The opinion of the Court was delivered by
GARIBALDI, J.In October 1994, the Legislature enacted a group of eleven bills, collectively known as “Megan’s Law,” in memory of a seven-year-old girl allegedly killed by a convicted sexual offender. In Doe v. Poritz, 142 N.J. 1, 109, 662 A.2d 367 (1995), we upheld the constitutionality of two of those bills, the Registration Law that *80requires certain sexual offenders to register with law enforcement, N.J.S.A. 2C:7-1 to -5, and the Community Notification Law, N.J.S.A. 2C:7-6 to -11 (collectively referred to as RCNL). In this appeal, we further interpret and apply those laws. Specifically, we consider whether a nonconviction offense may be considered in determining a convicted sex offender’s tier classification under the RCNL, and whether the prosecutor’s use of documentary hearsay evidence to prove that alleged offense offends procedural due process and the doctrine of fundamental fairness. We also address issues, not directly raised in this appeal, that relate to the validity of the Registrant Risk Assessment Scale that the Court sua sponte requested the parties and amici curiae to address.
I
We explained the purpose behind Megan’s Law in detail in Doe v. Poritz, supra, 142 N.J. at 12-20, 662 A.2d 367. The Legislature enacted Megan’s Law to protect the community from the dangers of recidivism by sexual offenders. N.J.S.A. 2C:7-1a. The Legislature also wanted to provide law enforcement agencies with additional information about sex offenders in their community because that information is “critical to preventing and promptly resolving incidents involving sexual abuse and missing persons.” N.J.S.A. 2C:7-1b. The statute provides that sex offenders shall register with local authorities. N.J.S.A. 2C:7-2a. The sex offenses that trigger required registration are set forth in N.J.S.A. 2C:7-2b.
After registration by the offender, the chief law enforcement officer in the municipality .where a registrant intends to reside is required to provide notification in accordance with N.J.S.A. 2C:7-8. That statute contemplates three levels of notification “depending upon the degree of the risk of re-offense.” N.J.S.A. 2C:7-8a.
(1) If risk of re-offense is low, law enforcement agencies likely to encounter the person registered shall be notified;
*81(2) If risk of re-offense is moderate, organizations in the community including schools, religious and youth organizations shall be notified in accordance with the Attorney General’s guidelines, in addition to the notice required by paragraph (1) of this subsection;
(3) If risk of re-offense is high, the public shall be notified through means in accordance with the Attorney General’s guidelines designed to reach members of the public likely to encounter the person registered, in addition to the notice required by paragraphs (1) and (2) of this subsection.
[N.J.S.A. 2C:7-8c.]
All registrants are subjected to at least Tier One notification, which requires registration with law enforcement agencies. Doe v. Poritz, supra, 142 N.J. at 22, 662 A.2d 367.
N.J.S.A. 2C:7-8a requires the Attorney General, after consultations with members of the advisory council, to “promulgate guidelines and procedures for the notification required pursuant to the provisions of this act. The guidelines shall identify factors relevant to risk of re-offense and shall provide for three levels of notification depending upon the degree of the risk of re-offense.” The Legislature instructed the Attorney General that
Factors relevant to risk of re-offense shall include, but not be limited to, the following:
(1) Conditions of release that minimize risk of re-offense, including but not limited to whether the offender is under supervision of probation or parole; receiving counseling, therapy or treatment; or residing in a home situation that provides guidance and supervision;
(2) Physical conditions that minimize risk of re-offense, including but not limited to advanced age or debilitating illness;
(3) Criminal history factors indicative of high risk of re-offense, including:
(a) Whether the offender’s conduct was found to be characterized by repetitive and compulsive behavior;
(b) WTiether the offender served the maximum term;
(c) WTiether the offender committed the sex offense against a child.
(4) Other criminal history factors to be considered in determining risk, including:
(a) The relationship between the offender and the victim;
*82(b) Whether the offense involved the use of a weapon, violence, or infliction of serious bodily injury;
(c) The number, date and nature of prior offenses;
(5) Whether psychological or psychiatric profiles indicate a risk of recidivism;
(6) The offender’s response to treatment;
(7) Recent behavior, including behavior while confined or while under supervision in the community as well as behavior in the community following service of sentence; and
(8) Recent threats against persons or expressions of intent to commit additional crimes.
[N.J.S.A. 2C:7-8b.]
Pursuant to that delegation of power, the Attorney General issued guidelines to determine the risk of re-offense and appropriate tier level of notification. The Attorney General’s guidelines currently provide that a registrant is rated by the local county prosecutor pursuant to the guidelines and the “Registrant Risk Assessment Seale” (Scale). To develop that Scale, the Attorney General convened a committee composed of mental health experts as well as members of the Law Enforcement Committee, which drafted the Scale and the accompanying Registrant Risk Assessment Manual (RRA Manual),1 which explains the Scale. The Seale focuses on two factors, “the seriousness of the offense should the offender recidivate,” and “the likelihood that the offender will recidivate.” Registrant Risk Assessment Manual at 2.
The Scale, attached hereto as Appendix A, identifies thirteen factors, including the statutory factors as well as other factors deemed relevant to the risk of re-offense. Those factors are assigned to one of four larger categories: “Seriousness of Offense,” “Offense History,” “Characteristics of Offender,” and “Community Support.” Each registrant is assigned a score for each of the thirteen factors depending on an objective appraisal of that registrant, with a zero for low risk, one for moderate risk, and three for high risk.
*83The Attorney General determined that “the historical factors— particularly those related to extensiveness of antisocial behavior— tend to be the most powerful predictors of future offenses.” Id. at 4, 662 A.2d 367. As a result, the Scale gives greater weight to the “Seriousness of Offense” and “Offense History” factors than to the “Characteristics of Offender” and “Community Support” factors. After adding up the total (adjusted) score for a registrant on every factor, a registrant who receives a score of at least 74 (out of a maximum of 111) is considered Tier Three, while one who scores at least 37 is considered Tier Two, and below 37 is Tier One.
II
In Doe v. Poritz, supra, 142 N.J. at 107, 662 A.2d 367, we concluded that a hearing is required prior to notification under Tiers Two and Three. The Court, recognizing that the RCNL impinges upon interests that trigger the protections of procedural due process and the State’s own fairness doctrine, required judicial review of those decisions to “assure that the risk of reoffense and the extent of notification are fairly evaluated before Tier Two [moderate risk] or Tier Three [high risk] is implemented.” Id. at 30, 662 A.2d 367. The RCNL judicial hearing process under Doe v. Poritz, supra, is not governed by the rules of evidence. The reviewing court may rely exclusively on documentary evidence “on all issues.” Id. at 31, 662 A.2d 367. The court has the authority to determine (1) the extent of witness production; (2) the extent of cross-examination; and (3) the use of expert testimony. Ibid. Those determinations are to be based on the “apparent complexity of the matter [and] the extent of doubt concerning the correctness of the level and manner of notification selected by the prosecutor.” Ibid. At the same time, Doe v. Poritz further prescribed a two-step procedure for evidence production. In the first step, the prosecutor has the burden of going forward with prima facie evidence that “justifies the proposed level and manner of notification.” Id. at 32, 662 A.2d 367. In the second step, assuming the *84prosecutor’s burden is met, the registrant then has the burden of producing evidence challenging the prosecutor’s determinations on both issues. Ibid. Once the State has satisfied its burden of going forward, the court “shall affirm the prosecutor’s determination unless it is persuaded by a preponderance of the evidence that it does not conform to the laws and Guidelines.” Ibid. The court’s determination is independent and based on its own review of the case on the merits.
