LENK, J.

A Superior Court jury found the defendant guilty of two counts of indecent assault and battery of a child, and one count of rape of a child. The victims are the defendant's children, George and Heather, who were eleven and ten years old, respectively, at the time of trial, and nine and eight years old when the crimes were committed. The defendant complains that he suffered a substantial risk of a miscarriage of justice because the trial court erred in (1) denying his request for a pretrial hearing concerning the competency and reliability of the children, given the nature of the interviewing techniques employed by Commonwealth investigators; (2) excluding defense expert testimony concerning the interview techniques employed; (3) improperly admitting expert testimony that vouched for the children's credibility; (4) allowing both stale fresh complaint testimony and the fresh complaint testimony of a child abuse investigator who had observed an interview with the children through a two-way mirror; and (5) refusing to give certain defense instructions to the jury concerning the reliability of children's testimony. Because there was no error, we affirm.

1. Pretrial hearing regarding the impact of the Commonwealth's interview techniques on the child-victims' competency and reliability. The defendant does not argue that, under Massachusetts law, the children were not competent to testify. Rather, he urges us to adopt the reasoning of a New Jersey case, State v. Michaels, 136 N.J. 299 (1994), and to reverse the convictions because, as a result of "coercive and suggestive" questioning of the complainants by the Commonwealth during the investigation of the case, he was deprived of his due process rights under the Fifth and Fourteenth Amendments to the United States Constitution and art. 12 of the Massachusetts Declaration of Rights, when denied a pretrial evidentiary hearing to determine whether the children's testimony had been so tainted as to render it unreliable.

In support of his motions in limine seeking a pretrial competency hearing, the defendant offered (1) a videotape of one such interview, between Susan Desrosiers, coordinator of the Sexual Assault Intervention Network in New Bedford and each of the children; (2) expert testimony to the effect that the techniques employed in the interview, including use of anatomically correct dolls and allegedly leading and suggestive questions, had affected the children's reliability; and (3) assertions in his offer of proof (a) that the children had been told in advance about the videotaped interview and (b) that prior to this interview, George and Heather had been interviewed eight and eleven times, respectively, by professionals such as Department of Social Service employees, police, investigators, and a psychotherapist. The trial judge denied the motions in limine, noting that he would view the videotape, and that he intended to review the children's competency in the course of the trial.

As the trial unfolded, the defendant's daughter was the first of his children to testify. A voir dire was conducted outside the hearing of the jury, with a full opportunity for questioning by all counsel. Both prongs of the competency test (see infra) were fully explored. The defendant's son, the older child, then testified. Without objection, the trial judge briefly examined George in the presence of the jury, to determine his competency. The boy was asked his name, age, grade in school, and the names of his school, his teacher, and his school principal. He was not asked any questions regarding his ability to understand the difference between the truth and a lie, or the importance of speaking only the truth. However, defense counsel made no objection to the competency of either child witness, nor did counsel request that the trial judge view the videotape before making a determination of the children's competency. We therefore consider whether the trial judge committed error in denying the motions for a pretrial hearing, and if so, whether it amounted to a substantial risk of a miscarriage of justice.

General Laws c. 233, § 20, as appearing in St. 1983, c. 145, provides, "Any person of sufficient understanding, although a party, may testify in any proceeding, civil or criminal . . . ." In this Commonwealth, there is a two-pronged test for competency. The witness must (1) have "the general ability or capacity to 'observe, remember, and give expression to that which she has seen, heard, or experienced'; and (2) . . . ha[ve] 'understanding sufficient to comprehend the difference between truth and falsehood, the wickedness of the latter and the obligation and duty to tell the truth, and, in a general way, belief that failure to perform the obligation will result in punishment.'" Commonwealth v. Gamache, 35 Mass. App. Ct. 805, 806 (1994), quoting from Commonwealth v. Brusgulis, 398 Mass. 325, 329 (1986), quoting from Commonwealth v. Tatisos, 238 Mass. 322, 325 (1921). A trial judge has broad discretion to determine whether a competency hearing is required, and whether a witness is competent to testify. "Whether the test is met is 'peculiarly for the trial judge, and his determination will be rarely faulted on appellate review.'" Gamache, supra at 806, quoting from Commonwealth v. Whitehead, 379 Mass. 640, 656 (1980).

Under Massachusetts law, there is no reason to conclude that the children were not competent to testify. Heather's response to the voir dire questions amply demonstrated her competence under both prongs of our competency test. Neither is there any reason for thinking that George was unaware of the difference between the truth and a lie, or the importance of speaking the former and not the latter, even though the judge did not explore the second prong of the competency test with him at trial. In fact, the videotaped interview with each child began with questions and admonitions to that effect. Finding no error under Massachusetts law, we consider the analysis set out in State v. Michaels, 136 N.J. 299.

