Defendant Phillip Rimmasch was charged and convicted after a bench trial of forcible sexual abuse, rape, forcible sodomy, and incest. These charges arose out of alleged incidents of sexual activity between defendant and his daughter. Rim-masch’s sentence was stayed, and he was placed on probation for eighteen months, on condition that he undergo psychological treatment. He appeals his conviction, arguing, inter alia, that the trial judge erred in admitting certain testimony by expert witnesses called on behalf of the prosecution in which one or more of them affirmed the truthfulness of his daughter’s version of events, testified to the psychological profile of the typical victim of child sexual *390abuse, compared his daughter’s characteristics with those of the profile, opined that his daughter was a victim of abuse on the basis of both the profile comparison and impressions gathered during interviews with the daughter, and in some instances, also opined that she was the victim of incest and that her father was the perpetrator. We agree in the main with Rim-masch’s contentions and reverse.1
We are aware that child sexual abuse is of great concern to the American public. The Utah legislature, like many around the country, has responded to this concern by enacting a number of measures designed to make apprehension and conviction easier and punishment more severe. Some of these measures have relaxed legal rules that might have operated to bar admission of some otherwise reliable testimony. See Utah Code Ann. §§ 76-5-409, -410, -411 (Supp.1988). Others have authorized special procedures to make less traumatic the giving of testimony by children. See Utah R.Crim.P. 15.5 (codified at Utah Code Ann. § 77-35-15.5 (Supp.1988)). Still others have provided sentences of severity unique in recent Utah law for those found guilty of such offenses. See Utah Code Ann. §§ 76-3-406, -407, -408 (Supp.1988); §§ 76-5-402.1, -402.3, -403.1, -404.1, -405 (Supp.1988). Nationally, some courts have joined the chase by easing rules that may appear to make conviction more difficult. See, e.g., Note, Liberalization in the Admissibility of Evidence in Child Abuse and Child Molestation Cases, 7 J. of Juv.L. 205, 206 (1983); J. Myers, Child Witness Law and Practice (1987) [hereinafter Myers].
However, the fact that child sexual abuse has emerged as a critical problem about which the public is seriously concerned does not mean all legal rules that may constitute obstacles to increasing the conviction and incarceration rates of those accused of such crimes, as opposed to those actually guilty, can properly be brushed aside. Fundamental changes in long-standing evidentiary rules should not be made lightly, even when it is claimed that those rules may stand in the way of convictions. See State v. Moran, 151 Ariz. 378, 380, n. 2, 728 P.2d 248, 250 n. 2 (1986). “No matter how defenseless the child, or how strong the policy of protecting victims of abuse, justice is not served by ‘proving’ sexual abuse through misleading and unreliable testimony.” Note, The Unreliability of Expert Testimony on the Typical Characteristics of Sexual Abuse Victims, 74 Geo.L.J. 429, 451 (1985) [hereinafter Georgetown Note].
There is no reason to recite the facts of this case in any great detail.- Rimmasch allegedly engaged in a long-term intimate relationship with his daughter beginning when the girl was approximately seven and continuing until she was thirteen or fourteen. The specific charges filed related to one incident that occurred during this period.
At trial, the daughter, then seventeen years old, described the alleged incident of sexual activity. The father denied the accusation, and various members of the family, as well as family acquaintances, testified in support of the father. In addition to the daughter’s testimony, the prosecution relied heavily on the testimony of four expert witnesses: Dr. William Palmer, Dr. Johanna McManemin, Dr. Ann Tyler, and Thomas Harrison. All four experts repeated the statements the daughter had made during interviews as to what had happened between her and her father. Each expert expressed his or her observations and opinions about whether the daughter was the victim of abuse. All four opined that the daughter had, in fact, been abused; some went further and said that she was a victim of incest and that her father was the culprit. The testimony of these four experts takes up almost two-thirds of the trial transcript and occupied several trial days.
At the conclusion of the trial, the judge found Rimmasch guilty. In so doing, he *391stated: “It is the opinion of the Court that the State through the testimony of the alleged victim together with corroborating testimony of the expert witnesses called by the State has made out a prima facie case against the defendant and in favor of a verdict of guilt.”
On appeal, the dispositive issues relate to the testimony of the experts. The specific issues we address concern the propriety of the expert witness testimony to the following effect: the daughter was telling the truth; there is a psychological profile of the typical child victim of sexual abuse; the daughter conformed to the profile; and on the basis of that conformance and/or a subjective assessment of the daughter’s credibility, the experts could opine that the daughter had been abused, that the daughter was a victim of incest, and that the father was the perpetrator of the abuse.
We first consider the claim that the experts improperly testified as to the truthfulness of the daughter’s version of events. We will then consider the rest of the issues together. This division is largely a matter of convenience, attributable to the fact that the first claim can be disposed of under rule 608(a) of the Utah Rules of Evidence, while the remaining issues are considered under a common analytical framework based on Utah Rule of Evidence 702.
I
The truthfulness issue presents two subsidiary questions: whether the experts actually testified that the daughter was telling the truth about the incident with her father; and, if so, whether the testimony was properly admitted.
Rimmasch contends that the experts testified to his daughter’s truthfulness by actually stating that she was telling the truth and by conveying to the fact finder, via their opinions on abuse, a belief in her truthfulness. Such testimony, Rimmasch argues, invades the province of the jury and is improper under Utah Rule of Evidence 608(a). The State responds that the experts expressed opinions not on the daughter’s veracity, but on whether she was a victim of sexual abuse and incest. In the alternative, the State contends that it is not prejudicial error for a psychiatric expert in a child sexual abuse case to testify that he or she finds the victim’s story believable.
We address the State’s alternative hypothesis first. Utah Rule of Evidence 608(a) provides in pertinent part:
The credibility of a witness may be attacked or supported by evidence in the form of opinion or reputation, but subject to these limitations: (1) the evidence may refer only to character for truthfulness or untruthfulness....
This rule permits testimony concerning a witness’s general character or reputation for truthfulness or untruthfulness but prohibits any testimony as to a witness’s truthfulness on a particular occasion. See, e.g., United States v. Azure, 801 F.2d 336, 341 (8th Cir.1986) (construing Federal Rule of Evidence 608(a), which is the model for Utah’s rule); United States v. Awkard, 597 F.2d 667, 671 (9th Cir.), cert. denied, 444 U.S. 885, 100 S.Ct. 179, 62 L.Ed.2d 116 (1979); Tevlin v. People, 715 P.2d 338, 341 (Colo.1986) (relying on Colo.R.Evid. 608, which is identical to Utah R.Evid. 608); People v. Koon, 713 P.2d 410, 412 (Colo.Ct.App.1985).
Some members of this Court have previously recognized that an expert may not express a direct opinion on whether a witness was truthful on a particular occasion. In State v. Lairby, 699 P.2d 1187 (Utah 1984), we were also presented with a child abuse case. In the lead opinion, authored by Justice Durham and joined by the Chief Justice, it is stated, albeit in dictum, that “Dr. Liebroder might have been able to testify about Wanda Lairby’s psychological capacity for untruthfulness or delusional testimony, but Dr. Liebroder could not have offered any views on the testimony actually given.” Id. at 1199 (emphasis added).
The State urges that we not follow this dictum in Lairby but instead follow State v. Kim, 64 Haw. 598, 645 P.2d 1330 (1982). Kim is the only decision we have found that holds an expert’s direct statement of *392an opinion on the believability of a witness’s version of the facts to be admissible under a state counterpart to Federal Rule of Evidence 608(a), from which Utah Rule of Evidence 608(a) was also taken. The Kim court said:
Arguably, [Haw.R.Evid. 608(a)(1) ] would preclude the admission of Dr. Mann’s statement since it arguably provided not only his opinion as to the character of the witness, but an opinion as to whether the witness spoke truthfully on a specific occasion....
We do not find this to be the case.... Essentially, the difference between an opinion as to character for truthfulness and an opinion as to the believability of a witness’s statement is the difference between “I think X is believable” and “X’s statement is believable.” We feel the admissibility of either statement should not turn on niceties of phraseology but on the probative value of the testimony.
64 Haw. at 609 n. 14, 645 P.2d at 1339 n. 14.
We disagree. Even if the distinction between opinions about a witness’s character for truthfulness and opinions about the truthfulness of a witness’s particular statements was based merely on “niceties of phraseology,” that distinction is embodied in the plain lauguage of rule 608(a) of the rules of evidence of both Hawaii and Utah and cannot be ignored. But, more importantly, we do not think the distinction is an empty one. It represents an important policy choice: it prevents trials from being turned into contests between what would amount to modern oath-helpers who would largely usurp the fact-finding function of judge or jury. State v. Moran, 151 Ariz. 378, 382, 728 P.2d 248, 252 (1986). The overwhelming majority of courts have expressly or impliedly rejected Kim.2 We join these courts and hold that rule 608(a)(1) bars admission of an expert’s testimony as to the truthfulness of a witness on a particular occasion.
