A jury convicted Richard Cosio of taking indecent liberties with a minor child, carnal knowledge, three counts of first degree child sexual abuse, and three counts of second degree child sexual abuse. D.C.Code §§ 22-3501(a), -2801 (1981); id., §§ 22-4108, -4109 (1996 repl.).1 In his direct appeal from these convictions, Cosio contends that the trial court abused its discretion in excluding a medical record of a physical examination of the complainant. After his conviction, Cosio filed a pro se motion under D.C.Code § 23-110 (2001) alleging ineffective assistance of counsel, which the trial court denied without prejudice. When later represented by counsel, Cosio filed a second § 23-110 motion on the same issue, which the trial court denied after a hearing on the merits. We consolidated Cosio’s appeal from the denial of his second § 23-110 motion with his direct appeal and, after reviewing his arguments, affirm all convictions.
I.
Cosio’s convictions were attributable to his relationship with his half-sister, the complainant A.A. In 1990, when Cosio was eighteen and A.A. about seven years old, they lived with their mother and three other siblings in an efficiency apartment. A.A. testified that Cosio would hit her and her siblings, and she specified at least one incident during 1990 when Cosio sexually abused her. A.A. testified that she had been afraid of Cosio during this time period, which lasted approximately one year, and that she also had been afraid that her mother would not believe her if she told her mother what had happened. A.A. left to attend a boarding school for a few years, but returned to live with her family in 1993 or 1994. She testified that Cosio began to sexually abuse her again in 1994, *169when she was eleven. A.A. said that this abuse, which included Cosio’s putting his penis in her vagina, occurred more than once a year between 1994 and 1997 (by which time A.A. was fourteen years old). The last incident was alleged to have occurred in October 1997 around the time of Cosio’s birthday.
After A.A. came home from boarding school she had remained silent about the renewed abuse, she said, because she had been “too afraid” of Cosio, meaning she believed that he might “go insane.” During 1997, however, when A.A. was in eighth grade, she met two lawyers — a husband and wife — who tutored her for school and eventually gained her trust. A.A. testified that she had revealed the abuse to the lawyers in November of that year because she hoped that “they could try to find [her] help.”
After disclosing the abuse, A.A. went to a hospital and was interviewed by the police. On December 15, 1997, Dr. Beverly Lindsay, a pediatrician in the Division of Child Protection, examined A.A. at Children’s Hospital. Dr. Lindsay was qualified at trial as an expert in pediatrics, child sexual abuse, and child sexual abuse examinations. Dr. Lindsay testified that her examination revealed an injury to A.A.’s hymen that was consistent with sexual abuse.
As his first witnesses, Cosio called two of his coworkers to testify solely as character witnesses. Jose Garcia and Nora Car-nathan both testified that Cosio had a “good reputation” and was a “law abiding person.” Cosio next called Dr. John Adams, a forensic pathologist, who testified about the details of Dr. Lindsay’s examinations. Cosio sought to introduce through Dr. Adams the medical record of a 1993 examination of A.A. (believed to be a routine physical required by A.A.’s school). The record included a line for “genital/urinary” on which the examining doctor had checked a box labeled “normal.” After a voir dire examination covering Dr. Adams’s qualifications and what his testimony about the 1993 record would be, the trial court ruled that Dr. Adams was not qualified to testify about the record, and that in any event the record was inadmissible.
II.
In his § 23-110 motion and again on appeal, Cosio argues that A.A.’s alleged fear of him — proffered by the government as a major reason why A.A. had failed to report the alleged sexual abuse until late 1997 — was an issue of central importance to the government’s case. He maintains that his trial counsel’s failure to rebut the evidence purporting to establish this fear was a deficiency in performance so prejudicial to his defense that the assistance he received from counsel was constitutionally ineffective.
A.