Ill
On July 22, 1988, C.A.2 entered the room of his sleeping victim, a friend’s girlfriend whom he had known for several months; when she awoke, he was digitally penetrating her and fondling her breasts. He pled guilty to fourth degree sexual assault and burglary and was sentenced on June 14, 1991 to a two year probationary term.
Two days later, on June 16, A.Z., “a middle class resident of the suburbs ... driving in a drug area late at night,” called the police from a telephone booth to report that she had been raped by C.A. In re Registrant C.A, 285 N.J.Super. 343, 350, 666 A.2d 1375 (App.Div.1995). When the police arrived, they searched the area and found C.A. hiding behind some wood; he was arrested after a brief struggle.
A.Z. told the patrol officer at the scene that
She was in her listed auto stopped at a red light on Johnston at Greenwood facing [sic] north.
At that time, [C.A.] opened her unlocked passenger door, got in, pulled out a small hunting type knife, held it up in front of him and told her that she was going to take him where he wanted to go. They drove north on Johnston ... [C.A] had *85her park in an alley then told her to get into the back seat and take off her pants, underwear and pull out her tampon. At no time did he threaten her or show her the knife. He then got in the back seat and had ... vaginal intercourse____
[C.A.] then choked her about the neck with both his hands. He let her go and told her to give him a 18" gold chain she wore around her neck. She gave it to him. He then got into the driver’s seat and drove to [another street] while she sat in the back. He parked the car, took the keys and told her not to move.
She told another detective the same story. She was eventually admitted to a hospital' and treated for depression and post-traumatic stress disorder. She told the doctors the same story that she had told police; she also told the doctors that she had' previously been raped eleven years earlier.
C.A. told police that he did not rape the victim but had known her for some time because she had often come to purchase drugs from him. He claimed that A.Z. had asked him to procure drugs, that he had driven her in her car to find drugs, and that they smoked the drugs and engaged in consensual sex. She gave him the gold chain to purchase additional crack, but, when he returned with the drugs, he hid because he saw the heavy police presence. C.A. also denied possession of a knife, and the police did not recover a knife from the scene.
When the police discussed this version with A.Z., she agreed to take a polygraph test. However, she never appeared for the test. Later, at the hospital, she admitted drug use since age sixteen (including the use of crack cocaine that she had purchased on the street.)
Based on A.Z.’s complaint, the State obtained an indictment charging C.A. with sexual assault, robbery, and weapons offenses.
On January 30, 1992, another woman called the police to report a rape committed by C.A. She told the police that C.A., whom she had seen before but did not know, had stopped her as she walked on the street and forced her into an alley. He took her money and demanded that she remove her clothing; he then had oral and vaginal intercourse with her after threatening to kill her. C.A. claimed, as he did with A.Z., that he was friendly with the victim and had engaged in consensual sex in exchange for drugs. The *86victim acknowledged prior drug use but insisted that she was raped. He was charged with aggravated sexual contact.
Pursuant to a plea agreement, C.A. pled guilty to third degree aggravated criminal sexual contact for the January 30 incident. Under that plea agreement, the earlier indictment charging him with sexual assault, robbery, and weapons offenses involving the victim A.Z. was dismissed. C.A. was sentenced to a five-year custodial sentence with a two and one-half year parole disqualifier.
In October 1995, as C.A. was preparing for his release from prison, the State notified him that, because he had received a score of 83 under the Seale, he would be classified in Tier Three. C.A. argued that his Tier Three classification was wrong and he should have been classified in Tier Two. Specifically, C.A. disputed the prosecutor’s scoring under factor one, “degree of force” (15 points); factor four, “victim selection” (9 points); factor five, “number of offenses/victims” (9 points); and factor nine, “response to treatment” (6 points).
C.A. requested a judicial hearing on his Tier Three classification. He alleged substantive and procedural deficiencies. First, he contended that the dismissed charge involving A.Z. was not an offense for purposes of coverage under the RCNL and could not be counted in computing his Scale score. Second, he contended that, even if the charge could be counted, A.Z.’s assertions were unreliable because of her history of alcohol and drug abuse and her admission that she had previously lied to hospital personnel. C.A. also denied that a knife was used, claiming that the incident was a consensual sex for drugs transaction. Thus, he claimed that the scoring was incorrect because he never used a weapon, because he was acquainted with all of his victims, and because, excluding A.Z., he only had two victims. C.A. further contended that the State’s documentary hearsay was not credible and requested an evidentiary hearing in which the victim, A.Z., and the State’s other hearsay declarants would testify. If successful in excluding his alleged offense against A.Z., C.A.’s score would have been reduced from 83 to below 73, moving him into Tier Two.
*87The prosecutor argued that the Scale specifically permits the consideration of documentation other than a criminal conviction. He relied on the RRA Manual, which provides that the scoring
may take into account any information available and encompass all credible evidence. Thus, a determination of the number of victims or offenses may be based upon documentation other than a criminal conviction. Such documentation may include, but is not limited to, criminal complaints not the subject of a conviction but which are supported by credible evidence, victim statements, admissions by the registrant, police reports, medical, psychological or psychiatric reports, pre-sentencing reports, and Department of Corrections discharge summaries.
[Registrant Risk Assessment Scale Manual at 5.]
The trial court found that the noneonvietion offense could be counted and further found that a hearing was not required. The court decided that the documentary evidence submitted was sufficiently reliable to establish that the incident occurred, especially “three separate statements to three separate persons ... that while she was in an emotional state, she gave information each of the three times that a knife was used in that incident.”
The Appellate Division reversed and remanded. In re C.A., supra, 285 N.J.Super. 343, 666 A.2d 1375. The panel agreed with the trial court that nonconviction offenses could be counted, and that documentary evidence could be used to evaluate nonconviction offenses. Id. at 347-48, 666 A.2d 1375. The court further observed that, because this was not a criminal ease, the Sixth Amendment Right to Confrontation did not apply. Id. at 348, 666 A.2d 1375. However, the panel decided that documentary evidence should only be admitted into evidence when adjudged reliable based on the totality of the circumstances surrounding the statements being considered. Ibid. If the State established a prima facie case based on those documents justifying the proposed level of notification, then the offender could rebut the claim; if there was a direct factual conflict, then a hearing would be appropriate. Id. at 349-51, 666 A.2d 1375. While providing for a hearing, the Appellate Division held that neither side could serve process upon the victim without leave of the court, “which shall be granted only upon a clear and convincing demonstration of a compelling need for that [witness’s] testimony.” Id. at 350, 666 *88A.2d 1375. Because the plausible story told by C.A. created a direct factual conflict, a hearing was required. Ibid.
