A jury convicted Richard Aspen (the defendant) of (1) one count of rape of a child under sixteen, (2) six counts of rape, (3) two counts of indecent assault and battery, and (4) one count of assault and battery. The complainant is the defendant’s stepdaughter. The convictions were upheld on direct appeal. See Commonwealth v. Aspen, 53 Mass. App. Ct. 259 (2001).1 The defendant moved for a new trial, asserting various claims, including ineffective assistance of trial counsel and ineffective assistance of appellate counsel. The motion was denied. The defendant appeals.2 We now reverse the order denying the motion for a new trial.
Discussion. 1. Ineffective assistance of trial counsel. The defendant contends that trial counsel was ineffective in (1) failing to call an expert witness to rebut the Commonwealth’s expert, (2) failing to call available character witnesses, (3) advising the defendant not to testify,3 and (4) failing to do more to impeach both the complainant and her mother.4
We apply the familiar test of Commonwealth v. Saferian, 366 Mass. 89, 96 (1974). “We also consider whether there is ‘some *280showing that better work might have accomplished something material for the defense.’ ” Commonwealth v. Baran, 74 Mass. App. Ct. 256, 272 (2009), quoting from Commonwealth v. Satterfield, 373 Mass. 109, 115 (1977). “A strategic or tactical decision by counsel will not be considered ineffective assistance unless that decision was ‘manifestly unreasonable’ when made.” Commonwealth v. Acevedo, 446 Mass. 435, 442 (2006), quoting from Commonwealth v. Adams, 374 Mass. 722, 728 (1978).
a. Failure to call an expert witness. First, the defendant asserts that trial counsel provided ineffective assistance in failing to call his expert witness, Dr. Reneau Kennedy, to rebut the testimony of the Commonwealth’s expert, Dr. Renee Brant. At trial, Dr. Brant testified as to general characteristics common to children who have been victims of sexual abuse. Those characteristics matched those collectively known as “Child Sexual Abuse Accommodation Syndrome” (CSAAS). Trial counsel, instead of calling a competing expert to rebut Dr. Brant’s testimony, attempted to discredit Dr. Brant through cross-examination. This strategy was not manifestly unreasonable.
In her affidavit, Dr. Kennedy asserted that had she been allowed to testify, she would have explained that CSAAS was developed as a clinical observation tool rather than as a diagnostic tool and that it was “not intended to be used as proof that a child has been abused” but that it rather “explains . . . reactions to the assumed abuse.” She would have testified also that it was not developed through any accepted scientific methodology and not peer reviewed. The defendant contends that without his expert, trial counsel was “unable to counterbalance the critically important and powerful opinions given by the Commonwealth’s expert[] • • • ■” Commonwealth v. Baran, 74 Mass. App. Ct. at 276.
Through thorough cross-examination, however, defense counsel brought out the facts that CSAAS is not a diagnostic device, and that it assumes the abuse has occurred and explains a child’s reaction to the assumed abuse. Essentially, defense counsel was able to elicit on cross-examination most of the same facts to which Dr. Kennedy would have testified.
It has not been shown that Dr. Kennedy’s testimony would *281have added materially to the defendant’s case, see Acevedo, 446 Mass, at 442, quoting from Commonwealth v. Satterfield, 373 Mass, at 115 (defendant must show that “better work might have accomplished something material for the defense”), or that the tactical decision not to call her was unreasonable.
b. Failure to call available witnesses. The decision not to call either the complainant’s ex-boyfriend5 or any of the defendant’s twenty-two character witnesses was also not manifestly unreasonable.
While “evidence of a defendant’s general reputation is admissible, evidence in the form of private opinions is not.” Commonwealth v. Belton, 352 Mass. 263, 269 (1967). Moreover, once a defendant calls a witness to testify to the defendant’s good reputation for a particular character trait, Massachusetts law allows the prosecutor to cross-examine that witness as to his or her awareness of any “rumors or reports of prior acts of misconduct by the defendant. . . that are inconsistent or conflict with the character trait to which the witness has testified.” Commonwealth v. Montanino, 27 Mass. App. Ct. 130, 136 (1989). See Commonwealth v. White, 409 Mass. 266, 276-277 (1991) (“any benefits that the defendant would have realized from this testimony could not have risen to the level of a ‘substantial ground of defence’ because they would have been largely offset by the harm that other aspects of the testimony would have done to his case”).
2. Ineffective assistance of appellate counsel. The defendant also alleges that appellate counsel provided ineffective assistance in failing to raise on appeal a violation of the Federico standard. Commonwealth v. Federico, 425 Mass. 844, 847 (1997).
