483 So.2d 1383 (1986)
Frank KRUSE, Appellant,
v.
STATE of Florida, Appellee.
No. 83-2364.
District Court of Appeal of Florida, Fourth District.
February 5, 1986.
Rehearing and Rehearing Denied March 26, 1986.
Gwendolyn Spivey, Tallahassee, for appellant.
Jim Smith, Atty. Gen., Tallahassee, and Richard G. Bartmon, Asst. Atty. Gen., West Palm Beach, for appellee.
Rehearing and Rehearing En Banc Denied March 26, 1986.
ANSTEAD, Judge.
This is an appeal by Frank Kruse from his convictions and sentences on two counts of lewd, lascivious, or indecent assault upon a child. We reverse and remand for a new trial because we believe errors were committed at trial that cannot be deemed harmless.
It is initially contended that the trial court erred by admitting the opinion testimony of Dr. Donna Holland, an expert in child and adolescent psychiatry, that the child-victim was suffering from a condition known as Post Traumatic Stress Syndrome. It is agreed by the parties that there was no demonstrable physical evidence of an assault on the alleged victim and that Kruse, who denied the allegations, and the victim were the only direct witnesses on the issue of whether an assault occurred. The child was seven years of age at the time of the alleged assault. Dr. Holland was allowed to describe the condition known as Post Traumatic Stress Syndrome and to correlate her observations of the victim's behavior with commonly observed behavior patterns of other Syndrome patients. She testified to the details of the victim's and the victim's parents' statements to her, including the victim's identification of Kruse as the assailant. This same information was revealed in testimony presented by the parents and victim at trial. Dr. Holland concluded that, in her opinion and based upon her psychiatric examination and the history of the child's behavior before and after the alleged assault, the child had suffered a sexual trauma.
The Florida Evidence Code became effective in criminal cases in 1979. Sections 90.401 and 90.402, Florida Statutes (1983), set out a general relevancy standard for the admission of evidence. Sections 90.702 and 90.703 deal specifically with expert testimony:
90.702 Testimony by experts. If scientific, technical, or other specialized knowledge will assist the trier of fact in understanding the evidence or in determining a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education may testify about it in the form of an opinion; however, the opinion is admissible only if it can be applied to evidence at trial.
90.703 Opinion on ultimate issue. Testimony in the form of an opinion or inference otherwise admissible is not objectionable because it includes an ultimate issue to be decided by the trier of fact.
In addition, section 90.403 provides:
90.403 Exclusion on grounds of prejudice or confusion. Relevant evidence is inadmissible if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of issues, misleading the jury, or needless presentation of cumulative evidence. This section shall not be construed to mean that evidence of the existence of available third-party benefits is inadmissible.
Section 90.702 contains three requirements: (1) that the opinion evidence be helpful to the trier of fact; (2) that the witness be qualified as an expert; and (3) that the opinion evidence can be applied to evidence offered at trial. These provisions embody a liberal policy on the admission of expert evidence, generally rendering such evidence admissible to the extent that it is helpful to the trier of fact. Section 90.403 adds a fourth test barring evidence that, although technically relevant, presents a substantial danger of unfair prejudice that outweighs its probative value.
In Brown v. State, 426 So.2d 76 (Fla. 1st DCA 1983), Judge Ervin discussed the evolution in Florida decisions, from a rigid test of admissibility of evidence relating to new scientific procedures, to the more generous relevancy standard contained in the evidence code. Id. at 85-90; see also Fay v. Mincey, 454 So.2d 587, 593-94 (Fla.2d DCA 1984), and Hawthorne v. State, 470 So.2d 770 (Fla. 1st DCA 1985) (Ervin, J., concurring in part and dissenting in part). The more rigid standard evolved from the decision in Frye v. United States, 293 Fed. 1013 (D.C. Cir.1923), which barred the admission of the results of a lie detector test because the test had not been generally accepted by the scientific community. Hence, the requirement of general acceptance was imposed. As Judge Ervin noted in his partial dissent in Hawthorne, the evidence code contains no reference to general acceptance in regard to the receipt of expert opinion evidence.
