter, therefore, is subject to judicial determination.
Defendant directs our attention to two cases in which video replay of children’s testimony during jury deliberation was found to be reversible error. The Ninth Circuit determined that allowing the jury to have a videotape of children’s testimony, with equipment for replaying it in the jury room, was an abuse of discretion. United States v. Binder, 769 F.2d 595, 601 (9th Cir.1985). The court opined that permitting the jury unrestricted access to the videotape during deliberations “was equivalent to allowing a live witness to testify a second time in the jury room.” Id. at 601 n. 1. The court held:
Videotape testimony is unique. It enables the jury to observe the demeanor and to hear the testimony of the witness. It serves as the functional equivalent of a live witness. Since there was no physical evidence, the only evidence of acts of molestation was presented through the children’s videotaped testimony. The defendant denied any criminal conduct, asserting that the children were displeased with him and their charges against him were vindictive. Credibility became a crucial issue. Under these circumstances the videotaped testimony may have taken on great significance. Allowing the jury to see and hear the children’s videotaped testimony a second time in the jury room during deliberations unduly emphasized their testimony. [Id. at 600-01 (footnote omitted).]
The majority in Binder felt that, if the children’s testimony was to be revisited during deliberations, it should have been by a reading of a written transcript. In no event should the video have been replayed. Id. at 601 n. 1.
Defendant also cites Martin v. State, 747 P.2d 316, 319-20 (Okla.Crim.App.1987), which held it was reversible error to allow the jury in a sexual-abuse case unrestricted access to children’s videotaped testimony during deliberations. However, the Martin court also held that a replay of the taped testimony in open court would, in the court’s discretion, be permissible at the jury’s request. Id. at 320.
We agree with the line of cases holding that it is error to allow the jury to have videotaped testimony and a means of playing it in the jury room. See Binder, supra, 769 F.2d at 601; *644 State v. Kraushaar, 470 N.W.2d 509, 516 (Minn.1991) (“preferable” for review to take place in courtroom); Taylor v. State, 727 P.2d 274, 276 (Wyo.1986); People v. Talley, 824 P.2d 65, 67 (Colo.Ct.App.1991) (audiotape); Martin, supra, 747 P.2d at 320. However, we refuse to hold that it is never permissible, at a jury’s request during deliberations, to replay a videotape of testimony in its entirety for the jury, with the defendant present, in open court. See United States v. Sacco, 869 F.2d 499, 502-03 (9th Cir.1989); United States v. McKinney, 822 F.2d 946, 951 (10th Cir.1987); Kraushaar, supra, 470 N.W.2d at 516; State v. Jennings, 815 S.W.2d 434, 440 (Mo.Ct.App.1991); Pfaff v. State, 830 P.2d 193, 195 (Okla.Crim.App.1992).
It is clear that videotaped testimony provides more than conventional, transcribed testimony. The witness’ actual image, available in a video replay, presents much more information than does a transcript reading. In essence, the witness is brought before the jury a second time, after completion of the defense ease, to repeat exactly what was testified to in the State’s case. The witness’ words and all of the animation, passion, or sympathy originally conveyed are again presented to the jury. It is difficult to deny that there is an advantage that may be gained in such circumstances. However, we cannot say that the replay of child-testimonial videotapes is prejudicial per se or that because of the impact of the visual image, the trial judge should be divested of discretion to accede to a jury’s request for a replay. A court exercises wide discretion in granting or denying a jury’s request to rehear testimony during deliberations. State v. Wilkerson, 60 N.J. 452, 460, 291 A.2d 8 (1972); State v. Wolf, 44 N.J. 176, 185, 207 A.2d 670 (1965); State v. DeBellis, 174 N.J.Super. 195, 201, 413 A.2d 986 (App.Div.1980).
We would, however, caution against routine replaying of such testimony. A trial judge should first seek to satisfy a jury request for playback of videotaped testimony by offering a reading of the transcript of the testimony. The trial judge should inquire of the jury as to whether there is something the jurors are seeking *645from the videotape which would be unavailable to them from an impartial reading of the witness’ testimony. If it is determined that the jury’s request for a replay of the tape appears reasonably necessary to its deliberations, then the trial judge should exercise discretion to balance that need against any possible prejudice in each particular case. See Evid.R. 4.
CONCLUSION
The prejudicial errors discussed in sections I and III, supra, require reversal of defendant’s convictions. We remand to the Law Division for further proceedings.