The next question is whether the experts offered testimony that runs afoul of rule 608(a). It can be a subtle business to determine whether a particular opinion, based only in part on an appraisal of veracity, runs afoul of rule 608(a) and is therefore flatly inadmissible. Cf. Annotation, Admissibility Of Opinion Of Medical Expert As Affected By His Having Heard The Person In Question Give The History Of His Case, 51 A.L.R.2d 1051 (1957) (permissibility of physician’s reliance on.patient statement as to symptoms); State v. Ward, 10 Utah 2d 34, 347 P.2d 865 (1959) (same).
To put this discussion in context, we set out two portions of Dr. Tyler’s testimony that Rimmasch contends go directly to his daughter’s credibility. In the first instance, Dr. Tyler was asked by the prosecutor to explain the basis for her opinion that the daughter had been sexually abused. Dr. Tyler responded: “Well, specifically, in *393my opinion, one does not give this kind of information with the amount of details and the amount of clarity unless one has experienced it.” Later, on redirect examination, the prosecutor asked Dr. Tyler to explain what the daughter had to gain by accusing her father of sexual abuse. Dr. Tyler stated:
One of the things that we do when we are doing a clinical interview and looking at the reality of the statements is to consider the alternative hypothesis, which is that the individual is not telling the truth; and based on the richness of the detail and the uniqueness of some of the incidents that she described, I can’t see that — well, then, if you would consider the alternative that [the daughter] is not telling the truth, then you would have to look at the consequences of the lie and what — why she would lie. And in talking with [the daughter] and reviewing her test data, I think she has only to lose.... [S]o I don’t know what she would have to gain.
Does this testimony amount to giving a direct opinion on the truthfulness of the daughter’s description of the incidents of sexual abuse at issue? Although it is arguable that this testimony amounts only to a statement of the basis of the expert’s opinion on abuse, we conclude that the prosecutor, in eliciting this testimony and focusing on the reasons why Dr. Tyler thought the daughter was telling the truth, crossed the line and elicited a direct opinion on the daughter’s truthfulness at the time she made her allegations of abuse. Through the first statement, the fact finder learned that Dr. Tyler, as an expert, had concluded that the daughter was telling the truth because of the amount of detail and clarity the daughter used in relating the alleged occurrences of sexual abuse. The second statement told the fact finder that one of the reasons Dr. Tyler thought the daughter’s accusations of abuse against her father were truthful was because the doctor could envision no motive for the daughter to lie. We hold that the court violated rule 608(a) in admitting into evidence Dr. Tyler’s testimony that the daughter told the truth when she described the alleged sexual abuse.
Rimmasch contends that the three other experts also gave testimony that was not admissible under rule 608(a). This testimony by Dr. Palmer, Dr. McManemin, and Mr. Harrison expressed opinions on whether the daughter had, in fact, been abused. These experts based their opinions, at least in part, on their appraisal of the veracity of the daughter during interviews, a fact they communicated to the trial court. We have reviewed the testimony of these experts and conclude that while it resembles Dr. Tyler’s in some particulars, it does not focus upon the truthfulness of the daughter’s statements to the same degree. We therefore conclude that it does not cross the line from being a statement of the basis of an expert opinion on abuse to being a comment on truthfulness on a particular occasion. The admissibility of this testimony depends, then, on whether it satisfies the rule on expert testimony considered in the next portion of this opinion. For the reasons discussed there, we conclude that it was not admissible.
II
We now address the remaining chal-lénges to the expert testimony. These concern the trial court’s admission of opinion testimony that the daughter was the victim of sexual abuse and, according to some experts, that the abuse was a result of incest committed by her father. These opinions were based largely upon comparisons of the daughter’s characteristics with those of the “typical” abused child and upon the experts’ subjective appraisals of the daughter’s truthfulness during interviews. It is the grounding of opinions on these comparisons and appraisals that defendant challenges.
A brief overview of the expert testimony and the trial court’s disposition of Rim-masch’s objections to this testimony will set the stage for evaluating the propriety of its admission.3
*394Dr. Palmer conducted a physical examination of the daughter and interviewed her concerning her sexual history. He testified that the physical examination neither proved nor disproved the allegations that abuse had occurred several years earlier. (The daughter admitted that she was sexually active with males of her own age by the time of Dr. Palmer’s physical examination.) However, Dr. Palmer concluded, based on his subjective judgment of the daughter’s affect during the interview and what she had told him about the alleged abuse by defendant, when considered in light of the results of the physical examination and his prior experience with abused children, that she had been abused by someone “within her family.”
Mr. Harrison was a member of a team of psychologists and social workers who first evaluated the daughter in 1983. At that time, she had been referred to the team because of conflicts with her family and misbehavior. The team administered a series of psychological tests to the daughter, and Mr. Harrison interviewed her twelve or more times immediately following the 1983 referral. Although he testified that the type of behavioral and psychological problems experienced by the daughter in 1983 can be caused by sexual abuse, the test results were consistent with other types of disorders and other causes.4 In 1983, the daughter did not allege sexual abuse, and the team did not conclude that she had been abused.
In 1984, Mr. Harrison again interviewed the daughter at the request of the prosecution. This time she alleged sexual abuse, and at trial, Mr. Harrison was of the opinion that she was the victim of incest and that she had been an incest victim in 1983 when he first saw her. His conclusion was based in part on her conformance with the psychological profile of a sexually abused child, in part on her ability to demonstrate the alleged acts of abuse with anatomically correct dolls, and in part on his subjective appraisal of her veracity during the interview in which she described the alleged incidents of abuse.
Dr. Tyler interviewed the daughter twice. She and Dr. McManemin administered various psychological tests to the daughter. In furtherance of their diagnosis and treatment of the daughter, they also conducted interviews with other members of the family.
In her testimony, Dr. Tyler described various behavioral and psychological traits that she characterized as being typical of sexually abused children. She said that the daughter evinced a number of these, including “the running away, the poor school behavior, the ambivalence, the self-destruction, the damaged goods self-blame, [the] nightmares, [and] the promiscuity.” She then opined that the daughter had been sexually abused “by her father,” that she was a “victim of incest.” This judgment was based on a comparison of the daughter’s traits, as evidenced by her behavior, the interviews, and the psychological tests, with those of the profile of a typical victim, as well as Dr. Tyler’s subjective judgment of the daughter’s credibility made during the interviews in which the daughter described the alleged incidents of abuse.
During her testimony, Dr. Tyler stated that the fact that the traits of an alleged victim conformed to the profile of a typical abused child would not be enough, without more, to warrant a conclusion that abuse *395had occurred. In fact, Dr. Tyler testified that conformance with the profile would not even justify one’s concluding that abuse more likely than not had occurred. She also testified that the profile alone is not capable of accurately segregating those who have been abused from those who have not. She claimed only that a statistical correlation existed between conformance with the profile and abuse.
Dr. Tyler’s testimony was similarly qualified regarding the results of the psychological tests administered to the daughter. The results of the tests could, at most, be said to have raised a red flag on the abuse issue, suggesting further research; the test results alone were not enough to permit one to say that abuse had occurred. She stated that her ability to render an opinion on abuse was heavily dependent on the interviews she had conducted with the daughter in which she judged the veracity of the daughter’s version of the facts against the backdrop of the profile comparison and the psychological test results.
Dr. McManemin was an associate of Dr. Tyler’s. In her testimony, she claimed more for the psychological test results than Dr. Tyler. Dr. McManemin asserted that the test results, in and of themselves, showed that it was “very likely” that the daughter had been abused. Like Dr. Tyler, Dr. McManemin compared the characteristics of the daughter with the profile “typical” of an abused child, specifically an incest victim. Among the characteristics she found to be exhibited by the daughter and which she stated were indicative of abuse were “[being] somewhat different than the other family members,” “looking for support outside the family,” “running away,” “stealing,” “a negative attitude toward sex,” and “bed wetting” at ages of less than seven or eight.
Dr. McManemin then went on to give an opinion based on the test data and the profile that the daughter had been “sexually abused repeatedly.” She opined, based on the additional information furnished by the interview with the daughter, that the abuser was her father. She even went further and testified that the whole family exhibited the characteristics of a family in which child sexual abuse was present.
Throughout the testimony of these key experts, which consumed almost two-thirds of the trial, little foundation was offered or demanded by the court as to the scientific basis for the profile of the typical sexually abused child, the ability of the profile to sort the abused from the nonabused with any degree of accuracy, or the ability of the experts to judge whether the daughter was telling the truth during the interviews.