With some qualification, we believe the relevancy approach set out in the evidence code is the appropriate standard for determining the admissibility of expert testimony on child sexual abuse. The statutory relevancy standard also comports with the holdings of the Florida Supreme Court in the area of expert testimony. The court has stated that while trial courts have broad discretion in determining the range of subjects on which an expert may testify, such testimony should usually be received only where the disputed issue for which the evidence is offered, is beyond the ordinary understanding of the jury. Johnson v. State, 393 So.2d 1069, 1072 (Fla. 1980). This view is consistent with the first requirement of section 90.702, that the opinion evidence be helpful to the trier of fact, as well as the provisions of section 90.403, that the danger of prejudice may outweigh the value of the evidence.
The highest courts of other states have cited evidence code provisions similar to section 90.702 in support of the admissibility of expert testimony on child sexual abuse; see State v. Myers, 359 N.W.2d 604, 609 (Minn. 1984); and State v. Middleton, 294 Or. 427, 657 P.2d 1215, 1219 (1983). In Terry v. State, 467 So.2d 761 (Fla. 4th DCA 1985), we cited the evidence code in support of our holding that evidence relating to battered women's syndrome was relevant to a defendant's claim of self defense in a prosecution for manslaughter. 467 So.2d at 764. In so holding, we noted that such evidence would be helpful to the jury in interpreting the circumstances surrounding the incident as they affected the reasonableness of the defendant's belief that she was in danger. Id. A similar rationale was employed by the Nevada Supreme Court when it held, in a factual context similar to this, that expert testimony on child sex abuse was admissible to explain the behavior of the victim after the incident. Smith v. State, 688 P.2d 326, 327 (Nev. 1984). We do not believe the issue presented here is much different than that decided in Terry, and we conclude that the trial court did not err in ruling that the opinion evidence on the Post Traumatic Stress Syndrome met the requirements of the evidence code.
Initially, in view of the absence of physical evidence of assault and the age of the victim-witness, it appears that testimony about the syndrome would be helpful to the jury in providing more information from which to decide whether the child had been a victim of sexual abuse. The expert's testimony connected the alleged victim's change in behavior, which was described by the victim's parents, to the trauma of the sexual assault, which was described by the victim. Dr. Holland initially indicated, and then affirmed on cross-examination, that her opinion was predicated on the validity of the history given to her by the child and the parents. Accepting the validity of the testimony of the victim and the parents, she stated that the happening of the trauma would explain the change in behavior. Hence, the change of behavior could be utilized by the trier of fact as relevant evidence that the trauma did occur. While we also believe that jurors would have some ability to decide for themselves whether the child's behavioral changes may be related to the trauma, we do not believe that the implications are so easily understood as to bar the receipt of a psychiatric expert's analysis thereof. Cf. Johnson v. State. Accordingly, in our view the first requirement of section 90.702, that of helpfulness to the trier of fact, was met here.
We also believe Dr. Holland's testimony, which outlined her formal training and experience, and her licensing as a physician in two states, with a specialty in child and adolescent psychiatry, established her qualifications to render an opinion, the second requirement of section 90.702. The third requirement, that of relating the opinion evidence to other evidence offered at trial, was met because Dr. Holland's opinion was predicated upon and intended to explain the evidence offered at trial of the victim's change in behavior and the victim's claim of trauma at the hands of the appellant.