In order to prevail on a claim of ineffective assistance of counsel, Cosio must show that (1) his trial counsel’s performance was “deficient,” meaning that the defaults were “so serious that counsel was not functioning as the ‘counsel’ guaranteed by the Sixth Amendment”; and that (2) counsel’s deficient performance so prejudiced the defense “as to deprive the defendant of a fair trial, a trial whose result is reliable.” Strickland v. Washington, 466 U.S. 668, 687, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984); Hill v. United States, 489 A.2d 1078, 1079 (D.C.1985). This court’s “scrutiny of counsel’s performance must be highly deferential.” Strickland, 466 U.S. at 689, 104 S.Ct. 2052. “[Strategic choices made after thorough investigation of law and facts relevant to plausible options are virtually unchallengeable.... ” Id. at 690, *170104 S.Ct. 2052. Furthermore, even if counsel’s representation was deficient, Co-sio must show enough prejudice that a reviewing court would have to conclude that, but for trial counsel’s deficiency, there is a “reasonable probability” that the result would have been different. Id. at 694, 104 S.Ct. 2052; Hockman v. United States, 517 A.2d 44, 51 (D.C.1986).
There can be no question on this record that A.A.’s alleged fear of Cosio played a central role in the government’s case, as the following pretrial and trial events make clear. The acts giving rise to Cosio’s convictions occurred while A.A. was between eleven and fourteen years old, but the government filed a notice of intent to introduce evidence of earlier, uncharged misconduct, as permitted by Drew v. United States, 118 U.S.App. D.C. 11, 331 F.2d 85 (1964) and Super. Ct.Crim. R. 12(c) (“Drew notice”). The Drew notice revealed that the government expected to show by clear and convincing evidence that Cosio had sexually and physically abused A.A. in 1990 and 1991 when she was seven and eight years old. The notice explained that this evidence was admissible for two reasons: first, to show Cosio’s “unusual sexual preference,” see Dyson v. United States, 97 A.2d 135 (D.C.1953); and second, to place the charged conduct in an understandable context, see Johnson v. United States, 683 A.2d 1087 (D.C.1996) (en banc). The prior physical abuse, argued the government, had “made [A.A.] fear [Cosio] and discouraged her from disclosing his acts of sexual abuse sooner.” According to the Drew notice, therefore, A.A.’s trial testimony about the earlier instances of abuse was necessary to rebut any importance the jury might place on the absence of contemporaneous reports of the abuse by A.A. Cosio’s trial counsel did not object to this'evidence until the morning of trial, more than four months after the Drew notice was filed. As a result, the trial court ruled that the objection was untimely and thus had been waived.
During his opening statement, the prosecutor referred to A.A.’s fear of Cosio three times. Then, before addressing the specific facts of A.A.’s case, Dr. Lindsay, the prosecution’s expert, testified that it was typical for a child not to disclose sexual abuse immediately both because she feels threatened and because she may not have a trusted adult in whom she feels comfortable confiding. Next, A.A. testified that she had not reported the sexual abuse, both when she was seven and eight and later from ages eleven to fourteen, because she had been “afraid” of Cosio. Finally, the prosecutor also mentioned A.A.’s fear of Cosio in the government’s closing argument and again in rebuttal closing argument.
Relying on A.A.’s and Dr. Lindsay’s testimony, the prosecutor also offered the jury several other theories as to why A.A. had delayed in reporting the abuse: (1) A.A. had not told schoolmates or teachers because she wanted to try to forget about what had happened and have a normal life; (2) A.A. may have thought that no one would believe her; and (3) A.A. may have been embarrassed or ashamed. The prosecutor’s central reason for the reporting delay, however, was A.A.’s fear of the consequences from Cosio.
B.