The trial court admitted the expert testimony over the objections of defense counsel on the basis of several assumptions. First, the trial court appears to have concluded that no detailed foundation was necessary because certain statements in the lead opinion in State v. Lairby, 699 P.2d 1187, 1198-1201 (1984), sanctioned the admission of expert testimony about the characteristics of sexually abused children, comparisons of those characteristics with the victim in the case at issue, and the giving of an opinion on whether abuse occurred, even when that opinion is based in part on a subjective judgment of the veracity of the child during interviews. Second, the trial court appears to have been persuaded that the testimony was analogous to “battered child syndrome” evidence that we held admissible in State v. Tanner, 675 P.2d 539 (Utah 1983). Finally, and perhaps most fundamentally, the court appears to have resolved any doubts about the admission of this expert psychological testimony on whether abuse had occurred, in the absence of significant physical symptoms, by accepting the view implicit in the prosecutor’s argument that any weaknesses in the foundation went to the testimony’s weight, not to its admissibility.
Before we can properly evaluate Rim-masch’s challenge to the trial court’s admission of the expert testimony, we must determine the standard by which the admissibility of expert scientific testimony is to be judged. And this determination requires that we address initially the question of whether the lack of an adequate foundation for expert testimony precludes the admission of that evidence or whether *396it simply weakens its probativeness. The parties do not agree on this point. Resolution of this question requires an examination of the interrelationship of the new rules of evidence and our old case law and raises issues that are the subject of continuing debate across the nation in states that have used the Federal Rules of Evidence as a model.
The admission of expert testimony is generally governed by Utah Rule of Evidence 702, which was adopted in 1983. This rule was copied verbatim from rule 702 of the Federal Rules of Evidence, which became effective in 1975. The Utah rule and its federal counterpart provide:
If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise.
While this is the general rule for the admission of all expert testimony, where expert testimony is based upon novel scientific principles or techniques, courts have long imposed additional tests of admissibility that antedate the federal rules. The most commonly used is that set forth in Frye v. United States, 293 F. 1013 (D.C.Cir.1923). Under Frye, “in addition to satisfying the traditional requirements of relevancy and helpfulness to the trier of fact, the proponent must show general acceptance of the principle or technique [upon which the testimony is based] in the scientific community.” McCormick, McCormick on Evidence § 203, at 605 (E. Cleary, 3d ed. 1984) [hereinafter McCormick]; see, e.g., Phillips v. Jackson, 615 P.2d 1228, 1233 (Utah 1980). As we noted in Phillips, the purpose of a more restrictive test for judging the admissibility of scientific testimony is to assure, as a threshold matter, that the evidence is sufficiently reliable to go to the finder of fact. Id.
One danger being guarded against is the tendency of the finder of fact to abandon its responsibility to decide the critical issues and simply adopt the judgment of the expert despite an inability to accurately appraise the validity of the underlying science. See McCormick at 607-08; 3 J. Weinstein & M. Berger, Weinstein’s Evidence If 702[03], at 702-42, -43 (1988) [hereinafter Weinstein], The Frye “general acceptance” test accomplishes this reliability screening because “[verification of the basic principle and its application through widespread replication and practical usage is an appropriate indicium of reliability.” Phillips v. Jackson, 615 P.2d at 1233 (citations omitted).
In recent years, the Frye test has been criticized as being too restrictive because, inter alia, in some circumstances it can operate to deny admission of evidence based on newly discovered principles that may be reliable but so new that they cannot be said to have attained general acceptance in what may be deemed to be the relevant scientific community. See Phillips v. Jackson, 615 P.2d at 1234-35; Kofford v. Flora, 744 P.2d 1343, 1361 (Utah 1987) (Durham, J., concurring in the result); McCormick at 606-08; McCormick, § 203, at 69 (Supp.1987). In Phillips, this Court was asked to determine whether proffered scientific evidence based on HLA tests purporting to determine accurately the probability that a specific individual is the parent of a specific child was sufficiently reliable to be admitted and to fix the standard by which such a determination would be made. There, we followed the modern trend and abandoned exclusive reliance on Frye.5 “Inherent reliability,” rather than “general acceptance,” became the touchstone of admissibility under Phillips. 615 P.2d at 1234. However, Phillips did state that even if satisfaction of the “general acceptance” test was not a sine qua non of admission, “a showing of gen*397eral acceptance would generally be sufficient” to show inherent reliability and to justify the admission of scientific evidence. Id.; see Kofford, 744 P.2d at 1346.
It is against this background that the parties in the present case make their contentions regarding the standard that governs the admission of expert scientific testimony. Rimmasch argues that the proper test is that established by Phillips: “An analysis of the admissibility of scientific evidence, while taking into account general scientific acceptance and widespread practical application, must focus in all events on proof of inherent reliability.” Id. at 1234. Rimmasch asserts that the trial court erred in admitting the expert opinion testimony that the daughter was abused based on the profiles and on the experts’ subjective appraisals of the truthfulness of the daughter because the court did not first require an adequate foundational showing that the scientific premises upon which such testimony was based are inherently reliable, as required by Phillips. See also Kofford, 744 P.2d at 1347. And, Rimmasch asserts, there was certainly no showing that the scientific principles or techniques in question were generally accepted in the relevant scientific community and therefore eligible for judicial notice. See Kofford, 744 P.2d at 1348.
The State responds that the “inherent reliability” threshold requirement for scientific evidence announced in Phillips in 1980 has been superseded as a result of the 1983 adoption of Utah Rule of Evidence 702. The State argues that the text of rule 702 does not require a showing of any particular level of reliability before expert scientific testimony, as opposed to all other expert testimony, is admissible. Instead, all relevant expert testimony that may be said to “assist the trier of fact” should be, admitted; according to the State, concerns about reliability go only to weight and are matters committed entirely to the finder of fact.6
The precise issue presented here — whether all threshold reliability requirements for scientific testimony, either of Frye or of Phillips, were done away with by the adoption of rule 702 — was addressed in our recent decision in Kofford. There we acknowledged that rule 702 phrases the question of admissibility only in terms of whether the proffered evidence “will assist the trier of fact to understand the evidence or to determine a fact in issue.” 744 P.2d at 1347. But we made it clear that regardless of how rule 702 phrases the general test for the admissibility of expert testimony, our case law superimposes a more restrictive test whenever scientific evidence is at issue, and that more restrictive test was set forth in Phillips. “However the test is formulated ... a foundation establishing the reliability of new scientific evidence must be established for it to be admissible.” Kofford, 744 P.2d at 1347. Casting the Phillips/Kofford standard in terms of the rubric of rule 702, it can be said that evidence not shown to be reliable cannot, *398as a matter of law, “assist the trier of fact to understand the evidence or to determine a fact in issue” and, therefore, is inadmissible. Utah R.Evid. 702. Accordingly, we reject the State’s contention that the Phillips inherent reliability threshold standard for admissibility has not survived the promulgation of rule 702 and that mere relevancy is sufficient.
To frame our analysis of Rim-masch’s claims, one more point needs to be made regarding expert scientific evidence. The question of the admissibility of such evidence may be presented in two different ways: a request that the trial court take judicial notice of the “inherent reliability” of the testimony’s foundational principles or techniques or, alternatively, a request that the trial court determine that these principles or techniques are inherently reliable after an evidentiary hearing addressing the issue.7
Kofford is a case in which the question of judicial notice was presented. There, we considered HLA test evidence for the first time since Phillips. Kofford demonstrates that a very high level of reliability is required before judicial notice can be taken. In Kofford, after reviewing the literature and the decisions of other courts, we concluded that, without exception, scientific scholars agreed that the HLA test was reliable as long as the test was conducted properly. 744 P.2d at 1348. We also noted that numerous courts had reached the same conclusion, i.e., that the HLA test was generally accepted in the scientific community as a reliable means for determining paternity. Id. Because the principles underlying HLA evidence met the Frye standard of general acceptance, we concluded that it was a suitable subject for judicial notice. Id.; see Utah R.Evid. 201; cf. State v. Long, 721 P.2d 483, 488-95 (Utah 1986) (notice taken of the fact that eyewitness identification is unreliable on basis of general acceptance of that view by scientific community). Once a scientific principle or test has achieved sufficient reliability that “judicial notice of that fact may be taken, ... foundational evidence as to the validity of the basic principles may be dispensed with in th[e] jurisdiction in the future.” Kofford, 744 P.2d at 1348.