The fourth test for admissibility contained in the provisions of section 90.403 presents a more difficult issue: Is the probative value of Dr. Holland's testimony substantially outweighed by its potential prejudicial effect? We have already discussed to some extent the probative value of Dr. Holland's testimony in assessing its helpfulness as required by section 90.702. We must also examine its costs. We are aware, for example, of the danger that the trier of fact may place undue emphasis on evidence offered by an expert, simply because of the special gloss placed on that evidence by reason of the witness' status as an expert. A similar concern is that the jury may infer that the expert, simply by virtue of his appearance for one party, is vouching for the credibility of that party. These are dangers present in every case involving an expert and should perhaps be the subject of instructions to the jury. However, they are not themselves sufficient reasons to exclude opinion testimony. We have already noted that Dr. Holland predicated her opinion on the reliability of the history obtained. In other words, the diagnosis was dependent on the accuracy of the reports of the assault and the child's changes in behavior. The jury was left free, as it should have been, to determine the validity of those reports in accepting, rejecting and weighing the testimony of the parents, the child, and the expert.
Another danger is that the opinion evidence may be unreliable. In this regard, we reaffirm what we view to be a fundamental requirement that the party seeking to introduce expert testimony first establish that the subject can support an expert opinion with a reasonable degree of reliability. Expert testimony in areas that are not sufficiently developed to support an expert opinion can present the kind of danger that section 90.403 was designed to prevent. While there is no requirement to demonstrate general acceptance, we believe that, without some indicia of reliability, opinion evidence on a particular subject could hardly be helpful to a jury as required by section 90.702. We note the concern raised by other state courts regarding the scientific validity of using the results of research on various stress disorders as the basis for expert cause-effect testimony in a criminal justice context. For example, courts considering the admissibility of testimony on rape trauma syndrome involving an adult victim in a prosecution for rape have questioned whether the syndrome is actually relied upon by the medical community to establish that a rape had in fact occurred; see People v. Bledsoe, 36 Cal.3d 236, 203 Cal. Rptr. 450, 459-60, 681 P.2d 291, 300-01 (1984); State v. Saldana, 324 N.W.2d 227, 229-30 (Minn. 1982). Here, Dr. Holland testified to the widespread use among psychiatrists of the Post Traumatic Stress Syndrome as a descriptive term to identify the psychological problems and pattern of behavior of a child who has suffered a traumatic experience, such as a sexual assault. Her testimony established that the syndrome was actually used to diagnose and treat patients. We believe her unrebutted testimony sufficiently established the reliability of the method of diagnosing the syndrome and its use in the medical community to permit the expression of an opinion under the statutory relevance standard.
It is also contended that Dr. Holland's testimony that the victim had suffered a sexual assault or been abused invaded the province of the jury. In Farley v. State, 324 So.2d 662 (Fla. 4th DCA 1975), cert. denied, 336 So.2d 1184 (Fla. 1976), an appeal from a conviction for rape, we held that the trial court erred in admitting the opinion testimony of a physician expert that the alleged victim had been raped. Our opinion was predicated upon the view that an expert may not state an opinion that a criminal violation had occurred or that the defendant was guilty of such violation. 324 So.2d at 663; see also Spradley v. State, 442 So.2d 1039, 1043 (Fla. 2d DCA 1983). While an expert may testify to matters within his or her expertise, "it is not the function of an expert witness to draw legal conclusions." Palm Beach County v. Town of Palm Beach, 426 So.2d 1063, 1070 (Fla. 4th DCA 1983). However, in Farley, we were careful to note that in North v. State, 65 So.2d 77, 87 (Fla. 1952), the following opinion had been allowed:
"Because of the multiplicity, nature and distribution of various wounds on this body, I concluded that they were most consistent with the person having been assaulted, principally by blunt force, and that the method of assault is most consistent with strangulation." Id. at 87.
Farley v. State, 324 So.2d at 663.