Cosio acknowledges that his trial counsel conducted an extensive pretrial investigation, including interviews with Cosio’s mother, brothers, sister Elizabeth, friends, school teachers, and coworkers. But he faults the questioning of Cosio’s coworkers that failed, he says, to reveal a powerful refutation of the government’s evidence that A.A. feared Cosio. Attached to the § 23-110 motion were affidavits from four *171coworkers who averred that A.A. had frequently visited Cosio where he worked, at International Data Processing (IDP), and that the two had appeared to have a “comfortable,” “friendly,” even “affectionate” relationship. Two of the affiants — Jose Garcia, who testified for Cosio at trial as a character witness, and Henry Estrada, a fellow employee who had not testified— also averred that they had spoken with trial counsel or his investigator but had been asked only about Cosio’s moral character, not about the nature of his relationship with A.A. At the § 23-110 hearing, four coworkers (three of whom had given affidavits as well as one who had not) said they could have testified at trial that A.A. would visit Cosio at work, and that the two of them appeared to have a “normal,” “comfortable,” and “friendly” relationship. Coworker Santos Rafael Villatoro testified that A.A.’s sister Elizabeth also worked at IDP, and that their brother, Dennis, as well as their mother would also spend time at IDP visiting Cosio.
Evidence produced on trial counsel’s behalf at the § 28-110 hearing established that four days after counsel had first met with Cosio, counsel had sent a memo to his investigator directing him, among other things, to investigate the nature of the relationship between Cosio and A.A., between A.A. and her family, and between A.A. and her tutor. The evidence also showed that Cosio had provided trial counsel with the names of many of his coworkers, a number of whom faxed statements to counsel in support of Cosio’s general good character. (After the guilty verdicts, but before sentencing, two coworkers also wrote letters to the trial judge that included remarks describing how Cosio’s relationship with his mother was dysfunctional.) None of these statements from coworkers mentioned Cosio’s and A.A.’s relationship with each other, except to point out generally that Cosio was concerned about the welfare of his family.
Cosio’s trial counsel testified at the hearing that although he had recognized that A.A.’s alleged fear of Cosio was an important issue, he had thought that the paramount concern was finding a reason why A.A. made the allegations that she did, in order to support a fabrication defense. Counsel also noted that the coworkers had informed him that Cosio’s mother was obsessive, that she would wait for him outside his place of employment for hours, and that A.A. seemed immature to them. From this testimony, therefore, we can be sure that trial counsel at least was aware that the coworkers had had a glimpse of the family’s dynamics — indeed, they had observed A.A. and may even have had some sort of interaction with her. On the other hand, the record does not indicate that trial counsel was informed by Cosio or by anyone else that any of Cosio’s co-workers had ever had ample opportunity to witness interactions between Cosio and A.A. Nor, according to trial counsel, had he received before trial any information from coworkers as to whether A.A. feared Cosio. (On the other hand, he acknowledged that he had not instructed his investigator to inquire of Cosio’s fellow employees about his relationship with A.A.) Trial counsel testified, in summary:
[I]n a case like this ... it’s almost ... assumed ... that I would investigate the relationship between [A.A.] and Mr. Cosio ... it is just implicit in this type of case that we’re going to do that. And there [were] documents that I saw where [the investigator] asked questions ... with regard to that issue to other witnesses, including family members [and A.A.’s sister’s friends] ... so I believe that — that issue was investigated, but that it wasn’t my focus....
*172C.
Cosio argues that coworker testimony about his “normal,” “comfortable,” “friendly,” even “loving” relationship with A.A., if presented at trial, would have impeached A.A.’s credibility by substantially rebutting the prosecution’s contention that A.A. had failed to report the abuse earlier because she was terrified of Cosio. We shall assume, for the sake of argument, that this testimony potentially would have been useful rebuttal evidence, as Cosio claims. But was counsel constitutionally deficient in failing to find out this information from Cosio’s former employees before trial?