C.A. filed a petition for certification, arguing that nonconviction offenses could not be counted and that, even if they could be counted, documentary hearsay evidence should not be sufficient to establish such incidents under his right to procedural due process and fundamental fairness. We granted C.A.’s petition for certification. 143 N.J. 328, 670 A.2d 1068 (1996).
IV
A
C.A. contends that the trial court’s inclusion of an alleged offense for which C.A. was never convicted was improper, and that only convictions should be counted in evaluating the risk of re-offense and determining the appropriate tier of notification. As discussed supra at 87, 679 A.2d at 1160, the RRA Manual specifically includes nonconviction offenses. Thus, the question is whether the RRA Manual’s inclusion of nonconviction offenses is authorized under the statute.
N.J.S.A. 2C:7-8a delegates the legislative power to identify factors “relevant to risk of re-offense” to the Attorney General and an advisory council. That statute enumerates certain factors relevant to risk of re-offense that “shall [be] include[d]” in the guidelines, but provides that the guidelines need “not be limited to” the enumerated factors. N.J.S.A. 2C:7-8b. The question presented is whether nonconviction offenses are one of the enumerated factors, and, if they are not, whether the inclusion of such offenses is an appropriate exercise of the Attorney General’s delegated power.
The notification statute does not list nonconvietion offenses as a separate enumerated factor. N.J.S.A. 2C:2-7-8b(1)-(8). However, the Legislature has identified those “criminal history factors” that it found “indicative of high risk of re-offense.” N.J.S.A. 2C:7-8b(3). They include:
*89(a) whether the offender’s conduct was characterized by repetitive and compulsive behavior;
(b) whether the offender served the maximum term;
(c) whether the offender committed the sex offense against a child.
[N.J.S.A. 2C:7-8b(3)(a) to -(c).]
“Other criminal history” factors identified by the statute to be considered in determining risk include:
(a) The relationship between the offender and the victim;
(b) Whether the offense involved the use of a weapon, violence or infliction of serious bodily injury;
(c) The number, date and nature of prior offenses.
[N.J.S.A. 2C:7-8b(4)(a) to -(c).]
Based on those statutory provisions, we find that the Legislature intended nonconviction offenses to be part of the statutory factors of “criminal history” and “other criminal history”. Therefore, nonconviction offenses are to be considered in evaluating a registrant’s risk of re-offense, provided there is sufficient evidence that the offense occurred.
Further, we observe that, even if nonconviction offenses were not deemed to be part of the enumerated statutory factors, the inclusion of such offenses in the Scale was still an appropriate exercise of the Attorney General’s delegated power. As an initial matter, the statutory delegation of power to the Attorney General to identify other factors relevant to risk of re-offense is constitutionally proper. “[T]he Guidelines were prepared in response to a specific statutory mandate and their contents are largely dictated either explicitly or implicitly by the language of the statute.” Doe v. Poritz, supra, 142 N.J. at 98-99, 662 A.2d 367. In reviewing the Attorney General’s decision to include certain non-statutory factors pursuant to this delegation, we “will not ... set [it] aside on the ground that it transgressed the statute unless the transgression is plain; the presumption is in favor of validity.” New Jersey Guild of Hearing Aid Dispensers v. Long, 75 N.J. 544, 561, 384 A.2d 795 (1978)(quoting Lane v. Holderman, 40 N.J.Super. 329, 335, 123 A.2d 56 (App.Div.1956), aff'd, 23 N.J. 304, 129 A.2d 8 (1957)).
*90As described infra at 105, 679 A.2d at 1170, experts generally agree that the best predictor of a registrant’s future criminal sexual behavior is that registrant’s prior criminal record. Accordingly, prior nonconviction offenses should be considered in the risk calculation provided that there is sufficient rehable evidence that the offense did happen.
Others have recognized, in different contexts, that noneonviction offenses are relevant in determining the risk of re-offense or recidivism. In United States v. Dunnigan, 507 U.S. 87, 97, 113 S.Ct. 1111, 1118, 122 L. Ed.2d 445, 455 (1993), the United States Supreme Court approved a sentence enhancement based on the defendant’s noneonviction offense of perjury by reasoning that
[i]t is rational for a sentencing authority to conclude that a defendant who commits a crime and then peijures herself in an unlawful attempt to avoid responsibility is more threatening to society and less deserving of leniency than a defendant who does not so defy the trial process. The perjured defendant’s willingness to frustrate judicial proceedings to avoid criminal liability suggests the need for incapacitation and retribution is heightened____
Similarly, others have argued that, “if incapacitation or reform is at issue, sentencing authorities must attempt predictors of an offender’s future behavior. In such systems, ... the best results depend upon an opportunity to take the defendant’s ‘whole life’ into view, especially significant events such as past crimes— whether or not they have ever ripened into convictions.” Kevin R. Reitz, Sentencing Facts: Travesties of Real-Offense Sentencing, 45 Stan. L.Rev. 523, 553 (1993)(rejecting that argument); accord State v. Green, 62 N.J. 547, 571, 303 A.2d 312 (1973)(permitting sentencing court to review defendant’s arrest record because “the sentencing judge might find it significant that a defendant who experienced an unwarranted arrest was not deterred by that fact from committing a crime thereafter”). See also Leonard N. Sosnov, Due Process Limits on Sentencing Power: A Critique of Pennsylvania’s Imposition of a Recidivist Mandatory Sentence without a Prior Conviction, 32 Duq.L.Rev. 461, 506 (1994)(justify-ing enhanced sentence based upon prior arrest resulting in pretrial intervention because “individuals who have previously accept*91ed [PTI] ... may present a somewhat higher risk of another incidence of DUI after conviction”).
Numerous state and federal courts have sustained civil remedies based on nonconvietion offenses. We held in Doe v. Poritz, supra, 142 N.J. at 40-74, 662 A.2d 367, that notification pursuant to Megan’s Law is not punishment for a criminal action but rather is a civil remedy to ensure public safety. See, e.g., In re McLaughlin, 144 N.J. 133, 143, 675 A.2d 1101 (1996)(denying application for admission to bar in part on the basis of prior act that was “illegal,” even though there was no conviction for that prior act); see also United States v. Ursery, _ U.S. _, _, _, 116 S.Ct. 2135, 2139, 2141, 135 L.Ed.2d 549, 557, 560 (1996)(approving of civil forfeiture for narcotics violation although conviction was not obtained until after civil proceedings were completed).