Initially, we observe that no objection was made to the particular question and answer that is now challenged as invading the province of the jury. In any case, however, when considered in the context of her entire testimony, we believe Dr. Holland's opinions were consistent with those permitted in North and distinguishable from those found objectionable in Farley. The North view is also consistent with the provisions of section 90.703 of the evidence code, which provides that expert testimony is not rendered inadmissible merely because it includes an ultimate issue to be decided by the trier of fact.[1] The Third District has held that section 90.703 permits an expert to testify in a prosecution for rape that the victim's injuries were consistent with forced sexual intercourse. Ferradas v. State, 434 So.2d 24 (Fla. 3d DCA 1983). In this case, the jury obviously had to decide the factual issue of whether the victim was actually subjected to a sexual assault, an issue similar to that involved in North and Ferradas. We also note that this view is not novel, since medical experts have traditionally been allowed to testify on numerous issues ultimately to be decided by the jury. For example, in Florida, when insanity is raised as a defense, the issues are (1) the individual's ability at the time of the incident to distinguish right from wrong; and (2) his ability to understand the wrongness of the act committed. Gurganus v. State, 451 So.2d 817, 820 (Fla. 1984). Experts are routinely called to testify regarding the defendant's sanity in precisely these terms, see Johnson v. State, 408 So.2d 813 (Fla. 3d DCA 1982).
In our view, the opinion of Dr. Holland in this case, was more in the nature of a medical opinion that trauma was responsible for the child's behavioral problems, than a legal conclusion that a criminal act had occurred. As in Ferradas, the gist of Dr. Holland's testimony was that the victim's reported behavioral changes were consistent with her report of sex abuse. Based upon all the considerations discussed above, we cannot say that the probative value of the opinion evidence on the Post Traumatic Stress Syndrome offered here was so substantially outweighed by the danger of undue prejudice as to bar its admission under section 90.403.
By our holding, however, we by no means recede from the position that expert testimony may not be offered to directly vouch for the credibility of a witness. A concern which underlied our decision in Farley was that the expert testimony in that case would be construed as a direct comment on the credibility of a witness, the alleged victim. At one point in the trial, Dr. Holland indirectly testified to the child's credibility when she stated that she believed the child was able to tell fantasy from reality. One could infer from this statement that the expert was of the opinion that the child was truthful in relating her account of the sexual assault. However, no objection to this statement was raised at trial and as such the objection must be considered waived. Witt v. State, 388 So.2d 1 (Fla. 4th DCA 1980); DeLuca v. State, 384 So.2d 212, 213 (Fla. 4th DCA 1980). In our view direct testimony on the credibility of witnesses is inherently unreliable. We have already discussed some problems relating to a juror's treatment of expert testimony. Some bolstering of a party's credibility cannot be helped. For instance, it is not impermissible for a physician to relate a party's complaints and history and then offer an opinion predicated in part on that information. Part of the reason that opinion is allowed is because the factual predicate upon which it is based, usually provided by the testimony of the party, is itself directly subject to the jury's scrutiny. However, as with the opinion of so-called "lie detector" experts, we have consistently rejected scientific or expert determinations of credibility. Knight v. State, 97 So.2d 115, 119 (Fla. 1957); Goldstein v. State, 447 So.2d 903, 905 (Fla. 4th DCA 1984); Rodriguez v. State, 413 So.2d 1303, 1305 (Fla. 3d DCA 1982); Holliday v. State, 389 So.2d 679, 680 (Fla. 3d DCA 1980). We caution trial courts to be careful that such opinions are not put before juries, including the one that may be impaneled to try this case again.
Notwithstanding our disagreement with appellant as to the admission of testimony on the Post Traumatic Stress Syndrome, we agree that other errors occurred that mandate reversal. First, we agree that the court erred by allowing the state to cross-examine the appellant's character witness as to his knowledge of other specific arrests and accusations against appellant for the same type of crime. The rule is clear that the state may only rebut testimony on reputation for good moral character, by reputation testimony as to bad moral character, not by cross-examination about prior arrests or specific bad acts. Dixon v. State, 426 So.2d 1258 (Fla. 2d DCA 1983); Michaels v. State, 429 So.2d 338 (Fla. 2d DCA 1983); §§ 90.404(1)(a), 90.405, Fla. Stat. (1983). Here the state was allowed, through the guise of impeaching a character witness, to inform the jury that the appellant had been accused of other specific child sexual abuse crimes, evidence that is admitted to be irrelevant and inadmissible on the present charges, and of such a nature that we could hardly hold harmless.
