The defendant, Gordon J. MacRae, was convicted of one count of felonious sexual assault and four counts of aggravated felonious sexual assault. See RSA 632-A:3 (1986); RSA 632-A:2, X (1986) (amended 1992). On appeal, the defendant argues that the trial court erred in admitting expert testimony allegedly offered by the State to prove that the victim’s allegations of sexual assault were true. He also asserts that the Superior Court (Brennan, J.) erred in refusing to allow him to cross-examine the victim about past juvenile convictions and in instructing the jury to disregard portions of the victim’s testimony. We affirm.
The defendant was convicted of acts that took place during “pastoral counselling” sessions he conducted with the victim while the defendant was serving as a priest at St. Bernard’s Parish in Keene. While the acts occurred between June and November 1983, the offenses were not reported until March 1993, when the victim *108contacted the Keene police department after learning that his brother had reported similar incidents of sexual abuse by the defendant.
The victim’s trial testimony lasted approximately four days, during which he was extensively cross-examined. The defense repeatedly attacked the victim’s credibility, and questioned him about the delay in reporting the assaults and his inability to recall specific details regarding those assaults. The defense also emphasized the victim’s apparent failure to actively resist the assaults, his continued close relationship with the defendant after the assaults, and his testimony that during the assaults he felt dissociated from his body.
Following the victim’s testimony, the State called Dr. Leonard Fleischer, a Ph.D. in counseling psychology, to testify about characteristic behaviors of sexual abuse victims in order to rebut many of the negative inferences that could be drawn from the victim’s cross-examination. The defendant argues that the trial court, over objection, allowed the State to admit testimony from the expert to prove that the victim’s allegations were true, in violation of our ruling in State v. Cressey, 137 N.H. 402, 628 A.2d 696 (1993).
As a preliminary matter, the State argues that this issue was not properly preserved for appeal. See N.H. R. Ev. 103. Defense counsel objected to Dr. Fleischer’s testimony by stating that “we note our exception and in particular for the record we would be relying upon the standards in Cressey which we discussed earlier this morning . . . regarding reliability, predictability and the ability to effectively challenge and cross-examine.” The State contends that the defense challenged only the reliability of the underlying psychological theories, and therefore cannot now base its claim on allegedly improper comments on the victim’s credibility by the State’s expert..
Assuming, arguendo, that the issue was properly preserved for appeal, see, e.g., State v. Roy, 138 N.H. 97, 98, 635 A.2d 486, 486 (1993), we address its substance. In Cressey and its progeny, see, e.g., State v. Chamberlain, 137 N.H. 414, 628 A.2d 704 (1993); State v. Luce, 137 N.H. 419, 628 A.2d 707 (1993), we set forth the circumstances under which expert testimony in sexual assault cases is sufficiently reliable to be admissible under New Hampshire Rule of Evidence 702. We held that
the State may offer expert testimony explaining the behavioral characteristics commonly found in child abuse victims to preempt or rebut any inferences that a child victim witness is lying. This expert testimony may not be offered *109to prove that a particular child has been sexually abused
Cressey, 137 N.H. at 412, 628 A.2d at 703. We allow such expert testimony to educate the jury about apparent inconsistent behavior by a victim following an assault that may be misinterpreted by the jury as evidence that the victim’s testimony was fabricated. See Chamberlain, 137 N.H. at 417, 628 A.2d at 706.
Expert testimony loses its educational value, however, and is inadmissible when it is “directly linked to a determination of the guilt or innocence of the defendant,” State v. Silk, 138 N.H. 290, 291, 639 A.2d 243, 244 (1994) (quotation omitted), as opposed to providing useful information that is beyond the common experience of an average juror. Cf. State v. Black, 537 A.2d 1154, 1156-57 (Me. 1988) (finding expert’s testimony to be admissible in part and inadmissible in part); State v. Spigarolo, 556 A.2d 112, 123 (Conn.) (critical distinction between admissible and inadmissible testimony is whether testimony directly concerns that particular victim’s credibility), cert. denied, 493 U.S. 933 (1989). The defendant argues that the State exceeded the permissible bounds of Cressey by examining Dr. Fleischer only about the specific behaviors that the victim had exhibited, which, the defendant contends, resulted in Dr. Fleischer giving the functional equivalent of a direct opinion about whether the victim’s allegations were credible. We disagree. We have stated that
when a [victim’s] actions after an alleged incident of sexual abuse have the potential to lead a jury to conclude that the [victim] is lying, the testimony of a qualified expert may be beneficial to offer an alternative explanation for the [victim’s] specific behavior so that the jury may more accurately judge the credibility of the [witness],
Chamberlain, 137 N.H at 418, 628 A.2d at 706-07 (emphasis added).