The second situation a court may encounter occurs when the basic principles underlying proffered expert scientific testimony are not suitable for judicial notice because their reliability is not beyond being reasonably questioned. Under Phillips and Kofford, the proponent of scientific evidence that does not qualify for judicial notice must make an initial foundational showing that convinces the trial court that the principles or techniques underlying the proffered testimony meet Phillips’ standard of inherent reliability before the trial court can proceed to consider the normal foundational questions appropriate to any expert testimony. Phillips, 615 P.2d at 1236; Kofford, 744 P.2d at 1347. In the absence of such an initial showing, the evidence is to be excluded. Moreover, a showing of the inherent reliability of the underlying principles or techniques must be made in each case in which similarly grounded evidence is offered.8 Cf. Kof ford, 744 P.2d at 1348.
In sum, if this Court’s position on the spectrum of opinions held by the various *399courts about the admissibility of scientific evidence could be fixed, see supra notes 5 & 6, it would probably be somewhere in the middle, where an awareness of the usefulness of scientific evidence is balanced by a concern that the door not be thrown wide open to any evidence that can be labeled “scientific.” We remain wary of the potential of such evidence to distort the fact-finding process by reason of its superficial plausibility and its potential for inducing fact finders to accept experts’ judgments on critical issues rather than making their own. And we are convinced that trial courts sometimes admit “scientific” evidence without scrutinizing its foundations carefully. It is for these reasons that we have imposed the threshold reliability requirement. See generally Note, United States v. Downing: Novel Scientific Evidence and the Rejection of Frye, 1986 Utah L.Rev. 839, 849-53.9
With the proper standard in mind, we now proceed to consider Rimmasch’s remaining challenges to the admissibility of the experts’ testimony. These challenges can be grouped around two somewhat interrelated issues. First, Rimmasch argues that the trial court erred when it permitted three of the four experts to give their opinions that the daughter was abused or was the victim of incest when their opinions were not based on physical evidence, but on her conformance with a profile typical of children who purportedly had been abused. This testimony, Rimmasch argues, is without adequate foundation in the record.
*400Second, Rimmasch asserts that even if all four experts were properly permitted to give an opinion on whether the daughter was abused, they should not have been permitted to give as one of the grounds for that opinion their subsidiary opinion that the girl told the truth when she recited her version of the alleged events.
Consideration of the first issue, the propriety of the profile testimony, is best broken down into two separate questions: first, whether a court may take judicial notice that the scientific principles and techniques underlying the experts’ profile testimony are inherently reliable; second, if not, whether the prosecution laid a sufficient foundation for the admission of such expert testimony. The reason we deal first with the question of judicial notice is that on appeal, we sustain a trial court’s eviden-tiary ruling on any available ground, even though it may be one not advanced below. See, e.g., State v. Gray, 717 P.2d 1313, 1316 (Utah 1986); State v. Gallegos, 712 P.2d 207, 209 (Utah 1985); State v. Bryan, 709 P.2d 257, 260 (Utah 1985). If the inherent reliability of the scientific principles and techniques underlying the testimony in question is eligible for judicial notice, even though the trial court did not recognize this fact, then we need not consider the adequacy of the foundation actually laid.
In Kofford, we held that the reliability of scientific principles and techniques underlying HLA testing were legitimate subjects of judicial notice because they had been recognized generally by the legal and scientific communities. 744 P.2d at 1348. Some courts have admitted profile testimony to prove that abuse has occurred; however, unlike the situation in Kofford, there is no unanimity in the legal community as to the inherent reliability of a child sexual abuse profile to show that abuse has actually occurred with respect to a specific alleged victim. Compare State v. Kim, 64 Haw. 598, 645 P.2d 1330, 1338-39 (1982) (admitting); Kruse v. State, 483 So.2d 1383, 1385-86 (Fla.Dist.Ct.App.1986) (admitting); Allison v. State, 179 Ga.App. 303, 308-09, 346 S.E.2d 380, 385 (1986) (admitting) with People v. Roscoe, 168 Cal.App.3d 1093, 1098-1101, 215 Cal.Rptr. 45, 49-50 (1985) (excluding); State v. Hudnall, 293 S.C. 97, 100, 359 S.E.2d 59, 61-62 (1987) (excluding); People v. Pullins, 145 Mich.App. 414, 420-21, 378 N.W.2d 502, 505 (1985) (excluding); see State v. Taylor, 663 S.W.2d 235, 240-42 (Mo.1984) (rape trauma syndrome evidence inadmissible to prove rape occurred); Myers at § 4.16; Comment, Syndrome Testimony in Child Abuse Prosecutions: The Wave of the Future?, 8 St. Louis U.Pub.L. Rev. 207, 218 (1989); Comment, The Admissibility of Expert Psychological Testimony in Cases Involving the Sexual Misuse of a Child, 42 U. Miami L.Rev. 1033, 1048-50 (1988); McCord, Syndromes, Profiles and Other Mental Exotica: A New Approach to the Admissibility of Nontraditional Psychological Evidence in Criminal Cases, 66 Or.L.Rev. 19, 41 (1987).
The trial court may have concluded, based on a discussion in the lead opinion in State v. Lairby, 699 P.2d at 1200-01, of the admissibility of expert psychological opinion testimony on whether a particular child had been sexually abused, that this Court had resolved any threshold reliability doubts and had determined that expert testimony of the type offered here is admissible in Utah trial courts. Lairby does not constitute such a determination. The referenced language in the lead opinion was dictum only.
The record also suggests that the trial court may have been persuaded that, as a matter of law, the threshold reliability of the proffered expert testimony was adequately established by analogizing it to “battered child syndrome” evidence held admissible in State v. Tanner, 675 P.2d 539 (Utah 1983). This analogy is not factually or legally convincing. The battered child syndrome consists of “a brief set of narrow, specific, and predominantly physical symptoms” that constitute an accepted medical diagnosis suggesting that the cause of the injuries in question was not accidental. Georgetown Note at 448; see State v. Tanner, 675 P.2d at 542-43. In Tanner, we concluded that “the concept of the battered child syndrome is grounded in scientific research and is widely accepted in *401the medical community” and that “all courts which have addressed the question have affirmed the admission of expert medical testimony regarding the presence of the battered child syndrome.” Id. at 543. Having found that the necessary threshold reliability had been established for that syndrome evidence, we concluded that it was admissible, subject always to a weighing in each case of its probativeness against its potential for unfair prejudice. Id. at 543-44.
In contrast, the child abuse profile consists of a long list of vague and sometimes conflicting psychological characteristics that are relied upon to establish the fact of injury in a specific case as well as the cause. And neither the record nor our independent research demonstrates that there is general acceptance of child abuse profile evidence as a determinant of abuse either by the legal community, as noted above, or by the scientific community, as discussed below. Tanner cannot therefore support the trial court’s admission of the profile evidence.
Not only is there a lack of any concensus about the ability of the profile to determine abuse, but the scientific literature raises serious doubts as to the reliability of profile testimony when used for forensic purposes to demonstrate that abuse actually occurred. Scholars acknowledge that no uniformly identifiable psychological profile applies to sexually abused children as a class.10 They make two related points. First, children may have widely varying reactions to abuse, so that “typicality” is hard to determine. The truth of this proposition is attested to by the fact that existing profiles use very general descriptive terms, such as “guilt” or “anxiety” or “nightmares,” to describe the characteristics of abused children. For similar reasons, some lists of characteristics of abused children include contrasting traits, such as “regressive behavior” and “pseudo-mature behavior,” “acting out” and “withdrawal.” Myers § 4.15, at 153; see Georgetown Note at 451-53 (discussion of several proposed profiles). The second criticism directed at profiles is that psychological characteristics often said to be indicative of abuse may also describe persons suffering from a wide range of emotional problems unrelated to sexual abuse. See McCord, Expert Psychological Testimony About Child Complainants in Sexual Abuse Prosecutions: A Foray into the Admissibility of Novel Scientific Evidence, 77 J.Crim.L. & Criminology 1, 23 (1986); Georgetown Note at 444-48. The vagueness and overinclusiveness of these profiles mean that in the view of many experts, they are not particularly helpful in deciding whether abuse has occurred.11
Suffice it to say, then, that the literature in the area is disparate and contradictory and that child abuse experts have been unable to agree on a universal symptomolo-gy of sexual abuse, especially a precise symptomology that is sufficiently reliable to be used confidently in a forensic setting as a determinant of abuse. See Cerkovnik, The Sexual Abuse of Children: Myths, Research, and Policy Implications, 89 Dick.L.Rev. 691, 705-08 (1985) (compilation of studies on victims of sexual abuse); Georgetown Note at 439-48. It is also fair *402to state that as a purely logical matter, we are not persuaded that, despite the lack of consensus, we can conclude that the profile evidence is inherently reliable. Its weaknesses are manifold. See infra note 13. If the profile relied upon by the experts in the present case was based on the same assumptions as those subject to debate within the legal and scientific communities, then we cannot take judicial notice of its reliability.