We have approved of civil penalties for conduct when the individual was acquitted on charges of committing that offense. Thus, in In re Pennica, 36 N.J. 401, 418, 177 A.2d 721 (1962), we held that “acquittal of a member of the bar following trial of a criminal indictment is not res judicata in a subsequent disciplinary proceeding based on substantially the same charge or conduct.” Similarly, in In re Darcy, 114 N.J.Super. 454, 458, 277 A.2d 226 (App.Div.1971), the Appellate Division allowed civil disciplinary proceedings against a civil servant despite a prior acquittal on those same charges. See also New Jersey Division of Youth and Family Services v. V.K., 236 N.J.Super. 243, 252, 565 A.2d 706 (App.Div.1989) (sustaining civil decision terminating parental rights due to abuse, despite parent’s prior acquittal on same charges), certif. denied, 121 N.J. 614, 583 A.2d 315, cert. denied, 495 U.S. 934, 110 S.Ct. 2178,109 L. Ed.2d 507 (1990); Township of East Hanover v. Cuva, 156 N.J.Super. 159, 163, 383 A.2d 725 (App.Div.1978) (approving civil injunction for violation of township ordinance despite prior acquittal on same charge); Kugler v. Banner Pontiac-Buick, Opel, Inc., 120 N.J.Super. 572, 579-80, 295 A.2d 385 (Ch.Div.1972) (allowing civil penalty for violation of Consumer Fraud Act despite acquittal on same charge); Freu*92denreich v. Mayor and Council of Borough of Fairview, 114 N.J.L. 290, 292-93, 176 A. 162 (E & A 1934) (permitting disciplinary action for police officer despite prior acquittal); cf. Helvering v. Mitchell, 303 U.S. 391, 397-99, 58 S.Ct. 630, 632-33, 82 L.Ed. 917, 920-22 (1938) (sustaining civil suit to recover tax deficiency despite prior acquittal).
Federal courts have enhanced criminal sentences based on nonconviction offenses whether the result of an acquittal, e.g., United States v. Manor, 936 F.2d 1238, 1243 (11th Cir.1991)(“an acquittal based on a reasonable doubt standard should not preclude a contrary finding [at sentencing] using the preponderance of the evidence standard”); cf. State v. Megargel, 278 N.J.Super. 557, 568, 651 A.2d 1051 (App.Div.1995)(sustaining enhanced Graves Act sentence for weapons possession despite acquittal on weapons possession charge because, inter alia, “the judge’s findings were based on a lower standard of proof’), rev’d on other grounds, 143 N.J. 484, 673 A.2d 259 (1996), a dismissed count pursuant to a plea bargain, e.g., United States v. Camuti, 950 F.2d 72, 74 (1st Cir.1991)(considering conduct involved in two counts dropped as result of plea bargain to enhance sentence), or conduct never charged, e.g., United States v. Andrews, 948 F.2d 448, 448-450 (8th Cir.1991)(considering defendant’s involvement in prior uncharged robberies to enhance sentence on conviction for robbery).
“Nearly all jurisdictions allow courts to consider such noneonviction charges when sentencing.” Reitz, supra, 45 Stan. L.Rev. at 534. One commentator has noted, “[constitutional challenges to [enhanced] punishment for nonconviction offenses have met with consistent rebuffs in the federal courts.” Elizabeth T. Lear, Is Conviction Irrelevant?, 40 U.C.L.A. L.Rev. 1179, 1183 (1993)(arguing that courts should not be allowed to consider noneonviction offenses).
The Scale’s inclusion of nonconviction offenses as relevant to the risk of re-offense accords with the views of those courts and commentators and is statutorily authorized and a rational imple*93mentation of the Attorney General’s delegated power pursuant to N.J.S.A. 2C:7-8. We therefore sustain it.
B.
If nonconviction offenses may be used, how should their existence be established? C.A. contends that a factual hearing must always be held in order to establish that such an offense did in fact occur, and that the use of only hearsay and documentary evidence at such a hearing deprives him of his right to procedural due process and fundamental fairness.
We have previously concluded that notification implicates “protectible liberty interests in privacy and reputation,” and therefore triggers the constitutional right to procedural due process. Doe v. Poritz, supra, 142 N.J. at 106, 662 A.2d 367. We have also held that New Jersey’s doctrine of fundamental fairness, which “require[s] procedures to protect the rights of defendants at various stages of the criminal justice process even when such procedures were not constitutionally compelled,” applies to notification hearings “to require procedural protections that will ensure that his classification ... [is] tailored to his particular characteristics and are not the product of arbitrary action.” Id. at 108-09, 662 A.2d 367.
The precise protections needed to ensure due process depend on a careful balancing of three factors known as the Mathews v. Eldridge test:
first, the private interest that will be affected by the official action; second, the risk of an erroneous deprivation of such interest through the procedures used, and the probable value, if any, of additional or substitute procedural safeguards; and finally, the Government’s interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedural requirement would entail.
[Doe v. Poritz, supra, 142 N.J. at 107, 662 A.2d 367 (quoting Zinermon v. Burch, 494 U.S. 113, 127, 110 S.Ct. 975, 984, 108 L. Ed.24 100, 115 (1990) (quoting Mathews v. Eldridge, 424 U.S. 319, 335, 96 S.Ct. 893, 903, 47 L. Ed.2d 18, 33 (1976))).]
*94The minimum requirements of due process are notice and an opportunity to be heard. Id. at 106, 662 A.2d 367. Applying the Mathews v. Eldridge test in Doe v. Poritz, supra, we added additional procedures and concluded:
We attempt by these procedures to reach a difficult accommodation between the State’s legitimate and substantial interest in effecting prompt notification and the offender’s legitimate interest in assuring accurate evaluation of the risk of re-offense and the proper determination of the manner of notification.
[Id. at 32, 662 A.2d 367.]
The Court vested reviewing courts with the obligation of providing procedural due process to ensure the appropriateness of a tier classification. Id. at 39, 662 A.2d 367. That obligation carries with it broad powers to control the summary hearing on the level and manner of the tier notification. Id. at 31, 662 A.2d 367. According to Doe v. Poritz, judges are to structure the hearing based “on the apparent complexity of the case [and] the extent of doubt concerning the correctness of the level and manner of notification selected by the prosecutor.” Id. at 31, 662 A.2d 367.
The hearing is not like an administrative hearing but is more like an evidentiary and investigatory hearing. It is civil, not criminal, and remedial, not adversarial. A fine balance must be drawn between the registrant’s rights to due process and fundamental fairness and the community’s right of protection against the registrant’s risk of re-offense, but the balance to be drawn is not the same balance as drawn in a criminal proceeding. In recognition of the countervailing governmental interests, Doe v. Poritz, supra, did not grant registrants the full panoply of rights applicable to a criminal proceeding. Id. at 34, 662 A.2d 367. Instead, we expressly held that the “rules of evidence shall not apply.” Ibid. The Court further held that the reviewing court may “rely on documentary presentations on all issues.” Ibid. In recognizing the need for procedural safeguards, the Doe Court did not analyze the circumstances under which admission of hearsay evidence is permissible, but simply authorized the trial court to
*95determine whether witnesses will be produced and cross-examinations allowed. Id. at 31, 662 A.2d 367.