Testimony of specific behavior is inadmissible, however, if its purpose is to prove that abuse occurred, or if the expert testifies that the particular victim’s behaviors were consistent with one who had been abused, Luce, 137 N.H. at 422, 628 A.2d at 709. Dr. Fleischer never interviewed or examined the victim in this case for evidence of sexual abuse, and never stated that he thought the victim’s behavior was consistent with one who had been abused. But see Silk, 138 N.H. at 291, 639 A.2d at 244; Chamberlain, 137 N.H. at 418, 628 A.2d at 706; Cressey, 137 N.H. at 406-07, 628 A.2d at 699-700. Indeed, Dr. Fleischer testified that because he had never met the victim and had only a general knowledge of the facts of the *110case, he was in no position to judge the victim’s credibility. His testimony was carefully limited to informing the jury about characteristics and behaviors exhibited by victims of child sexual abuse in order to rebut the inference that the victim’s testimony was not credible.
The defense argues further that Dr. Fleischer’s testimony was not generalized because the questions posed by the State were closely tailored to the victim’s earlier testimony. The questions posed by the State, while inquiring generally about alternative explanations for behaviors exhibited by the victim, were not so specific as to amount to a direct comment on the defendant’s guilt. But see State v. Pansegrau, 524 N.W.2d 207, 210-11 (Iowa Ct. App. 1994) (finding error because hypothetical questions posed by State simply reiterated events that victim testified preceded rape); State v. Foret, 628 So. 2d 1116, 1130 (La. 1993) (finding error in expert testimony that included victim’s own description and concluded that victim was telling the truth).
Therefore, to the extent that the purpose of Dr. Fleischer’s testimony was educational and was related to the testimony of this particular victim, but was not offered as proof that the victim in fact had been abused, it was proper. In doing so, we emphasize that
[t]he fact that the jury, if it believes the expert’s testimony, may draw inferences which would tend to bolster the victim’s credibility does not make the evidence inadmissible. So long as the expert does not render an opinion on the accuracy of the victim’s recitation of facts, his or her general testimony on the dynamics of sexual abuse does not prejudice the jury.
State v. Hicks, 535 A.2d 776, 778 (Vt. 1987) (quotation, brackets, and ellipsis omitted); see Foret, 628 So. 2d at 1130.
One aspect of Dr. Fleischer’s testimony, however, must be carefully considered. The victim, on direct examination, stated that he had been admitted to Beech Hill Hospital, a drug and alcohol rehabilitation facility, in 1985. Dr. Fleischer later testified that it was his experience that between seventy and eighty percent of the males who had been treated for substance abuse problems at Beech Hill between 1989 and 1990 had been sexually abused. The defendant alleges, and the State concedes, that this was inadmissible expert testimony because it improperly provided statistical evidence that the victim more probably than not had been abused. See State v. Wheat, 527 A.2d 269, 274-75 (Del. 1987). While the court erred by admitting this testimony, we find it harmless beyond a *111reasonable doubt. Similar statistical evidence regarding patients at the Derby Lodge facility, which the victim had also attended, was first presented to the jury by defense counsel during cross-examination of the victim in an effort to show that earlier disclosures by the victim were the result of peer pressure in group counseling sessions. But see Silk, 138 N.H. at 291-92, 639 A.2d at 245 (inadmissible expert testimony was not merely a single question and answer). Because the inadmissible testimony presented by Dr. Fleischer only reiterated similar statistics already presented to the jury by the defense, the error was harmless.