The standard for determining ineffective assistance of appellate counsel is the same standard set forth in Commonwealth v. Saferian, 366 Mass, at 96: “whether there has been serious incompetency, inefficiency, or inattention of counsel — behavior of counsel falling measurably below that which might be expected from an ordinary fallible lawyer — and, if that is *282found, then, typically, whether it has likely deprived the defendant of an otherwise available, substantial ground of defence.” See Commonwealth v. Cardenuto, 406 Mass. 450, 453-454 (1990). See also Commonwealth v. Phinney, 446 Mass. 155, 162-163, 167-168 (2006). When assessing whether appellate counsel’s behavior fell below the standard of an ordinary, fallible lawyer, we focus on whether appellate counsel “failed to raise a significant and obvious issue . . . which . . . may have resulted in a reversal of the conviction, or an order for a new trial.” Commonwealth v. Sowell, 34 Mass. App. Ct. 229, 232 (1993), quoting from Gray v. Greer, 800 F.2d 644, 646 (7th Cir. 1986).
Here, defense counsel at trial had objected to testimony of the Commonwealth’s expert as being in violation of Commonwealth v. Federico, supra. Expert testimony “is admissible whenever it will aid the jury in reaching a decision, even if the expert’s opinion touches on the ultimate issues that the jury must decide.” Commonwealth v. Federico, 425 Mass, at 847, quoting from Simon v. Solomon, 385 Mass. 91, 105 (1982). “[Testimony on the general behavioral characteristics of sexually abused children may properly be the subject of expert testimony because behavioral and emotional characteristics common to these victims are ‘beyond the jury’s common knowledge and may aid them in reaching a decision.’ ” Commonwealth v. Federico, supra at 847-848, quoting from Commonwealth v. Colin C., 419 Mass. 54, 60 (1994). “Such evidence must, however, be confined to a description of the general or typical characteristics shared by child victims of sexual abuse,” Commonwealth v. Federico, supra at 848, and “[djeference must be preserved for the role of the jury as the final judge of credibility . . . .” Ibid. “Evaluations of credibility are . . . within the exclusive province of the trier of fact,” Commonwealth v. Montanino, 409 Mass. 500, 504 (1991); “witnesses may not offer their opinions regarding the credibility of another witness.” Ibid., quoted in Commonwealth v. Federico, supra.
The Supreme Judicial Court in Commonwealth v. Federico, 425 Mass, at 849, “identified several broad areas of expert testimony that are most likely to run afoul of” the prohibition *283against witnesses offering their “opinions regarding the credibility of another witness”: First, an expert may not directly refer to or compare “the behavior of the specific child complainant to general characteristics of sexual abuse victims because ‘[s]uch testimony impermissibly intrudes on the jury’s province to assess the credibility of the witness.’ ” Id. at 848-849, quoting from Commonwealth v. Trowbridge, 419 Mass. 750, 759 (1995) . Second, the expert may not give an opinion as to whether the alleged victim was in fact subject to sexual abuse. Commonwealth v. Federico, supra. “Third, with respect to the accused, the expert may not provide profiles or testimony as to the typical attributes or characteristics of the perpetrators of sexual abuse.” Id. at 850, citing Commonwealth v. Day, 409 Mass. 719, 724 (1991) (testimony concerning “battered child syndrome” admissible, but testimony concerning “profile” of child batterers inadmissible).
Thus, testimony that “explicitly links [the expert witness’s] opinion to the experience of the witness child” is inadmissible. Commonwealth v. LaCaprucia, 41 Mass. App. Ct. 496, 498 (1996) , quoting from Commonwealth v. Richardson, 423 Mass. 180, 186 (1996). Further, “[i]n some cases opinion testimony that does not explicitly link the opinion to the child witness nevertheless constitutes impermissible vouching.” Commonwealth v. Richardson, supra. Commonwealth v. LaCaprucia, supra. See Commonwealth v. Perkins, 39 Mass. App. Ct. 577, 582-584 (1995) (experts’ testimony, in response to hypothetical questions “whose components mirrored the underlying facts of the case, would inescapably have the same impact on the jury as a direct reference to and comparison with the child witnesses in the case”; such testimony was “tantamount to an endorsement of the credibility of the complaining child witness”).
In this case, the defendant does not allege that Dr. Brant did not testify to child victims’ general characteristics.6 Instead the defendant asserts that the Commonwealth and Dr. Brant “cherry-pick[ed]” characteristics that were similar to those that the complainant experienced, leaving out other characteristics: that is, that they “tailored] ” the characteristics mentioned to fit the *284evidence in this case. This type of testimony can be problematic, because “opinion testimony that does not explicitly link the opinion to the child witness [may] nevertheless constitute[] impermissible vouching.” Commonwealth v. Richardson, 423 Mass, at 186, and cases cited. Commonwealth v. LaCaprucia, 41 Mass. App. Ct. at 498.