A brief review of the experts’ testimony confirms that they based their opinions that the daughter was abused on a psychological profile indistinguishable from those discussed above. The characteristics of the daughter relied upon as points of conformance with the profile included the following: running away, poor school behavior, ambivalence, self-destructive attitudes, nightmares, promiscuity, poor self-image, negative attitude toward sex, and bed wetting at a young age.12 Illustrative of the generality of the profile is Dr. McMane-min’s statement that “toileting accidents are usually a symptom of sexual abuse.” 13
*403It is clear from the foregoing that the reliability of the scientific principles and techniques underlying sexually abused child profile testimony of the sort given in this case is not a fact of which this or any court may take judicial notice. In saying this, we do not pretend that we have undertaken a comprehensive analysis of all the literature in the field, and we do not mean to imply that profile testimony is unreliable as a matter of law. We simply conclude that the necessary threshold reliability of such testimony has yet to be established. Therefore, the issue must be explored in individual cases by trial judges, who must determine whether the profile testimony is reliable. If it is found reliable, an analysis of whether the potential for unfair prejudice resulting from the evidence outweighs its probativeness must be made before such evidence can be received and expert opinions based upon it. See supra note 8.
Having concluded that the trial court here could not properly have taken judicial notice of the inherent reliability of the scientific principles and techniques upon which the profile testimony was based, we must next determine whether the trial court properly admitted the evidence and the opinion grounded upon it based on the foundational showing made by the state. According to Phillips, that foundational showing must explore with careful precision such questions as the correctness of the scientific principles underlying the testimony, the accuracy and reliability of the techniques utilized in applying the principles to the subject matter before the court and in reaching the conclusion expressed in the opinion, and the qualifications of those
actually gathering the data and analyzing it. See Phillips, 615 P.2d at 1235.
In fine, the trial court should carefully explore each logical link in the chain that leads to the expert testimony given in court and determine its reliability. Only with such information can the overall decision on admissibility be made intelligently. In the absence of such a showing by the proponent of the evidence and a determination by the court as to its threshold reliability, the evidence is inadmissible. See Phillips, 615 P.2d at 1233-36; Kofford, 744 P.2d at 1347; State v. Maule, 35 Wash.App. 287, 295-96, 667 P.2d 96, 100 (1983); In re Amber B., 191 Cal.App.3d 682, 691, 236 Cal.Rptr. 623, 629 (1987).
The difficulty presented by the record in the instant case is that the trial court did not follow the approach dictated by Phillips and the State made no real effort to lay a foundation that would permit a determination of reliability. Instead, the court appears to have accepted the State’s argument that reliability concerns went only to the weight to be given the evidence, not to its admissibility. As noted above, we reject that view. But even though the trial court followed the wrong standard, is there sufficient record evidence so that we can say Phillips was satisfied and that an adequate foundation was laid for a determination that the evidence was reliable? The answer is no.
Dr. Tyler was the first expert witness to testify. When defense counsel called for a proffer concerning Dr. Tyler’s testimony, the prosecutor stated that Dr. Tyler would *404render an opinion on “[wjhether or not [the daughter] has been sexually abused, number one, whether [the daughter] in Dr. Tyler’s opinion is a victim of incest, specifically.” During the short voir dire examination, Dr. Tyler explained that her conclusions would be based upon psychological test results and the interviews with the daughter. She did not describe the profile of a sexually abused child, and she admitted that she knew of no studies testing the hypothesis that those diagnosed as sexually abused through the use of a psychological profile were, in fact, abused. Dr. Tyler further conceded that conformance with the profile alone was not enough to support an opinion on abuse. According to her testimony, such an opinion depended heavily on the subjective impressions of the interviewer gained through the interviews with the alleged victim.
At the conclusion of the voir dire, the State asked the court to rule that Dr. Tyler was an expert and could give the proffered opinion. The court made such a finding, over the objection of defense counsel.14 No additional foundation for the sexually abused child profile testimony was laid by any of the other expert witnesses. In each case, defense counsel renewed his foundational objection to the opinion testimony and had that objection overruled.
Although the court made a finding that Dr. Tyler and the other expert witnesses were experts, it made no determination that the scientific principles and techniques upon which their testimony was to be grounded were reliable. This is understandable because there was no evidence before the court from which any such determination could be made. In fact, the little information given about the strength of the profile by Dr. Tyler should have led the court to conclude that the profile testimony was insufficiently determinative of abuse to be found reliable enough to pass the threshold admissibility requirement of Phillips, which is subsumed by rule 702.
For the foregoing reasons, we find that the trial court violated Utah Rule of Evidence 702 when it admitted the expert testimony founded upon the psychological profile.15
*405We next address defendant’s contention that the trial court also erred in permitting the experts to base their opinions on abuse in part on their assessments of the daughter’s veracity.
Each expert, in explaining the basis for his or her opinion that the daughter had been abused, told the court that he or she relied not only on the daughter’s conformance with an abuse profile, but also on his or her individual appraisal of the daughter’s demeanor and affect during interviews in, which she told her version of the events. We have already determined that there is insufficient foundation in the record to permit a finding of the necessary threshold reliability of the abuse profile. The next question is whether the experts’ appraisal of the daughter’s credibility is based on scientific principles or techniques that are sufficiently reliable to permit the admission of testimony based upon that appraisal.
It is helpful to review the experts’ testimony on this particular point before analyzing its admissibility. Dr. Palmer had found the physical evidence to be inconclusive of abuse or even intercourse. After the examination, he interviewed the daughter about her sexual history, at which time she described the alleged incidents of abuse by her father. When asked by the prosecutor whether he thought that the daughter was a victim of incest, Dr. Palmer stated, “My opinion would be that based upon what she told me during the evaluation, the appropriateness of her affect m my experience that she has been the victim of sexual abuse within her family.” (Emphasis added.)
Mr. Harrison was a member of a team of psychologists and social workers who evaluated the daughter in 1983. The psychological tests and interviews constituting this evaluation showed symptoms that were consistent with not only sexual abuse but other causes and types of disorders. The daughter did not allege sexual abuse at that time, and the team did not conclude that she had been abused.
In 1984, Mr. Harrison again interviewed the daughter at the request of the prosecution. This time, the daughter did allege sexual abuse. During the course of the interview he had her replicate the sex acts with anatomically correct dolls.16 Mr. Harrison concluded that she was a victim of incest. When asked to explain what constituted the basis for this opinion, Mr. Harrison stated:
The information [the daughter] gave me ... from the interview, that she was able to not just give me cognitive, verbal data, not just able to tell me about an act, but she was able to put it in the context of a relationship, a family, a chronological order, and she showed the appropriate affect.
Dr. McManemin administered psychological tests to the daughter and interviewed her. She opined that the daughter had *406been sexually abused by her father. Dr. McManemin stated her expert opinion as follows: “My opinion based on the test data is that [the daughter] has been sexually abused repeatedly. This information was clarified by the statement she made that the perpetrator was her father. I, therefore, believe that [she] has been sexually abused by her father repeatedly.” 17 When asked the basis for her expert opinion, Dr. McManemin discussed the results of psychological tests and then stated: “I was also relying on the way [the daughter] related during the interviews, on her ability to convey information, on her ability to be open, on her affect, in general how she came across and how trustworthy she came across.” (Emphasis added.)
Dr. Tyler’s testimony has been reviewed in some detail earlier. Her opinion that the daughter had been abused and that the father was the abuser was grounded largely on her appraisal of the daughter’s credibility. In fact, she testified that absent the interviews, she would not have had enough information to form an opinion.
The foregoing demonstrates that for each expert, his or her opinion that the daughter had been abused and that the father was the perpetrator was based on a subsidiary opinion that the daughter was telling the truth when she made statements about the abuse. The admissibility of this opinion must be appraised under rule 702 just as was the profile-based opinion. The first question is whether the testimony was founded upon scientific principles or arrived at through scientific techniques that satisfy the requirement of inherent reliability; if so, then we proceed to balance the probativeness of the evidence against its potential for unfair prejudice. See supra note 8.
On the first issue, there is no basis for taking judicial notice of the reliability of the principles or methodologies underlying the experts’ credibility appraisals. As a general matter, scientific expert testimony that purports to determine whether a witness is truthful on a particular occasion is not admissible, largely because there has been no demonstration that such a determination can be accurately made. See, e.g., State v. Rebeterano, 681 P.2d 1265, 1268-69 (Utah 1984) (polygraph examination results admitted only on stipulation waiving objection to admissibility); State v. Tuttle, 106 Utah Adv.Rep. 6, 10 (April 12, 1989) (hypnotically enhanced testimony not admissible); State v. Rammel, 721 P.2d 498, 500-01 (Utah 1986) (error to admit testimony of police officer as expert on likelihood of suspect lying).