The defendant next argues that the court erred by precluding him from cross-examining the victim about his juvenile convictions. New Hampshire Rule of Evidence 609(d) states that “[e]vidence of juvenile adjudications is generally not admissible.” The court may admit such evidence only if “conviction of the offense would be admissible to attack the credibility of an adult” and the court is convinced that the evidence “is necessary for a fair determination of the issue of guilt or innocence.” N.H. R. Ev. 609(d).
The trial court has discretion in determining whether to allow cross-examination about prior criminal convictions, see State v. Chaisson, 123 N.H. 17, 30, 458 A.2d 95, 102 (1983), and we will not overturn the trial court’s ruling absent an abuse of discretion. See State v. McGann, 128 N.H. 186, 190, 514 A.2d 1247, 1250 (1986). To warrant a reversal, the defendant must show that “the trial court’s ruling was clearly untenable or unreasonable to the prejudice of his case.” State v. Gooden, 133 N.H. 674, 677, 582 A.2d 607, 609 (1990) (quotation omitted).
The defendant alleges that “the trial judge effectively declared that it was his policy not to allow impeachment by cross-examination with juvenile convictions.” See Chaisson, 123 N.H. at 30, 458 A.2d at 102 (trial court’s refusal to allow any witness to be impeached by a prior criminal conviction constituted reversible error). The defendant’s claim is without merit. Unlike Chaisson, there was no blanket prohibition against introduction of the victim’s prior juvenile convictions. The court allowed the defense to question a police officer about whether the victim’s entry into a drug and alcohol rehabilitation center had “something to do with a juvenile offense.” In fact, despite the State’s assertion during a pretrial hearing that there was a blanket prohibition against the admission of juvenile convictions, the court disagreed and stated that “there’s always room when justice is involved,” and agreed to revisit the issue, if necessary, later in the trial.
Nevertheless, the defendant faults the court for refusing to admit evidence of the victim’s juvenile convictions without making specific *112reference to Rule 609(d). When the defense attempted to question the victim about his juvenile convictions, however, the stated purpose of the inquiry was not impeachment, see N.H. R. Ev. 609(d), but rather to establish why the victim was in counseling with the defendant and to help ascertain whether the incidents of sexual abuse occurred during the dates charged in the indictments. In response, the court properly ruled that questions about specific juvenile convictions were not relevant to this stated line of inquiry, but did allow questioning about whether the victim was having legal problems at that time.
Finally, the defendant has failed to satisfy the requirements of either Rule 609(d) or our abuse of discretion standard. He has not shown that the proffered evidence was either “necessary for a fair determination of the issue of guilt or innocence,” N.H. R. Ev. 609(d), or that the trial court’s ruling unduly prejudiced his case, see Gooden, 133 N.H. at 677-78, 582 A.2d at 609. During the victim’s four days of testimony, the defense had ample opportunity to impeach his credibility. In addition to the extensive cross-examination noted above, the victim was questioned about his history of alcohol and drug abuse, a past admission that he had “lied a lot to get what [he] wanted” and past criminal behavior. The State had introduced evidence of the victim’s prior adult felony conviction. Evidence of the victim’s juvenile convictions would have.been of minimal impeachment value, and therefore we find no abuse of the trial court’s discretion. See id. at 678, 582 A.2d at 610.
The defendant’s final argument concerns the court’s instruction that the jury disregard the victim’s answers to a series of questions posed by the defense. A central theme of the defense was that the victim fabricated the accounts of abuse to further a civil suit against the Catholic Church. During cross-examination, defense counsel attempted to establish that the victim had contacted William Cleary, a private attorney, about the incidents of abuse prior to reporting them to Detective James McLaughlin of the Keene Police Department. The instruction at issue was made in response to a series of questions by defense counsel that included the following exchange:
Q: So, sir, from my understanding of your earlier testimony, you talked to William Cleary before you talked to Detective McLaughlin, not vice versa, isn’t that true?
A: No, you asked me — whatever your question was, who I talked to about the incidents. ... I said I talked to Jim McLaughlin in detail about what happened. I talked to Bill Cleary — maybe it was before I talked to Jim, but it was only out of a friendship and as my former lawyer to *113ask for his advice. You have to understand there were other circumstances going on at the time which aren’t able to —
Q: Yes. But would you agree with me you talked to Bill Cleary before you talked to Detective McLaughlin as McLaughlin indicates?