Although it is undisputed that the framework contemplated by Doe v. Poritz permits the State’s use of hearsay evidence, nevertheless, to protect the registrant’s constitutional rights, “relaxed standards for admissibility are not to be equated with automatic admissibility,” State v. Davis, 96 N.J. 611, 623, 477 A.2d 308 (1984). Our rules of evidence insist that only statements subject to cross-examination, or other statements where “circumstantial guarantees of trustworthiness” exist, should be admitted as evidence. 2 McCormick on Evidence § 253, at 130 (citation omitted) (4th ed. 1992); see also Idaho v. Wright, 497 U.S. 805, 820, 110 S.Ct. 3139, 3149, 111 L. Ed.2d 638, 655 (1990) (explaining that “if the declarant’s truthfulness is so clear from the surrounding circumstances that the test of cross-examination would be of marginal utility, then the hearsay rule does not bar admission.”) C.A. contends that, because hearsay is so unreliable, there is an increased risk of error and the additional safeguard of requiring admissible evidence would increase accuracy and should be required.
Because the main difficulty with hearsay is reliability, hearsay that is reliable, even though not deemed sufficiently reliable to be admitted under our rules of evidence, should be admissible and sufficient to allow the State to sustain its burden of presenting a prima facie case. If the hearsay is reliable, then the risk of error is low and the utility of additional safeguards (i.e., excluding hearsay because unreliable) is low and does not outweigh the important government interests mitigating in favor of the admission of hearsay. Admission of reliable hearsay is compatible with C.A.’s right to procedural due process and fundamental fairness. In the future, when prosecutors plea-bargain sex offenses that may later form the basis for tier classification under the RCNL, they should take care to ensure that the factual basis of the sexual offenses that are dropped pursuant to the plea bargain are established by reliable evidence.
*96We expect that trial courts will carefully evaluate the proffered hearsay to ensure that the hearsay is reliable under the totality of circumstances surrounding that statement. In re C.A., supra, 285 N.J.Super. at 348, 666 A.2d 1375 (citing Idaho v. Wright, supra, 497 U.S. at 819, 110 S.Ct. at 3148, 111 L. Ed.2d at 654-55). Based on the circumstances surrounding the statement and any corroborating evidence, trial courts should determine whether the hearsay is sufficiently reliable to accept it as substantive evidence. A determination about the reliability of hearsay evidence should be made on a case-by-case basis. The determination about use of an anonymous phone tip to police will certainly be different than the current case, where A.Z. voluntarily identified herself and provided a report immediately after the incident that was consistent with her other statements. Each case will depend on its facts, and, in order to protect the constitutional rights of all registrants, only hearsay evidence that appears reliable should be admitted.
Other courts have approved of the use of hearsay in somewhat analogous situations. For example, in Zannino v. Arnold, 531 F.2d 687, 692 (3d Cir.1976), the Third Circuit approved of the use of hearsay by a parole board to find that a prisoner had been part of a large-scale conspiracy and not worthy of parole. “Though hearsay, they could legitimately be taken into account by the Parole Board.” Ibid. Similarly, “the admission and consideration of reliable hearsay evidence in probation violation proceedings is both fair and practical.” State v. Reyes, 207 N.J.Super. 126, 139, 504 A.2d 43 (App.Div.)(emphasis added), certif. denied, 103 N.J. 499, 511 A.2d 671 (1986). Likewise, N.J.S.A 52:14B-10 provides that our administrative agencies may admit hearsay evidence, even though many administrative cases involve the deprivation of liberty or property, e.g. In re Polk, 90 N.J. 550, 562-63, 449 A.2d 7 (1982), and implicate the constitutional right to due process.
C.
Once the State has proven its prima facie case, the burden of proof and persuasion shifts to the registrant, and the *97court “shall affirm the prosecutor’s determination unless it is persuaded by a preponderance of the evidence that it does not conform to the laws and Guidelines.” Doe v. Poritz, supra, 142 N.J. at 32, 662 A.2d 367. If the defendant produces proof, whether in the form of reliable hearsay, affidavit, or an offer of live testimony, that is sufficient to raise a “genuine issue of material fact,” that the tier classification and the manner of notification are inappropriate, then the trial court should convene a fact-finding hearing and permit live testimony. Brill v. Guardian Life Ins. Co. of America, 142 N.J. 520, 534, 666 A.2d 146 (1995). “This is not to suggest that every facially unbelievable and bizarre story advanced by a registrant as a basis for an evidentiary hearing will compel such a hearing. A judge might well conclude that a registrant’s unsubstantiated claim that an elderly grandmother or nun was engaged in a consensual sex for drugs transaction is unworthy of belief.” In re C.A., supra, 285 N.J.Super. at 350, 666 A.2d 1375.
We adopt the hearing format suggested in Reyes, supra, 207 N.J.Super. 126, 504 A.2d 43, for probation violation hearings as the appropriate procedure in notification cases when the offender has raised a genuine issue of material fact. At the hearing, “the State may be permitted to rely initially on hearsay evidence but supplement its proofs ... concerning matters which the [offender] contests by eliciting evidence on cross-examination or by the introduction of contrary evidence.” Reyes, supra, 207 N.J.Super. at 139, 504 A.2d 43. The State may, if it wishes, decide not to provide additional proof and rely solely on its hearsay evidence, and the “court may conclude that the State’s hearsay evidence is sufficient to overcome live testimony offered on behalf of the” offender. Id. at 139 n. 5, 504 A.2d 43.
We also agree with the Appellate Division' that neither side may compel the victim’s testimony without leave of the court. In re C.A., 285 N.J.Super. at 350, 666 A.2d 1375. The State often avoids a trial, particularly in a sex offense ease, so that the victim will not be forced to testify. We are therefore reluctant to compel *98a victim to testify unless it is absolutely necessary. The trial courts should only seek to compel such testimony when there is a real need for the testimony that cannot be met in an alternative manner. We expect that only in the rarest of cases will a court compel the testimony of a victim. In those cases, we suggest that, when possible, the trial court itself conduct all questioning of the victim.
V
We have little difficulty concluding that the proffered documentary hearsay evidence in this ease was reliable and was properly considered by the trial court. The State offered three reports (two by police officers and one by a doctor at a hospital) detailing the incident as related by A.Z. We readily accept these reports as reliable insofar as they relate that A.Z. actually told that story to these individuals. We presume that police officers and medical doctors will accurately report on the statements given to them. Indeed, because of that indicia of reliability, those reports would ordinarily be admitted in court as an exception to the hearsay rule under N.J.R.E. 803(c)(6), although A.Z.’s statements inside those reports would not be admitted under that rule, see Sets v. Strelecki, 110 N.J.Super. 14, 22, 264 A.2d 247 (App.Div. 1970).