On a more specific level, nothing has come to our attention suggesting a general acceptance of the proposition that those who regularly treat symptoms of sexual abuse are capable of determining with a high degree of reliability the truthfulness of allegations that one has been abused. In fact, the better view of the matter is to the contrary. As Professor Myers states in his treatise, “The more persuasive decisions reject expert testimony on the truthfulness of allegations of abuse or on the credibility of a particular child.” Myers at 161; see id. at 166. Cases that support this position are plentiful. See, e.g., State v. Lindsey, 149 Ariz. 472, 475-76, 720 P.2d 73, 76 (1986); State v. Taylor, 663 S.W.2d 235, 239 (Mo.1984) (en banc); State v. Fitzgerald, 39 Wash.App. 652, 656-57, 694 P.2d 1117, 1121 (1985); Kruse v. State, 483 *407So.2d 1383, 1387-88 (Fla.Dist.Ct.App.1986). The Arizona Supreme Court put it well when it stated: “Psychologists and psychiatrists are not, and do not claim to be, experts at discerning truth. Psychiatrists are trained to accept facts provided by their patients, not to act as judges of patients’ credibility.” State v. Moran, 151 Ariz. 378, 385, 728 P.2d 248, 255 (1986) (citing People v. Bledsoe, 36 Cal.3d 236, 250, 681 P.2d 291, 300, 203 Cal.Rptr. 450, 459 (1984); United States v. Azure, 801 F.2d 336, 341 (8th Cir.1986)).
It should not be surprising that those who undertake to treat persons who may have suffered sexual abuse have no peculiar competence to judge the credibility of their patients. Symptoms may be treated effectively without knowing for certain their precise cause. This is not to dene-grate the talents of experts such as those who testified in this case. Working hypotheses relied upon by therapists may be useful for treatment purposes; however, that does not mean that they are reliable enough to be used for forensic purposes. See McCormick § 206, at 636 n. 97; Georgetown Note at 452.
Since the inherent reliability of the principles underlying the credibility determinations of the experts in this case have not been shown to be eligible for judicial notice, the next question is whether the record contains enough foundation to permit us to conclude that the evidence was properly admitted. Under our decisions in Phillips and Kofford, the burden is on the party proffering the evidence to demonstrate that it has the requisite degree of reliability. In the present case, the State made no effort to show that these experts were capable of reliably determining whether the daughter was telling the truth either from the content of the story or from the daughter’s affect when telling it. And the testimony of the experts themselves does not support such a finding. They only testified that they had experience dealing with abused children and that on the basis of that experience, they had formed an opinion that the daughter had been abused based in part on their judgment of her credibility.
The trial court admitted the evidence over objection, apparently on the erroneous assumption noted earlier, that the lack of foundation went to the weight, not to the admissibility, of the evidence. As a result, a foundation demonstrating the necessary reliability is entirely lacking. We are therefore forced to conclude that the trial court erred in permitting the experts to give opinions on abuse based on their purportedly scientific appraisals of the daughter’s truthfulness when she made her allegations. This testimony should have been excluded under rule 702.
We now consider the harmfulness of the evidentiary errors. If, in the absence of the evidentiary errors, there is a reasonable likelihood of a more favorable outcome for defendant, we must reverse the conviction. State v. Verde, 770 P.2d 116, 122 (Utah 1989); State v. Bell, 770 P.2d 100, 106 (Utah 1988); State v. Hackford, 737 P.2d 200, 204 n. 1 (Utah 1987); State v. Knight, 734 P.2d 913, 919-20 (Utah 1987); Utah R.Evid. 103; Utah R.Crim.P. 30 (codified at Utah Code Ann. § 77-35-30 (1982)).
Erroneous admissions of evidence are not as critical in a bench trial as where a jury is involved;18 however, the errors in this case had a pervasive impact on the trial. This case hinged on a determination of credibility. The daughter accused her father of sexual abuse. There was no physical evidence to corroborate the daughter’s charges. The father denied the accusations, and witnesses testified in support of his version of the facts. On the other hand, the daughter’s version of the events wqs bolstered principally by the testimony of the four experts challenged here. Their testimony was the primary focus of the trial. In substance, all four experts told the court that in their respective opinions, the daughter had been sexually abused by *408her father. They testified that their opinions were the product of their expertise and that they were grounded on an evaluation of the daughter’s truthfulness made during sessions in which she described the alleged abuse and on her conformance with a profile of sexually abused children. This expert testimony was admitted without an adequate foundational showing as to the reliability of the scientific principles and techniques upon which it was based and without any balancing of its probative value against its danger of unfair prejudice.
The foregoing factors alone almost dictate the conclusion that the improperly admitted testimony had a substantial impact on the verdict. Our judgment that the improper expert testimony likely affected the outcome is confirmed by the trial judge’s statement at the conclusion of the case. “It is the opinion of the Court that the State through the testimony of the alleged victim together with corroborating testimony of the expert witnesses called by the State has made out a prima facie case against the defendant and in favor of a verdict of guilt.” (Emphasis added.) Accordingly, we find that absent the error in admitting the testimony of the experts, there is a reasonable likelihood of a result more favorable to Rimmasch. The conviction must be reversed and the matter remanded for retrial.
HOWE, Associate C.J., and STEWART, J., concur.. Because our resolution of these issues is determinative, we do not address defendant’s other contentions.
. E.g., United States v. Azure, 801 F.2d 336, 341 (8th Cir.1986); United States v. Binder, 769 F.2d 595, 602 (9th Cir.1985); State v. Moran, 151 Ariz. 378, 382-86, 728 P.2d 248, 252-56 (1986); Johnson v. State, 292 Ark. 632, 639-40, 732 S.W.2d 817, 819-21 (1987); People v. Roscoe, 168 Cal.App.3d 1093, 1098-99, 215 Cal.Rptr. 45, 49-50 (1985); Tevlin v. People, 715 P.2d 338, 341 (Colo.1986); Wheat v. State, 527 A.2d 269, 275 (Del.1987); Kruse v. State, 483 So.2d 1383, 1387-88 (Fla.Dist.Ct.App.1986); Simmons v. State, 504 N.E.2d 575, 579 (Ind.1987); State v. Myers, 382 N.W.2d 91, 97 (Iowa 1986); Commonwealth v. Carter, 9 Mass.App. 680, 681-82, 403 N.E.2d 1191, 1193 (1980), aff’d, 383 Mass. 873, 417 N.E.2d 438 (1981); People v. Walker, 150 Mich.App. 597, 389 N.W.2d 704, 707 (1985); State v. Jackson, 239 Kan. 463, 470, 721 P.2d 232, 238 (1986); State v. Myers, 359 N.W.2d 604, 611 (Minn.1984); State v. Taylor, 663 S.W.2d 235, 239 (Mo.1984); In re J.W.K., 724 P.2d 164, 166 (Mont.1986); Townsend v. State, 103 Nev. 113, 734 P.2d 705, 709 (1987); People v. Reid, 123 Misc.2d 1084, 1087, 475 N.Y.S.2d 741, 743 (1984); State v. Staples, 120 N.H. 278, 281-82, 415 A.2d 320, 322 (1980); State v. Heath, 316 N.C. 337, 341 S.E.2d 565 (1986); State v. Milbradt, 305 Or. 621, 628-29, 756 P.2d 620, 624 (1988); Commonwealth v. Seese, 512 Pa. 439, 517 A.2d 920, 922 (1986); State v. Castore, 435 A.2d 321, 326 (R.I.1981); State v. Fitzgerald, 39 Wash.App. 652, 656-57, 694 P.2d 1117, 1121 (1985); State v. Haseltine, 120 Wis.2d 92, 96-97, 352 N.W.2d 673, 676 (1984); Brown v. State, 736 P.2d 1110, 1115 (Wyo.1987); see United States v. Earley, 505 F.Supp. 117 (S.D.Iowa 1981); State v. Black, 537 A.2d 1154, 1156-57 (Me.1988); State v. Buell, 22 Ohio St.3d 124, 489 N.E.2d 795, 803-04, cert. denied, 479 U.S. 871, 107 S.Ct. 240, 93 L.Ed.2d 165 (1986); State v. Logue, 372 N.W.2d 151, 157 (S.D.1985); see also Feeney, Expert Psychological Testimony on Credibility Issues, 115 Mil.L.Rev. 121, 134-35 (1987) (discussing same prohibition in military courts).
. The defense did not call an expert witness to rebut the profile testimony of these experts. *394The fact that there were no opposing experts does not justify an inference that the experts’ testimony was reliable. Rather, the lack of defense experts highlights the importance of imposing on the proponent of scientific evidence the burden of establishing a proper foundation as a prerequisite to admissibility, especially in a criminal case. The prosecution has ready access to expert witnesses, whereas criminal defendants are generally "‘without the economic means to marshall scientific witnesses for a battle of the experts.’ ” Giannelli, The Admissibility of Novel Scientific Evidence: Frye v. United States, A Half Century Later, 80 Colum.L.Rev.1197, 1244 (1980) [hereinafter Giannelli] (quoting State v. Williams, 388 A.2d 500, 506 (Me.1978) (Nichols, J., concurring)).