After the defense completed its cross-examination, the State requested a ruling from the court as to whether the defense, by inquiring about the civil suit against the church, had “opened the door” to the admission of evidence that the defendant had sexually abused other young men, including the victim’s brother. The court had earlier ruled that this evidence was inadmissible pursuant to New Hampshire Rule of Evidence 404(b).
During the course of this particular hearing, however, the court noted the evasive testimony of the victim reproduced above, and asked whether it was in part caused by the limitations imposed by the Rule 404(b) order. The court conducted a voir dire of the victim, during which he testified that he was aware of the court’s order regarding prior bad acts of the defendant, and as a result, was hesitant in answering defense counsel’s questions. The victim stated that he had contacted Attorney Cleary to discuss incidents of sexual abuse by the defendant reported by his brother, but felt unable to testify to that fact given the court’s order.
The trial .'judge ruled that the defense had gained a misleading advantage because the victim’s credibility “was damaged” due to the court’s “order on his inability to answer fully the questions.” Rather than admitting evidence of the defendant’s other bad acts, however, the court issued the following curative instruction to the jury:
Ladies and gentlemen, you may not consider those statements for the purpose of determining whether [the victim] told the truth when he testified to those statements. You may not consider those statements for the purposes of determining whether [the victim] is telling the truth during the rest of his testimony. At the time [the victim] made those statements, he could not fully answer [defense counsel’s] questions. There was a standing Court order in this case that prevented [the victim] from answering the questions completely.
The defendant argues that this series of questions did not mislead the jury and, therefore, the court erred in issuing the curative *114instruction. When a defendant leaves the trier of fact with a false or misleading impression, the State is entitled to counter with evidence to refute the impression created by the defendant and cure the misleading advantage. See State v. Mello, 137 N.H. 597, 601, 631 A.2d 146, 148 (1993). This rule prevents a defendant from successfully excluding inadmissible evidence favorable to the State and “then selectively introducing pieces of this evidence for his own advantage, without allowing the prosecution to place the evidence in its proper context.” State v. Benoit, 126 N.H. 6, 20, 490 A.2d 295, 305 (1985) (quotation omitted). Since the trial court is in the best position to judge the witness’s demeanor and the impact on the jury, we will not disturb its finding absent an abuse of discretion. See State v. Taylor, 139 N.H. 96, 100, 649 A.2d 375, 377 (1994).
Here, the court found that the testimony was confusing and illogical, and had an impact on the jury. After a review of the record, we find that the court did not abuse its discretion and could have reasonably concluded that the victim’s inability to answer fully may have left the jury with the false impression that the victim was being evasive for reasons other than the court’s order.
Finally, the defendant argues that the instruction itself impermissibly invaded the province of the jury to determine the victim’s credibility. The scope and wording of jury instructions are within the sound discretion of the trial judge and are evaluated as a reasonable juror would have interpreted them. See State v. Cegelis, 138 N.H. 249, 251-52, 638 A.2d 783, 784 (1994). The scope of the instruction was narrow, charging the jury to disregard only those statements that had created the misleading impression. We find that a reasonable juror would not have interpreted the instruction as commenting on the victim’s overall credibility, cf. King, 136 N.H. at 677, 621 A.2d at 923, but rather explaining why the victim’s answers to this particular line of questioning appeared hesitant and evasive. The court emphasized that only “these parts of the testimony that I point out here are not to be used by you.” The court’s instruction merely placed the victim’s responses in the proper context, and did not expound on the victim’s veracity.
Further, having found that the defense had created a misleading impression, the court’s only alternative to the curative instruction was to admit potentially prejudicial evidence of the defendant’s prior bad acts. Cf. State v. Trempe, 140 N.H. 95, 99, 663 A.2d 620, 622 (1995) (fact that door has been opened does not permit all evidence to pass through; the purpose of the doctrine is to prevent prejudice, *115not inject it). We find that the limiting instruction was not only proper, but prudent as well.
Affirmed.
All concurred.