The Michaels decision focused on the reliability of children's testimony. In keeping with a host of psychological research, the Michaels court acknowledged that leading or coercive questioning can distort a child's memory. (In Massachusetts, this would implicate the first prong of the competency test.) To provide defendants a means for raising this issue of memory distortion, the Michaels court provided for a "pretrial taint hearing" when a defendant shows "some evidence" that a victim's statements were the product of suggestive or coercive interview techniques. Michaels, supra at 320. Such is not the case here. Even if we were to embrace the Michaels analysis, a matter that we do not reach today, it would not change the result, as the defendant's offer of proof did not reach the threshold required by Michaels to trigger a pretrial hearing.

In the case at bar, both children made their initial allegations of abuse spontaneously, rather than in response to questioning. Heather made her first allegation when she discovered she was about to go on a camping trip without her mother, and feared her father might know where she was. George made his first allegation right after the defendant, who no longer lived with the family, appeared at the children's home with no warning and allegedly hit George's bicycle with his truck. The videotape of each child's separate interview with Desrosiers reveals that her questions were not particularly leading, nor were they coercive. In contrast to the Michaels interviews, there was no vilification of the defendant, no incessant questioning, no references to statements made by the other complainant, and no use of threats, bribes, or cajoling. While the defendant asserted in his motions that some eleven or more interviews had been conducted with the children, evidence of only one interview was proffered. We note that neither the initial interview, nor subsequent interviews (with the exception of the one set of interviews with Desrosiers) were videotaped or recorded, and that, without such recordings, it is difficult for the defendant to substantiate an offer of proof in support of a request for a pretrial competency hearing. Nevertheless, on the record before us, there is no reason to think that the children's allegations or testimony were coerced.

2. Exclusion of expert opinion on the interview techniques employed in questioning the complainants. The trial judge ruled that the defendant would be permitted to show the videotaped interview with the complainants, and to put on expert testimony regarding proper and improper interview techniques to be employed with children alleged to be the victims of sexual abuse. Further, the trial judge ruled that defense counsel would be permitted to argue in his closing that the techniques employed on the video were of the improper type described by the expert witness, and thus the children's testimony was unreliable. However, the trial judge refused to allow the defense expert to comment specifically on the questions employed in the videotape itself, on the basis that such opinions would impermissibly comment on the children's credibility. Apparently deciding as a matter of strategy that it would be better not to show the jury the videotape, defense counsel never showed the tape, and was arguably unable to put squarely before the jury the nature of the interview techniques used with the complainants.

Again relying on Michaels, the defendant complains that his expert should have been permitted to testify about the questions and techniques used in the videotaped interview. He argues that the jury would be aided by such testimony, and that the testimony would not be commentary on the complainants' credibility, but rather, an opinion on the soundness of the interview techniques themselves.

We need not decide today whether this is an appropriate distinction since the facts before us do not even reach the threshold established in Michaels for triggering admission of such testimony. Moreover, the defendant obtained much of what he sought. His expert testified generally about proper and improper interview techniques, and told the jury that the best practice is to videotape all interviews with a complainant, because, without such documentation, "you simply don't know the behavior of adults toward the child, nor can you tell what kind of influence was exerted on the child to produce an account that conforms to what the adult [interviewer] wants rather than what the child may have to tell you, if anything." The expert further testified that children are more suggestible than adults and criticized the use of anatomically correct dolls in interviews because they suggest to the complainant that the interviewer wants to hear about genitalia or sexual acts. The defendant was able to introduce evidence that not all of the interviews with the complainants had been videotaped, and that anatomically correct dolls had been used in the Desrosiers interviews with the children. The defendant was further permitted, despite contrary authority (see Commonwealth v. Ianello, 401 Mass. 197, 201-202), to introduce testimony that children in families undergoing divorce (as the Allens were) are more likely to fabricate allegations of sexual abuse. We accordingly find no error in the exclusion of the subject expert testimony.

3. Admission of expert testimony. The defendant's next claim of error is that two expert witnesses were improperly permitted to vouch for the complainants. The first expert, Dr. Van Deven, is a medical doctor at Children's Hospital whose job is to examine children when there is a question of child abuse or child sexual abuse. Doctor Van Deven provided no fresh complaint testimony. In fact, she made clear that she conducts only physical examinations; she does not interview the children she sees. That task is left to a psychologist in the hospital. Doctor Van Deven testified that, ". . . I do not reinterview children as to what they say may have happened or why, because that's not my practice. Usually children I have seen have already been interviewed by other authorities." Doctor Van Deven described generally how she conducts a physical examination in cases of alleged sexual abuse, and described what she observed in examining Heather, including her finding that Heather's hymen was "normal shape," that it was not scarred, and that the hymen was a bit fuller or stretched out, which could be normal for Heather or could reflect "some kind of minor injury" to the hymen. Doctor Van Deven concluded her testimony by noting that there is a "normal exam" in the majority of girls examined with regard to a finding of sexual abuse. The defendant objected to this last statement at trial, and timely raised the issue on appeal. He argues that the statement was improperly admitted because it was tailored to vouch for Heather's credibility. We do not agree.