However, we must also determine whether there is sufficient evidence of reliability in A.Z.’s statement — i.e., we know that she said it to the police and hospital, but is what she said true? Again, we find that the proffered evidence was sufficiently reliable. A.Z. reported the story immediately after the incident occurred, when she voluntarily telephoned the police, while in an emotional and upset state, to report a rape. Cf. Biunno, Current N.J. Rules of Evidence, Comment 1 on N.J.R.E. 803(c)(2) (1994-95) (“Excitement caused by the observation of a startling event insures the reliability of a spontaneous statement about it made at or near the time of the event’s occurrence.”) She then told the story again when she was admitted to a hospital for post-traumatic stress *99disorder, when she “was more interested in obtaining a diagnosis and treatment culminating in a medical. recovery than ... in obtaining a favorable medical opinion culminating in a legal recovery.” Biunno, supra, Comment on N.J.R.E. 803(c)(4).3 The time and place when A.Z. made her statements lend credence to their reliability.
Furthermore, the fact that A.Z. reported the same story to several police officers, as well as the hospital, is further proof of the statement’s reliability. “As the trial judge stated, nothing on the face of the documentary evidence suggested inconsistencies or credibility problems.” In re C.A., 285 N.J.Super. at 348, 666 A.2d 1375. Because the hearsay statements were reliable, they were properly admitted and would be sufficient to sustain the State’s burden of proof in presenting a prima facie case that C.A. deserved to be classified as Tier Three.
Based on our review of the record, however, C.A. does raise a genuine issue of material fact in his claim that he did not use a knife and that A.Z. engaged in a consensual trade of sex for drugs. He should therefore be granted a hearing. At that remand hearing, the trial court should allow C.A. to present his version of the incident, either through reliable hearsay, an affidavit, live testimony, or other credible evidence. The State can decide to rely on the hearsay evidence that it has already presented to establish its prima facie case. The State may also present additional evidence if it deems such evidence necessary, and, if C.A. presents live testimony, it may challenge his story on cross-examination. After the hearing, the trial court should decide whether C.A. has established, by the preponderance of evidence, that the offense did not occur or that he did not possess a knife *100during the offense. If C.A. establishes those facts, then the Scale must be recalculated.
This procedure is in accordance with Doe v. Poritz and the Registration and Notification Laws and properly and fairly balances the procedural due process rights of the registrant with the needs of the community, as well as protecting the rights of the victim.
VI
Although the validity of the Scale was not directly raised in this appeal, the Court sua sponte requested the parties and various amici curiae to address whether the Scale and the current guidelines comport with the RCNL and Doe v. Poritz. We now discuss the concerns expressed by the parties and the amici curiae about the validity and reliability of the Scale. Generally, those concerns fall into two categories: whether the quality and nature of the offense (seriousness of the offense) is an appropriate consideration of risk of re-offense under the RCNL and Doe v. Poritz, and whether the authors of the Scale assigned the proper weight to each of the four basic categories. We also examine the proper use of the Scale.
A
As previously stated, supra, at 82, 679 A.2d at 1158, the Attorney General, pursuant to N.J.S.A. 2C:7-8, adopted guidelines. In Doe v. Poritz, supra, 142 N.J. at 24 n. 2, 662 A.2d 367, we found these original guidelines to be deficient because they failed to include several risk factors required by the Community Notification Law. On September 14, 1995, the Attorney General promulgated new guidelines to meet the requirements contained in Doe v. Poritz. As discussed supra at 82-83, 679 A.2d at 1158—1159, the new guidelines include the Registrant Risk Assessment Scale and the RRA Manual. The Seale and the guidelines were designed to provide prosecutors with an objective standard on which to base the community notification decision mandated by *101statute and to assure that the notification law is applied in a uniform manner throughout the State.
B
Quality of Offense
All of the parties and amici agree that the Scale and RRA Manual are premised on the belief that the statutory measurement, the risk of re-offense, includes a combination of two factors: the damage likely to be caused by re-offense, if it occurs (the quality of the re-offense), and the likelihood of re-offense. The authors of the Scale contend that the quality of re-offense is in large part determined by the seriousness of the registrant’s prior offense record. The Registrant and amici, however, dispute the State’s contention that the use of quality of re-offense criteria conforms with the risk factors provided by the Legislature in the RCNL. According to the Public Defender, the Legislature did not envision the creation of a scale that would classify registrants according to the potential gravity or nature of future sex offenses. The Public Defender argues that it was the Legislature’s intent that tier classification be based solely on the likelihood of re-offense without regard for the degree of harm or lack of harm that might result from a particular future offense. Consequently, the Public Defender contends that the “Seriousness of Offense” category should not be treated as an independent category on the Scale, but instead should be considered in connection with and as a subset of the statutory criminal history factors listed in N.J.S.A. 2C:7-8(b)3 and (b)4.
The Legislature, however, did intend for the gravity or nature of past offenses to be considered as a component of the risk of re-offense. In Doe v. Poritz, supra, 142 N.J. at 74, 662 A.2d 367, we observed that the focus of Megan’s Law in determining the appropriate notification level was the risk of re-offense. For the purposes of community notification, part and parcel of the risk of re-offense is the gravity of the crime the particular registrant *102might re-commit. We clearly stated that the factors making up tier determinations are “strongly related to the risk of re-offense and the consequent need for greater or lesser notification.” Ibid. The need for greater or lesser notification is directly related to the gravity of the offense to be re-committed along with the risk that the registrant will re-commit whatever crime the registrant committed before.
When we said in Doe v. Poritz that the statutes “were designed simply and solely to enable the public to protect itself from the danger posed by sex offenders,” we did not rule out the understanding that the level of notification required for the public to protect itself varies according to what crime the public must guard against. Id. at 73, 662 A.2d 367. The Scale was therefore appropriately designed so that people who are very likely to recommit their offense, but whose offenses are low in seriousness, such as exhibitionists, would not be ranked as high on the Scale as people who commit fewer, but more serious, offenses such as rapists or sexual thrill killers. The Attorney General’s decision to include factors in the Scale that relate to the quality or nature of a re-offense is consistent with the RCNL and Doe v. Poritz; accord W.P. v. Poritz, 931 F.Supp. 1199, 1221 (D.N.J.1996).