We also agree that it was reversible error for the trial court to refuse to poll the jury to determine their exposure to two admittedly prejudicial and inaccurate media reports, published mid-trial and also referring to other charges pending against the appellant. This question was addressed in Robinson v. State, 438 So.2d 8 (Fla. 5th DCA), petition for rev. denied, 438 So.2d 834 (Fla. 1983), wherein it was held that a new trial was required where the trial court failed to take any action to determine whether the jurors had been exposed to and prejudiced by certain newspaper articles published after jury selection and relating to the appellant's trial. We agree with the holding in Robinson requiring a new trial as applied to the facts here.
In view of our resolution of the issues discussed above, it is unnecessary to address the other points on appeal. Accordingly, for the reasons set out above we reverse and remand for a new trial.
HURLEY, J., concurs.
DELL, J., concurs in part and dissents in part.
DELL, Judge, concurring in part and dissenting in part:
I agree that this case must be reversed and remanded for a new trial because the trial court admitted testimony concerning other criminal charges against appellant and failed to poll the jury regarding media articles related to those charges. I write this special concurrence to express my concern regarding the expert testimony of Dr. Holland. While I agree that testimony from a qualified expert witness may be introduced in a child abuse case to explain changes in a child's behavior and to relate such changes to the issue of whether that child has been subjected to child abuse, such testimony must be scrupulously limited to this narrow area of the witness's expertise.
Dr. Holland first related the child's history which included the child's accusation that appellant had sexually abused her. Dr. Holland further testified that in her opinion the child had been sexually assaulted and abused. This testimony, coupled with the history given by the child, established not only that a sexual assault had occurred, but that appellant had committed the crime.
Such an opinion is not permissible; Gibbs v. State, 193 So.2d 460 (2d DCA Fla. 1967):
"The opinion of a witness as to the guilt or innocence of an accused person is not admissible in evidence." Id. at 463.
Farley v. State, 324 So.2d 662, 663-64 (Fla. 4th DCA 1975), cert. denied, 336 So.2d 1184 (Fla. 1976).
Contrary to the majority opinion, I find no support in North v. State, 65 So.2d 77 (Fla. 1952), for the admission of Dr. Holland's testimony. The court in North permitted the testimony quoted by the majority because:
This question did not call for an answer from Dr. Mills as to whether or not there was a felonious assault or who made the assault.
65 So.2d at 87 (emphasis supplied).
Another problem occurred when Dr. Holland expressed her opinion that the child was not indulging in fantasies. Such testimony did not constitute mere bolstering of the child's testimony. Rather, it constituted a clear statement on the credibility of the child. The credibility of a witness is a question for the jury. See Farley v. State. In this case, Dr. Holland's testimony left nothing for the jury to decide.
I am not sure what the majority opinion will permit upon retrial, but if it can be interpreted to mean that such testimony may be admitted under the guise of an expert medical opinion, I must respectfully dissent.
NOTES
[1] Even prior to the enactment of the evidence code, experts in civil cases were allowed to offer opinions as to ultimate facts in issue; see Buchman v. Seaboard Coast Line Railroad Company, 381 So.2d 229 (Fla. 1980) ("human factors" expert in negligence case); see also School Board of Broward County v. Surette, 394 So.2d 147 (Fla. 4th DCA 1981) ("safety engineer" testified to dangerous condition of school bus stop). Such testimony is permissible so long as the trier of fact is not directed to arrive at a conclusion which it should be free to determine from the facts presented. Town of Palm Beach v. Palm Beach County, 460 So.2d 879, 882 (Fla. 1985).