. Among the behavioral and psychological problems the daughter was experiencing in 1983 were the following: acting out at school; low self-esteem; confused sexual identity; anger; difficulty with peers; difficulty in resolving issues of closeness and bonding; and love/hate relationship with parents.
. For a general discussion of the various divergent trends of departure from the Frye test, see Giannelli, supra note 3, at 1228-1231; Note, United States v. Downing: Novel Scientific Evidence and the Rejection of Frye, 1986 Utah L.Rev. 839, 840-44 [hereinafter Utah Note]; McCormick, McCormick on Evidence § 203, at 606-08 (E. Cleary, 3d ed. 1984) [hereinafter McCormick].
. There is significant dispute among the courts and commentators, paralleling that between the parties in this case, over whether relevant scientific evidence should be viewed like any other and excluded only when the potential for unfair prejudice concerns of rule 403 preponderate or, on the other hand, whether some significant threshold requirement of reliability must be met — whether it is the Frye standard or some other lesser test — before the rule 403 analysis is performed. See, e.g., 2 G. Joseph, S. Salzburg & Trial Evid.Comm. of the A.B.A. Section of Litigation, Evidence in America: The Federal Rules in the States § 51.5 (1987) [hereinafter Evidence in America ]; Utah Note, supra note 5, at 840-44. In the present case, the State is pressing the former position, and defendant, the latter.
To the extent that my separate opinion in Kofford, 744 P.2d at 1361, suggests endorsement of the position advocated by the State, I withdraw that endorsement. Further reflection convinces me that there is a continuing need for a threshold reliability requirement applicable to scientific evidence to ensure that trial courts do not uncritically accept as sufficiently trustworthy for submission to the trier of fact the supposedly scientific assertions of anyone who can qualify as an expert in a particular field of scientific endeavor. The Frye test may be too rigid to be the only threshold reliability test, as critics have suggested and as Phillips acknowledged, but the more flexible test articulated in Phillips seems fully capable of performing the necessary screening function without unduly impeding the flow of reliable scientific evidence to the fact finder.
. Of course, even if the court ultimately decides, either on grounds of judicial notice or after an evidentiary hearing, that there is an adequate foundational basis for finding that the scientific principles or techniques are inherently reliable, that determination will not resolve the question of whether a particular piece of expert testimony may be admitted. The trial court must still make a separate determination that there is an adequate foundation for the proposed testimony, i.e., that the scientific principles or techniques have been properly applied to the facts of the particular case by qualified persons and that the testimony is founded on that work. See Kofford, 744 P.2d at 1354, 1356.
. Of course, once the threshold requirement of inherent reliability imposed on scientific evidence is satisfied, expert testimony based upon it is not automatically entitled to admission. There remains the question that must be posed when any expert evidence, scientifically based or otherwise, is offered under rule 702: whether, on balance, the evidence will be helpful to the finder of fact. This determination requires the trial court under the guidance of rule 403 of the Utah Rules of Evidence to balance the pro-bativeness of the proffered evidence against the dangers its admission poses. That rule provides:
Although relevant, evidence may be excluded if its probative value is substantially out*399weighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence.
In making this determination, the relative probative value of the proffered scientific evidence of a fact in issue becomes important. For example, if the scientific proof is based on undeniably valid scientific premises, has a high degree of power to accurately determine the existence or nonexistence of a fact in issue, and is easily replicable and its application to similar situations has been tested and validated often, then the dangers of unfair prejudice, confusion of the issues, misleading the jury, etc., attendant to its introduction would have to be great indeed to preclude its admission. See Phillips v. Jackson, 615 P.2d at 1236-37. But if there were weaknesses in the testimony on some or all of these points, then it would be relatively easier to show that the dangers of admission outweighed the probativeness of the testimony.
Similarly, the potential for unfair prejudice, etc., posed by the admission of various types of expert scientific evidence can vary widely and must be considered in making the helpfulness determination. Among the important variables are the nature of the evidence offered, the quality of the other evidence available to the finder of fact, and the centrality of the issue to which the scientific evidence is directed. For example, when the principles underlying scientific evidence are easily demonstrable or are readily understood by lay persons, there is relatively less danger that the finder of fact will be confused by the presentation or unduly impressed with the apparent “scientific” nature of the evidence. McCormick at 609; 3 J. Weinstein & M. Berger, Weinstein’s Evidence ¶ 702[03], at 702-42 to -43 (1988) [hereinafter Weinstein]. This may justify the admission of evidence with relatively less probative power. "On the other hand, when the nature of the technique is more esoteric, as with some statistical analyses and seriologic tests, or when the inferences from the scientific evidence sweep broadly or cut deeply into sensitive areas, a stronger showing of probative value should be required." McCormick at 609. Such a "sensitive area” is one central to the core of the fact-finding process — whether one witness or another is telling the truth. Id.; Evidence in America § 51.3.
. It has been suggested that any test focusing on the threshold reliability of "soft” scientific evidence may tend to unduly restrict the admission of such evidence. McCord, Syndromes, Profiles and Other Mental Exotica: A New Approach to the Admissibility of Nontraditional Psychological Evidence in Criminal Cases, 66 Or.L.Rev. 19, 82-88 (1987) [hereinafter McCord, A New Approach ]. A multi-factor balancing approach to determining "helpfulness” has been proposed as an alternative. Id. at 94-101. Justice Durham has adopted this approach in her separate opinion. It is little more than a sophisticated version of the rule 403 balancing test advocated by the State here. As such, it is readily susceptible of being applied in a fashion that would admit proffered expert evidence on nothing more than a mere relevance showing. Our approach has the practical advantage of very explicitly focusing the attention of the court and the parties on a point that is often slighted by nonscientifically trained judges and lawyers in the rush to accept the views of "experts," i.e., the strength of the scientific premises upon which they base their testimony. See Golding, Mental Health Professionals and the Courts: The Ethics of Expertise, at 18-26 (to be published in_Int’l J.L. Psychiatry (1989)) [hereinafter Golding, Mental Health Professionals and the Courts],
. The views of a number of these scholars are set out at length in McCord, Expert Psychological Testimony About Child Complainants in Sexual Abuse Prosecutions: A Foray into the Admissibility of Novel Scientific Evidence, 77 J.Crim.L. & Criminology 1, 18-24 (1986) [hereinafter McCord, 77 J.Crim.L. & Criminology]; Note, The Unreliability of Expert Testimony on the Typical Characteristics of Sexual Abuse Victims, 74 Geo.L.J. 429, 440-43 (1985) [hereinafter Georgetown Note]; and in J. Myers, Child Witness Law and Practice § 4.15, at 151-52 & n. 107 (1987) [hereinafter Myers]. See also Golding, Mental Health Professionals and the Courts, supra note 6, at 32 n. 8 ("Currently, absent certain strong physical signs (e.g. presence of venereal disease or blatant vaginal or anal trauma), there are no scientifically acceptable scientific data on 'sexual abuse profiles’ based upon psychological data.")
. While such generally worded and overinclu-sive psychological profiles may be of little utility in a forensic setting for determining whether abuse has actually occurred, that does not mean that they may not be of use in a therapeutic setting when the person suffering from these symptoms is being treated. See McCormick, supra note 5, at 636 n. 97; Georgetown Note, supra note 10, at 452.
. The only other evidence cited in support of the experts’ opinions on abuse, besides their judgment of the daughter’s veracity, was Mr. Harrison’s reference to the daughter’s ability to demonstrate the alleged acts of abuse with anatomically correct dolls and Dr. Palmer’s conclusion after an examination that his physical findings were consistent with sexual activity. It is hard to understand how either fact can be probative of whether the daughter was abused at an earlier age when the subject at the time of the physical examination and the doll play is a seventeen-year-old girl who has admittedly been intimate with males of her own age. See Myers, supra note 10, at 158.
. Professor Myers, an avowed "child advocate,’’ has stated that opinions of abuse that are based on just such generalized symptoms are not logically supportable.
The most casual examination of these symptoms reveals, however, that many of them are associated with other developmental and psychological problems of childhood and adolescence. For example, the fact that a child suffers from nightmares, loss of appetite, regression, and depression says very little, if anything, about sexual abuse. A myriad of other factors can cause such symptoms, and it would be improper for an expert to base an opinion relating to sexual abuse on such ambiguous symptoms alone.