Further evidence that the gravity of the offense involved relates to tier classification is how every factor in the “Seriousness of Offense” category relates to the RCNL’s enumerated risk factors. Supra at 81-82, 679 A.2d at 1157-1158. “Whether the offender committed the sex offense against a child” is a statutory risk factor, and thus the inclusion of the age of victim factor in the “Seriousness of Offense” category is appropriate and warranted. N.J.S.A. 2C:7-8b(3)(c). Similarly, consideration of the “degree of force” (low risk for no physical force or threats, moderate risk for threats and minor physical force, and high risk for violence, use of weapon, and significant harm to the victim), the second qualitative factor contained in the Scale under the “Seriousness of Offense” category, was mandated by the Legislature. N.J.S.A. 2C:7-8b(4) *103requires consideration of “[t]he relationship between the offender and victim,” “[w]hether the offense involved the use of a weapon, violence, or infliction of serious bodily injury,” and “[t]he number, date, and nature of the offenses” (emphasis added). The degree of force used by a registrant in prior sex offenses describes the nature of those offenses, and thus its inclusion in the Scale is reasonable. The third qualitative factor contained in the “Seriousness of Offense” category is “degree of contact” involved in prior sexual offenses (low risk for no contact or fondling over clothing, moderate risk for fondling under clothing, and high risk for penetration). Although this factor was not specified by the Legislature, it too relates to the nature of the offense, which is a specified risk factor. N.J.S.A. 2C:7-8b(4)(c). In addition, this factor is relevant to the determination of whether a registrant has a duty to register. Under Megan’s Law, offenders who select adult victims do not have a duty to register unless the nature of their sexual offense involves penetration or sexual contact accomplished by means of violent, dangerous or abusive conduct. N.J.S.A. 2C:7-2b(1); N.J.S.A. 2C:14-2; N.J.S.A. 20:14-3.
C
Weight of Categories
The Scale is divided into four basic categories. The first two categories, “Seriousness of Offense” and “Offense History,” are considered static categories because they relate to the registrant’s prior criminal conduct. The factors comprising “Seriousness of Offense” category include: degree of force, degree of contact, and age of victim. The factors comprising the “Offense History” category include: victim selection, number of offenses/victims, duration of offensive behavior, length of time since last offense, and history of anti-social acts. The remaining two categories, “Characteristics of Offender” and “Community Support” are considered to be dynamic categories, because they are evidenced by current conditions. The factors constituting the “Characteristics of Offender” category include response to treatment and sub*104stance abuse. The final category, “Community Support,” measures the registrant’s therapeutic support, residential support, and employment/educational stability.
Each of the Scale’s thirteen factors is assigned a risk level: low risk (= 0), moderate risk (= 1), or high risk (= 3). The total for all levels within a category provides a score that is then weighted based on the particular category. The point total of the “Seriousness of Crime” category, which is designed to predict the nature of any re-offense, if it occurs, is multiplied by five. The “Offense History,” “Characteristics of Offender,” and “Community Support” categories are multiplied by three, two, and one respectively.
The registrant and amici contend that both the static factors are weighted too heavily in the Scale. Further, they assert that the dynamic factors, those that emphasize the subjectiveness of the offender and reflect his recent behavior, are not given sufficient importance in the Scale. Various amici contend that the Scale fails to give proper consideration to assessments of an offender’s characteristics and post-sentence behavior. In addition, several amici argue that, in order to accurately predict risk of re-offense, any scale should be combined with a clinical interview of the offender.
A Committee of mental health professionals and legal experts (Committee) developed the Scale. They examined risk assessment scales being used in the United States and Canada. After reviewing the scientific literature, the Committee selected for inclusion in the Scale those factors that met two conditions. First, all of the factors selected had to be empirically supported in the risk assessment field as criteria positively related to the risk of re-offense. Second, all of the factors selected had to be fairly concrete criteria that could be gathered in a consistent and reliable manner. In preparing the Seale, the Committee also conducted clinical interviews. Criteria that were too cumbersome, too expensive to ascertain, or too difficult to gather in a reliable manner were not used.
*105The Committee’s decision to weight the static factors, particularly the Seriousness of Offense category, higher than the dynamic factors is consistent with the majority of scientific literature in this area. See, e.g., Vernon L. Quinsey et al, Predicting Sexual Offenses, in Assessing Dangerousness, 114 (Jacquelyn C. Campbell ed., 1995); Joseph J. Ranero & Linda Meyer Williams, Recidivism Among Convicted Sex Offenders: A 10-Year Followup Study, 49 Federal Probation 58 (1985); Frank Tracy et al., Program Evaluation: Recidivism Research Involving Sex Offenders, in the Sexual Aggressor, 198 (Joanne G. Greer & Irving D. Stuart eds., 1993). Most studies show that the static factors used in the Scale are the best predictors of risk of re-offense. Ibid. Risk assessment experts generally agree that the best predictor of a registrant’s future criminal sexual behavior is his or her prior criminal record. We emphasize that the focus on prior offenses is not due to any attempt at punishment but is rather a scientific attempt to better protect the public safety from registrants likely to re-offend. Although the static factors are weighted the highest, the Scale also includes various dynamic factors, such as the registrant’s likely response to treatment, residential support, employment/educational stability, and therapeutic support. Those factors were weighted less heavily in the Scale because the scientific literature generally shows that they are not as good predictors of risk of re-offense as are the static factors. Ibid. Although the weights assigned to the categories in the Scale have not been scientifically proven to be valid, the State has produced sufficient evidence to convince us that the factors used in the Seale are reliable predictors of recidivism and are weighted in the Scale according to their relative effectiveness as predictors. The greater weight attached to the static categories is in accord with expert opinion on criminal sexual behavior. Cf. W.P. v. Poritz, supra, at 1222 (finding that higher weight accorded to past offenses does not violate due process).
Although using both the Scale and clinical interviews of registrants may on some occasions provide a more accurate pre*106diction of risk of re-offense than use of the Scale alone, the Committee decided that clinical interviews of each registrant were not necessary. That decision was based on scientific literature that showed that the use of actuarial concrete predictors is at least as good, if not in most cases better, in terms of reliability and predictability than clinical interviews. After conducting a cost-benefit analysis of requiring clinical interviews of each offender, the Committee determined that the minimal improvement in predictability that might be gained by using clinical interviews could not justify the high cost associated with conducting such interviews for each registrant. Based on the record presented, that decision appears reasonable.
The registrant and amici curiae also question whether the new guidelines properly consider “whether psychological or psychiatric profiles indicate a high risk of recidivism.” N.J.S.A. 2C:7-8b(5). That concern is part of the larger problem of what role expert testimony related to psychological or psychiatric profiles should play in Megan’s Law cases. In In re G.B., 286 N.J.Super. 396, 407, 669 A.2d 303 (1996), the Appellate Division was persuaded, in “light of the circumstances surrounding the offense,” to remand the case to permit the registrant to present expert testimony demonstrating that the variable factors that related to him should result in a lower tier classification. We granted the State’s petition of certification, 143 N.J. 328, 670 A.2d 1068 (1996), limited solely to the issue of the use of experts in a Megan’s Law hearing. Accordingly, we defer our decision with respect to expert testimony, including psychological and psychiatric testimony, until we have an opportunity to hear that case. We do suggest, however, that the Attorney General be prepared to show that the State has properly responded to the Court’s directive in Doe v. Poritz, supra, 142 N.J. at 24, n. 5, 662 A.2d 367, that in the court-mandated revised guidelines “psychological or psychiatric profiles” be considered in support of a low-risk assessment as well as a high-risk assessment.