Doctor Van Deven testified within her competence as an expert regarding the findings from her examination of Heather, and regarding general syndromes (here, a lack of physical findings) associated with sexual abuse. An expert may testify about general syndromes associated with sexual abuse. See Commonwealth v. Dockham, 405 Mass. 618, 628 (1989); Commonwealth v. Mamay, 407 Mass. 412, 421-422 (1990). However, in so doing, the expert must not connect the complainant to the syndrome. Commonwealth v. Trowbridge, 419 Mass. 750, 759 (1995). Commonwealth v. Perkins, 39 Mass. App. Ct. 577, 582-583 (1995). Doctor Van Deven did not connect Heather with the general syndrome. In fact, the trial judge is to be credited for having ruled in advance of Dr. Van Deven's testimony that she would not be permitted to conclude her general syndrome testimony with her note in Heather's medical record, "So today's normal findings do not exclude the possibility of sexual abuse." Such testimony might well have crossed over the line into vouching. Doctor Van Deven was not asked to render an opinion as to whether Heather had been sexually abused, nor did she give such an opinion in her testimony. Contrast Commonwealth v. Colin C., 419 Mass. 54, 61 (1994); Commonwealth v. Trowbridge, supra at 760; Commonwealth v. Perkins, supra at 583-584; Commonwealth v. Brouillard, ante 448 (1996). We find no error in the admission of Dr. Van Deven's testimony.

The defendant next claims that the trial judge erred in allowing the complainants' treating therapist to testify as an expert, giving testimony about general syndromes associated with sexual abuse. We disagree. Doctor Solomone gave expert testimony that children's disclosures of sexual abuse "[o]ften . . . come out piecemeal over an extended period of time." Only towards the end of his direct testimony did the prosecution bring out the fact that Dr. Solomone was the children's treating therapist, as a prelude to a single question asked to rehabilitate Heather's testimony. The trial judge admitted Dr. Solomone's testimony not as fresh complaint, but to rehabilitate Heather. While the prosecutor touched only briefly on Dr. Solomone's relation to the complainants as treating therapist, the defendant, on cross-examination, focused primarily on this relationship, trying to show that the children's testimony might have been influenced by Dr. Solomone as a result of their frequent discussions with him about the defendant, and their alleged preparation with Dr. Solomone for the videotaped interview.

We have suggested that it is the better practice to avoid using the treating therapist as an expert on syndromes associated with sexual abuse, as it gives rise to the risk that such an expert's testimony "can be construed as impliedly supporting the truthfulness of [the complainant]." Commonwealth v. McCaffrey, 36 Mass. App. Ct. 583, 591 (1994). See also Commonwealth v. O'Brien, 35 Mass. App. Ct. 827, 832 n. 6 (1994); Commonwealth v. Rather, 37 Mass. App. Ct. 140, 148 n. 4 (1994); Commonwealth v. Brouillard, supra at 451. However, we have not gone so far as to hold that it would never be permissible for a treating therapist to give expert testimony. See, for example, Commonwealth v. Hudson, 417 Mass. 536, 541 (1994); Commonwealth v. O'Brien, supra at 834 (1994). Here, as in Hudson and O'Brien, Dr. Solomone's testimony about what the complainant told him was quite limited. Other witnesses (the complainants' mother and a child abuse investigator) supplied the fresh complaint testimony detailing the specifics of the alleged acts of abuse. As in Commonwealth v. O'Brien, supra at 833-834, it was the defendant's decision to solicit extensive testimony from Dr. Solomone regarding his therapy sessions with the children, their number, frequency, and subject matter. Similarly, the defendant here, as in O'Brien, argues that someone other than the children's therapist should have been brought in to testify as the prosecution's expert. We agree with the O'Brien court that, while it would have been preferable to have someone other than Dr. Solomone provide expert testimony, ". . . on the whole record we are satisfied the jury could find their way and the defendant was well defended and fairly tried." Commonwealth v. O'Brien, supra at 834. Moreover, on direct examination, the defendant's own expert gave testimony to the same effect as that complained of, stating that victims of child sexual abuse often disclose in bits and pieces. The complained of testimony was therefore cumulative. See Commonwealth v. Montmeny, 360 Mass. 526, 529 (1971).