Dr. Brant’s testimony also went beyond the description of general characteristics of child abuse victims and expanded into profile testimony relating to “intrafamily sexual abuse” that mirrored the complainant’s family makeup and dynamic.7 See Commonwealth v. Poitras, 55 Mass. App. Ct. 691, 693-694 (2002) (expert’s testimony virtually mirrored evidence proffered by prosecution; “[t]he admission of such testimony that effectively provides a profile of typical abusers or abuse situations has been condemned by the appellate courts of this Commonwealth as error, especially where the particular characteristics the expert testified to could be taken by the jury to identify the defendant as fitting the profile and therefore more likely than not to have committed the crime”). Dr. Brant’s testimony then moved to the general characteristics of disclosure, using factors that resembled the complainant’s testimony.8 This testimony essentially signaled to the jury that the complainant’s testimony was credible. See Commonwealth v. LaCaprucia, 41 Mass. App. Ct. at 499 (expert essentially gave profile testimony that presented defendant’s family situation as prone to abuse, suggesting to jury that this profile evidence was reliable factor in determining whether abuse had occurred). We have previously held that testimony such as this, “while not directly vouching *285for the complainant[’s] credibility nor expressly connecting the childfs] claim of sexual abuse to the defendant, served as the functional equivalent.” Commonwealth v. LaCaprucia, supra at 501, citing Commonwealth v. Trowbridge, 419 Mass, at 759-760.
Trial counsel, who appropriately objected to Dr. Brant’s testimony, suggested to appellate counsel that he raise a violation of the Federico standard on appeal. Appellate counsel, in his affidavit, averred that he did not pursue this issue because he “did not believe it would be successful.” This decision by appellate counsel was manifestly unreasonable. See Commonwealth v. LaCaprucia, 41 Mass. App. Ct. at 497-502; Commonwealth v. Poitras, 55 Mass. App. Ct. at 694-695. See also Commonwealth v. Day, 409 Mass, at 723 (such testimony “inherently prejudicial to the defendant”).
Appellate counsel’s behavior deprived the defendant of an available, substantial ground of defense by “fail[ing] to raise a significant and obvious issue . . . which . . . may have resulted in a reversal of the conviction, or an order for a new trial.” Commonwealth v. Sowell, 34 Mass. App. Ct. at 232, quoting from Gray v. Greer, 800 F.2d at 646. If the issue had been argued on appeal and Dr. Brant’s testimony had been held inadmissible, the question would then have been whether there was reversible error.9 Given the lack of physical evidence, the issue of guilt here ultimately turned on credibility.10 See and compare Commonwealth v. Poitras, 55 Mass. App. Ct. at 694-695 (“The harm is particularly pronounced when ... the issue of guilt ultimately turn[s] on . . . credibility . . . and the prosecutor, in closing argument, remindfs] the jury of the very testimony by the expert that provided profile characteristics matching those of the defendant”).
*286Also, “[p]art of th[e] determination turns on what the prosecutor argued to the jury.” Commonwealth v. LaCaprucia, 41 Mass. App. Ct. at 501. Here, during closing argument, the prosecutor suggested the complainant’s family dynamic fit a characteristic that Dr. Brant had opined was indicative of child sex abuse. In addition, the prosecutor specifically referenced Dr. Brant’s testimony when she urged the jury to consider the relationship between the defendant and the complainant.11 Furthermore, the judge merely gave the standard instructions regarding expert testimony and credibility. Without specific instructions, it is more likely than not that the jury considered Dr. Brant’s testimony “as supportive of the accuracy, reliability, and truthfulness” of the complainant’s testimony. Id. at 502. See Commonwealth v. Rather, 37 Mass. App. Ct. 140, 150-151 (1994). As a sufficient basis appears on this record to constitute reversible error, we conclude that appellate counsel’s failure to argue this issue satisfies both prongs of the Saferian test, and that “better work might have accomplished something material for the defense.” Commonwealth v. Satterfield, 373 Mass, at 115.
We acknowledge that the strategy of “ ‘[wjinnowing out weaker arguments on appeal and focusing on’ those more likely to prevail ... is the hallmark of effective appellate advocacy.” Commonwealth v. Sowell, 34 Mass. App. Ct. at 233, quoting from Smith v. Murray, 477 U.S. 527, 536 (1986). Here, however, appellate counsel, instead of arguing that it was error to admit Dr. Brant’s testimony, chose to challenge the jury selection process and the admission of fresh complaint evidence.12 See Commonwealth v. Aspen, 53 Mass. App. Ct. at 261-270. In short, appellate counsel did not raise the most winnable argument on appeal.