Myers, supra note 10, at 157; accord Golding, Mental Health Professionals and The Courts, supra note 9, at 32 n. 28; Georgetown Note, supra note 10, at 442 (quoting Rosenfeld, The Clinical Management of Incest and Sexual Abuse of Children, 22 Trauma 2, 3 (Oct.1980)), and Johnson, The Sexually Mistreated Child: Diagnostic Evaluation, 3 Child Abuse and Neglect 943, 947 (1979); McCord, supra note 10, at 23. Myers goes on to state that in the context of a prosecution for sexual abuse of a child, "a finding of sexually abused child syndrome [based on such ‘highly ambiguous symptoms'] should be regarded as of de minimis evidentiary value but of great potential prejudice.” Myers, supra note 10, at 158.
Profile testimony that portrays the characteristics of the typical victim of sexual abuse also has a tendency to mislead and confuse a finder of fact by suggesting that the issue to be decided is whether the accusing child possesses these characteristics, rather than whether the child experienced the specific instances of abuse described. See State v. Rammel, 721 P.2d 498, 501 (Utah 1986) ("Even where statistically valid probability evidence has been presented],] ... courts have routinely excluded it when the evidence invites the jury to focus upon a seemingly scientific, numerical conclusion rather than to analyze the evidence before it and decide where the truth lies.”); State v. Lindsey, 149 Ariz. 472, 474-75, 720 P.2d 73, 75 (1986) (expert’s opinion as to credibility may not be indirectly adduced by allowing the expert to quantify the percentage of victims who are truthful in their initial reporting of abuse despite subsequent recantation); Powell v. State, 527 A.2d 276, 279 (Del. 1987) (admission of expert testimony that 90 percent of child abuse victims were telling the truth was error).
Furthermore, the issue in cases where child sexual abuse has been charged is not whether the child has been abused, but whether the child has been abused by this particular defendant. See Bussey v. Commonwealth, 697 S.W.2d 139, 141 (Ky.1985) (where evidence showed that child may have been abused by someone other than defendant, profile testimony was immaterial). When the fact finder is allowed to focus on the child’s psychological characteristics rather than on the alleged incidents of sexual abuse, the danger exists that the defendant will be convicted merely because the child appears to have been sexually abused; the fact finder will have a tendency to presume that the person on trial was the abuser. In State v. Saldana, 324 N.W.2d 227 (Minn.1982), a rape prosecution, the court stated:
The jury must not decide this case on the basis of how most people react to rape or on whether Fuller’s reactions were the typical reactions of a person who has been a victim of rape. Rather, the jury must decide what happened in this case, and whether the elements of the alleged crime have been proved beyond a reasonable doubt.
Id. at 230.
*403In fairness to the experts here, Mr, Harrison and Dr. Tyler both acknowledged that the symptoms upon which they based their finding of conformance with the profile were ambiguous and could be attributable to other causes. Indeed, Mr. Harrison had done just that in 1983. And Dr. Tyler testified that conformance with the profile alone would not justify concluding that abuse was more likely than not the cause of the daughter’s psychological and behavioral problems. Unless the symptoms and characteristics manifested by a particular child are likely caused only by sexual abuse, it is a faulty syllogism to permit an expert to base an opinion that the child was sexually abused on such evidence. Myers, supra note 10, at 158. For example, in In re Cheryl H., 153 Cal.App.3d 1098, 200 Cal.Rptr. 789 (1984), an expert testified that the three-year-old complaining witness had exhibited behavior with anatomically correct dolls and talked about sex in ways that indicated that she was a victim of sexual abuse. The appellate court noted that such behavior "appears to be unique to children subjected to child abuse." 153 Cal.App.3d at 1117, 200 Cal.Rptr. at 800. If the court was correct and if a foundation was established showing that such behavior is unique to child victims of sexual abuse, then it was logical to conclude that a child manifesting that behavior was a victim of sexual abuse, and the court propefly admitted such testimony. See Myers, supra note 10, at 158.
. The court's ruling was made in the course of the following discussion:
Mr. O’Connell: Maybe with the expertise, okay. I think she is probably an expert in the field. I do not think that maybe she can draw this sort of a conclusion. I would object to her drawing the conclusion until we know that it is a conclusion based on some kind of data that one can draw a conclusion. I don’t know there’s any foundation on that. I don’t dispute she is a psychologist or she studied in this area.
The Court: I think her testimony to this has been that she can logically draw that kind of opinion and conclusion from the data.
[[Image here]]
Ms. Lewis: Could I ask the Court, then, at this time to make a finding with reference to this witness’ expertise and ability based upon that to render an opinion.
The Court: All right. I will make that sort of finding that, based upon her education, experience, she’s an expert in the field of clinical psychology, that's what you are?
The Witness: I am a licensed marriage and family therapist with training in psychology.
The Court: That, then.
Mr. O'Connell: Just for the record I object to the finding, if it goes to — that she is qualified to render a conclusion as to whether or not sexual abuse actually occurred in this case without further foundation.
The Court: Your objection will be noted. You may proceed.
. As noted earlier, even if a finding of sufficient reliability is made to satisfy the threshold requirements of the Phillips/KofforcL standard, before expert evidence based on scientific principles is admitted, the trial court still must appraise the potential for unfair prejudice and the likelihood that it outweighs the probativeness of the evidence. See supra note 8. In performing that evaluation when profile evidence is proffered, the court must consider not only the probative strength of the profile, but also the means used to determine the characteristics of the victim that are used to compare him or her with the profile. In making this determination, experts necessarily rely heavily on subjectively evaluated psychological tests, such as the Thematic Apperception Test and the Rorschach Inkblot Test used in the present case. The reliability of these tests is not universally accepted. See, e.g., J. Ziskin, Coping with Psychiatric and Psychological Testimony (2d ed. 1975 & Supp. 1977); Comment, The Psychologist as Expert Witness: Science in the Courtroom?, 38 Md.L. Rev. 539, 565-88 (1979) (citing studies about the validity and reliability of psychological testing and diagnoses). Therefore, the admission of profile evidence will necessarily require an appraisal of the relative ability of the expert to determine the characteristics upon which the comparison with the profile is made.
*405As has been noted by one authority, “[ojpin-ions which are based in large measure on a subjective analysis may have less probative value because it may be difficult to evaluate the skill of the expert in extrapolating a judgment from the scientific data.” Weinstein, supra note 8, at 702-42 (citing United States v. Williams, 583 F.2d 1194, 1198 (2d Cir.1978) (spectrography is qualitatively different than polygraph evidence, because in polygraph analysis, the examiner must extrapolate from the data a judgment of something that is not directly measured by the polygraph test, whereas spectrography involves no such subjective interpretations of the data), cert. denied, 439 U.S. 1117, 99 S.Ct. 1025, 59 L.Ed.2d 77 (1979)); see also McCormick, supra note 5, at 609-10.
. Under cross-examination, Mr. Harrison admitted that the mere fact that a seventeen-year-old, sexually active girl could represent sex acts with anatomically correct dolls was not, in and of itself, as significant as it would be if a younger child had that ability. Mr. Harrison’s testimony shows that the daughter was asked to use the dolls not to determine whether she knew about sex, but as a test of her credibility, i.e., to see whether she would become confused in trying to illustrate her narrative. In dealing with a much younger child, anatomically correct dolls might be used to test something other than credibility, specifically, whether the child has a knowledge of sexual matters beyond the norm. The opinion for two members of the Court in State v. Lairby implied that expert testimony concerning a child’s verbal indicators of sexual knowledge beyond the norm might be admissible. 699 P.2d at 1200, 1201 & n. 15; see supra note 12. Of course, a proper foundation would have to be laid.
. It is worth noting that the ability to determine who abused a child based only on the statements of the child appears to be far beyond the scope of expertise claimed by the fields of psychiatry and psychology.
In In re Cheryl H., 153 Cal.App.3d 1098, 200 Cal.Rptr. 789 (1984), the court stated:
Dr. Powell’s opinion that Cheryl's father was the abuser extends the subject matter of a psychiatrist’s expert opinion far beyond anything that has yet been approved by California courts. It assumes peering into the mind of one person allows a psychiatrist not only to draw valid conclusions about the mental state and behavior of that person but to draw equally valid conclusions about the conduct of another person.
153 Cal.App.3d at 1118, 200 CaI.Rptr. at 801. The court went on to conclude that the psychiatrist’s opinion that the father abused the child was inadmissible. Dr. McManemin's opinion and that of the other experts purporting to identify the culprit would appear to be subject to the same objection.
. Sweeney v. Happy Valley, Inc., 18 Utah 2d 113, 118, 417 P.2d 126, 130 (1966); In re Baxter's Estate, 16 Utah 2d 284, 288, 399 P.2d 442, 445 (1965). However, erroneous admissions of evidence in a bench trial may rise to the level of reversible error.