The defendant's last complaint about Dr. Solomone's testimony is that the trial judge improperly left the question of Dr. Solomone's qualification as an expert to the jury, and that Dr. Solomone lacked the requisite background, training, and experience to testify as an expert. The qualification of a witness to offer an expert opinion is a preliminary issue of fact within the discretion of the trial judge. Commonwealth v. Garabedian, 399 Mass. 304, 310 (1987). Commonwealth v. Benoit, 410 Mass. 506, 520 (1991). A trial judge's decision on a witness's qualifications to testify as an expert will rarely be overturned on appeal. Commonwealth v. Seit, 373 Mass. 83, 92 (1977). Cronin v. McCarthy, 22 Mass. App. Ct. 448, 449 (1986). Doctor Solomone's background and experience were such that the trial judge was within the bounds of his discretion in allowing Dr. Solomone to testify regarding the common phenomenon of delayed disclosure in cases of child sexual abuse.

The defendant argues that the trial judge abdicated his responsibility to determine Dr. Solomone's qualification as an expert. We disagree. After the comment complained of, the judge immediately gave the jury an expert witness instruction, which he repeated in his instructions to the jury at the end of the trial. Such instructions indicate that the judge determined that Dr. Solomone was qualified as an expert. Leibovich v. Antonellis, 410 Mass. 568, 572-573 (1991). Furthermore, it is assumed that a trial judge has decided to qualify an expert when she accepts the expert's credentials and allows him to testify, as the trial judge did here. Commonwealth v. Boyd, 367 Mass. 169, 183 (1975). While the judge's comment (see note 15, supra) may have been ill-advised, it did not constitute an abdication of his responsibility to determine whether Dr. Solomone was qualified to testify as an expert.

4. Admission of fresh complaint testimony. The defendant argues that the children's complaints, some three to four months after the alleged abuse occurred, were not seasonably made, and thus fresh complaint testimony from their mother and a child abuse investigator should have been barred. We do not agree.

"There is no absolute rule of law as to the time within which a sexual assault victim must make [a] first complaint for that complaint to be admissible in evidence as a fresh complaint." Commonwealth v. Amirault, 404 Mass. 221, 228 (1989). Commonwealth v. Montanino, 409 Mass. 500, 508 (1991). Commonwealth v. Fleury, 417 Mass. 810, 813 (1994). The test is whether a victim's delay in reporting alleged abuse was reasonable under the circumstances. "Courts have not insisted on great promptness for fresh complaints in prosecutions involving child sexual abuse." Commonwealth v. Amirault, supra at 229. Here, the children, eight and nine years old at the time of the abuse, were threatened by the defendant that he would harm them and members of their family if they disclosed the abuse. The children disclosed the alleged abuse to their mother only when they feared they were again in danger of imminent harm from the defendant. The relationship of the complainant to the defendant, the complainant's age, the defendant's control over the complainant, and any threats against or intimidation of the complainant, are factors to be weighed in assessing the reasonableness of a delayed complaint. Commonwealth v. Dockham, 405 Mass. at 626. Commonwealth v. Fleury, supra at 814-815. Given the children's fear of reprisals from the defendant, and the reasonableness of their belief that their own father could gain access to them, the victims' complaints were seasonably made. See Commonwealth v. King, 387 Mass. 464, 473-474 (1982); Commonwealth v. Rockwood, 27 Mass. App. Ct. 1137, 1139-1140 (1989); Commonwealth v. Titus, 32 Mass. App. Ct. 216, 222-223 (1992).

The defendant next complains that it was error to admit the fresh complaint testimony of Tom Carroll, a child abuse investigator who witnessed the children's interviews with Desrosiers through a two-way mirror. The defendant argues that, because the children did not address their complaints to Carroll, he should not have been permitted to testify. We do not agree. A fresh complaint witness need not be the intended recipient of a victim's statement. The witness can be a third party such as Carroll who was present when the complaint was made. Commonwealth v. Owens, 402 Mass. 639, 640-641 (1988). Even a videotape itself is sometimes admissible as fresh complaint. See Commonwealth v. Lavalley, 410 Mass. 641, 645 (1991); Commonwealth v. Jerome, 36 Mass. App. Ct. 59, 62-63 (1994). There was no error in admitting Carroll's testimony.

5. Jury instructions. Citing to no authority, the defendant complains that the trial judge erred by refusing to deliver the defendant's requested instructions concerning the reliability of children's testimony. This argument does not rise to the level of appellate argument, and therefore need not be considered here. Mass.R.A.P. 16 (a)(4), 367 Mass. 921 (1975).

Judgments affirmed.