*2873. Other issues. We do not reach the remaining arguments raised by the defendant, as they are either without merit or unlikely to result in the granting of a new trial. Deciding as we do, we do not need to discuss the defendant’s right to a public trial.13
Conclusion. The order denying the motion for a new trial is reversed, and a new order shall enter allowing the motion for a new trial. The judgments are reversed, and the verdicts are set aside.
So ordered.
The facts on which the defendant’s convictions were based are recited in Commonwealth v. Aspen, 53 Mass. App. Ct. at 260-261.
The Commonwealth argues that because the defendant did not raise the claims of ineffective assistance of trial counsel in his direct appeal, they are waived. However, as we have often stated, “the preferred method for raising a claim of ineffective assistance of counsel is through a motion for a new trial.” Commonwealth v. Zinser, 446 Mass. 807, 810 (2006). See Commonwealth v. Saferian, 366 Mass. 89, 90 n.l (1974).
In an affidavit in support of his motion, the defendant stated that he “reluctantly” took counsel’s advice not to testify. The defendant must “prove, by a preponderance of the evidence, that, but for his counsel’s erroneous advice ... he would have testified in his own defense.” Commonwealth v. Marrero, 459 Mass. 235, 241-242 (2011), quoting from Commonwealth v. Freeman, 29 Mass. App. Ct. 635, 642 (1990). “It is not enough to say that counsel had discouraged him from testifying.” Commonwealth v. Marrero, supra at 242. The defendant has not met this burden.
“Generally, failure to impeach a witness does not amount to ineffective assistance of counsel.” Commonwealth v. Fisher, 433 Mass. 340, 357 (2001). The decision to impeach a witness is “fraught with a host of strategic considerations, to which we will . . . show deference.” Ibid. Trial counsel thoroughly impeached both witnesses. It is speculative to presume that more could have been done to further undermine their credibility. See ibid, (rejecting claim of ineffective assistance where line of questioning that counsel failed to pursue was not “obviously powerful form of impeachment”).
The defendant did not submit an affidavit of the ex-boyfriend in support of the new trial motion, but only an affidavit of a defense investigator stating that in 2009 the ex-boyfriend did not want to be involved as a witness in connection with the new trial motion.
We note that the complainant here, who testified that the alleged abuse began when she was a teenager, was testifying at trial as an adult.
Dr. Brant testified that sexual abuse generally occurs in families involving a “parent, parent substitute, a boyfriend, [or] a caretaker with the family,” with an “attachment” to the victim. She went on to testify that the victim will often keep the abuse secret because children may “fear that something terrible will happen to the family”; “[m]aybe [the abuser] will have to leave the family,” and perhaps the abuser is an “important bread winner in the family,” such as a husband or father.
Dr. Brant testified that disclosures of sexual abuse are often made by “older children." She testified that with regard to younger children “[sjomeone will notice something unusual” about the child’s behavior, and that victims will often initially disclose the abuse to a “peer.” She also testified that a victim may disclose abuse out of “anger” (adding that “anger sometimes is a very legitimate precipitant of an abuse disclosure”) or sometimes disclosure is made out of concern for the “welfare or the safety of others in the family.”
Our courts have held that profile testimony “is inherently prejudicial to the defendant. . . and cannot be deemed harmless error when the Commonwealth’s case against the defendant is not strong, because of the reasonable possibility that it contributed to the jury’s conclusion that the defendant was guilty.” Commonwealth v. Poitras, 55 Mass. App. Ct. at 694-695.
While there was corroboration of the complainant’s testimony, by her brother, that the stepfather used physical punishment on the children, and of her testimony that there were guns in the house, the question whether the sexual acts occurred turned on credibility.
The prosecutor argued, “And if you take a look at that card [a card from the complainant to the stepfather], there is language in the card where [the complainant] writes, ‘I love you. I love you more than anyone else in the world.’ And I want you to think about Dr. Brant’s testimony in this — in relationship to what Dr. Brant told you, and this whole business about the card.”
The defendant did not object to the composition of the deliberating jury, and this court held that the fresh complaint testimony admitted, even if in error, was not unduly prejudicial. See Commonwealth v. Aspen, 53 Mass. App. Ct. at 266-267.
Although the defendant established that the jury selection process was impermissibly closed, he does not allege that it was unfair, nor has he identified any unfairness that would have been prevented had the voir dire been performed in public. See Commonwealth v. Horton, 434 Mass. 823, 832 (2001). The issue also appears to have been waived. See Commonwealth v. Morganti, 467 Mass. 96 (2014); Commonwealth v. Alebord, 467 Mass. 106, 111-114 (2014).