*107D
Reliability of the Seale
Neither C.A. nor amici have presented us with an alternative method on which to classify registrants, yet they urge us to delay use of the Scale until it is empirically validated through years of field tests. They assert that the Scale is unreliable, untested, and operates in an arbitrary and capricious manner.
Although the Seale was not field-tested, it was subjected to intense scrutiny by experts. Empirical validation of the Scale is neither feasible nor practicable. Researchers would have to release offenders and then wait for five or ten years until they have enough data to determine which factors were the best predictors of recidivism. Obviously, it was not the Legislature’s intent for the Attorney General to wait ten years before assigning offenders to tier levels. The Legislature concluded that the need to protect children from the risk of re-offense by future offenders and previously convicted offenders warranted registering sex offenders as soon as possible. Thus, the Legislature mandated that the Attorney General promulgate guidelines and procedures for the notification required pursuant to the RCNL within sixty days of the effective date of Megan’s Law. N.J.S.A. 2C:7-8.
By analyzing the scientific literature on valid and reliable predictors of recidivism, the Committee, within the time constraints that it faced, created a useful and rational scale that can be used as a tool for deciding tier classification. Although the Scale has not been empirically validated through scientific field studies, the factors that comprise the Scale have been shown to be the best indicators of risk of re-offense.
Based on the record presented by the parties and amici, we find that the Scale is an appropriate and reliable tool whose use is consistent with the requirements of the RCNL and Doe v. Poritz. In that ease, we held that tier classification under the RCNL requires a grouping of offenders into three categories depending upon comparative risk of re-offense. Doe v. Poritz, *108supra, 142 N.J. at 32-33, 662 A.2d 367. We explained that a “substantially higher” risk of re-offense must distinguish one group of offenders from those offenders in proximate groups. Ibid. However, because of the “unavoidable uncertainties in this entire area,” we did not “believe it is realistic to impose requirements of proof of some statistical differention of the risk of re-offense between the classes or between the offender before the court and the typical offender of other classes.” Id. at 33-34, 662 A.2d 367. The Scale gives prosecutors a reasonably objective measure on which to assign registrants to the low, moderate, or high tier classification. Indeed, one of the great strengths of the Scale is that it can provide consistent measures of risk of re-offense. Because the Seale is concrete and simple to use, if the same individual were hypothetically being evaluated in different counties, he should receive a close, if not identical, score.
E
Use of Scale
The Scale, however, is not a scientific device. It is merely a useful tool to help prosecutors and courts determine whether a registrant’s risk of re-offense is low, high, or moderate. Yet, the Scale is just that — a tool. Although a tier classification made on the basis of the Scale should be afforded deference, a court should not rely solely on a registrant’s point total when it conducts a judicial review of a prosecutor’s tier level classification or manner of notification decisions.
Many of the concerns expressed by the registrant and amici curiae are substantially addressed by the judicial review procedures established under Doe v. Poritz, supra, 142 N.J. at 83-84, 679 A.2d at 1158-1159. Under that case, before a Tier Two or Tier Three classification is implemented, a court must review the correctness of the level and manner of notification selected by the prosecutor. The State, which has the burden of going forward with prima facie evidence that justifies the proposed level and manner of notification, may use the tier classification suggested by the Seale in making out its case. However, tier classifications *109based on the Scale are subject to judicial review and modification on a case-by-case basis. Because it is impossible to create an all-inclusive scale, any classification based on the Scale should not be viewed as absolute.
Although the Scale provides a useful guide for the prosecutors and courts to evaluate risk of re-offense there is still a value judgment that must be made when determining a registrant’s risk of re-offense and proper classification. In most cases, we expect that the tier classification suggested by the Scale will be the same classification recommended by the prosecutor and approved by the court. However, there may be cases in which the registrant presents subjective criteria that would support a court not relying on the tier classification recommended by the Scale. In those cases, we do not expect the court to blindly follow the numerical calculation provided by the Scale, but rather to enter the appropriate tier classification. We recognize that subjective accomplishments, such as an individual registrant’s positive response to treatment, may warrant a lower classification than the Scale recommends. However, we believe that those determinations are best made on a case-by-case basis within the discretion of the court.
If after reviewing the State’s prima facie ease, the trial court is concerned that the proposed tier classification may be inappropriate, the court can, if necessary, secure its own subjective evaluations, appoint its own experts, and order further submission of documentation by the prosecutor. Any classification that is inconsistent with the classification based on the Scale is subject to judicial review by either side through appeal and any finding will have to be supported on the record. Therefore, the court’s findings of the appropriate tier classification and manner of notification should be clearly set forth in the record.
VII
Balancing C.A.’s procedural due process rights and right to fundamental fairness with the community’s right to protect itself *110against the risk that he may commit another sexual offense, we find that C.A.’s prior nonconviction offense may be considered in determining his tier classification and that the State may prove that offense solely by reliable documentary hearsay evidence. We further find, however, that C.A. presented sufficient evidence to create a material factual question about the nature and circumstances of the offense that may have a significant effect on his tier classification. Therefore, we affirm the Appellate Division’s remand of the case to the trial court for a hearing on those points.
We also find on the record that the Scale is a reliable and useful tool that the State can use to establish its prima facie case concerning a registrant’s tier classification and manner of notification. The procedures provided in a civil hearing concerning a registrant’s tier classification and manner of notification together with the requirement for judicial review of those decisions adequately protect the registrant’s rights to procedural due process and to fundamental fairness.
The judgment of the Appellate Division in In re C.A., supra, 285 N.J. Super. 343, 666 A.2d 1375, is affirmed.
STEIN, J., concurring in result.
For affirmance — Justices HANDLER, POLLOCK, O’HERN, GARIBALDI, STEIN and COLEMAN — 6.
*111APPENDIX A
REGISTRANT RISK ASSESSMENT SCALE
[[Image here]]
The Registrant Risk Assessment Manual (RRA Manual) is also included in the Registration and Community Notification Laws Bench Manual, October 16, 1995.
The initials used are fictitious. This Court has, for the purposes of confidentiality, refrained from identifying the names of those involved as well as the municipality and county in question. Initials of victims and witnesses are also fictitious.
We note that, even if we were to hold hearsay non-admissible, A.Z.'s statement to the police would possibly be admissible as an excited utterance under N.J.R.E. 803(c)(2), and her statement to the hospital might be admitted under N.J.R.E. 803(c)(4) as a statement for purpose of medical diagnosis and treatment.