It is far from self-evident that, in order to meet the standard of performance required by the Constitution, counsel should have asked Cosio’s coworkers whether they had ever seen A.A. and Cosio together at his workplace or elsewhere and, if so, what their relationship appeared to be. Coworkers are a logical source of information for evidence of good character; they are less easily thought of as sources of information about one’s family relationships. Nonetheless, trial counsel here had been notified about the government’s intention to show A.A.’s fear of her half-brother, and he also had learned that Co-sio’s coworkers had observed and perhaps even interacted with A.A. Under these special circumstances, then, would the most competent of defense counsel have considered Cosio’s coworkers to be potential fact witnesses and raised family relationship questions with them? Probably yes. Although coworkers might not be a typical source of information about a colleague’s family relationships, it at least is not unusual for some employees to share closely-held thoughts about personal relationships with friends they see every day at work. Furthermore, because several of Cosio’s coworkers had volunteered at least some observations about A.A., it would have been logical for Cosio’s lawyer to have followed up that information by asking how much his fellow employees knew about his relationship with A.A.
But how much would that pursuit have been likely to inform counsel and help the defense? Cosio almost certainly would not have shared information with co-workers tending to show that he was abusing his half-sister. On the other hand, any information he shared tending to show how much he cared for her would have been expected of a brother and thus of little, if any, real probative value on the “fear” issue. So the question narrows to whether trial counsel, to avoid constitutional deficiency, should have been expected to inquire of Cosio’s coworkers, as possible fact witnesses, whether they had any objective information, based on personál observation, that would characterize the kind of relationship Cosio had with A.A.
“Counsel has a duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary.” Strickland, 466 U.S. at 691, 104 S.Ct. 2052. The record of the § 23-110 hearing confirms that trial counsel did investigate the relationship between Cosio and A.A. by questioning the parties who had the greatest opportunity to witness it, namely, Cosio’s and A.A.’s family and friends. The record also shows that the information gathered from these most obvious sources did not refute the government’s evidence that A.A. feared Cosio. The coworkers’ observations, therefore, if trial counsel had learned about them, would have been the best evidence Cosio could have found for rebutting A.A.’s alleged fear of him. Trial counsel is not helped by the fact that Cosio apparently failed to tell him that fellow employees would be able to confirm the good relationship he claimed to have with A.A. A client *173should not be expected to anticipate the best defense and then volunteer every kind of information he or she has in support of it. It is the lawyer’s job to ask the right questions of the client, and there is no evidence that trial counsel asked Cosio whether his coworkers could help establish important facts.
All this said, it is not at all clear that counsel’s failure to pursue Cosio’s coworkers as fact witnesses amounted to constitutionally deficient preparation. In the first place, trial counsel’s basic strategy was a fabrication defense: motivated by resentment, A.A. made everything up to get out from under Cosio’s persistent efforts to control her life — in particular, his efforts to prevent her from spending time with her lawyer-tutors, and especially with the male tutor, whom Cosio suspected of having a romantic connection with A.A. That defense would to some extent have been inconsistent with, or at least in tension with, a simultaneous defense effort to demonstrate (through the testimony of Cosio’s fellow employees) that A.A. and Cosio had a close, friendly personal relationship — the kind of relationship necessary to rebut the government’s contention that A.A.’s driving emotion was fear of Cosio.
Second, although the government did stress that A.A. had not reported her sexual abuse because she feared Cosio, other reasons for her failure to come forward were offered as well. A.A. testified that she had not told schoolmates or teachers because she wanted to try to forget about what happened and have a normal life. And the government’s expert, Dr. Lindsay, testified that A.A. might have concealed the abuse because she was embarrassed or ashamed and, in any event, thought that no one would believe her. Thus, even the suggested coworker testimony would not have entirely rebutted the government’s “fear” argument; the deficiency of investigation was less telling than a failure to investigate would have been in another context where the government had no alternative basis for its contention.
Finally, coworker testimony that A.A. and Cosio appeared to have a normal, friendly, and comfortable relationship would not necessarily have been inconsistent with the government’s emphasis on A.A.’s fear of Cosio. The government could have argued, with some force, that a young girl truly afraid of her half-brother would have acted friendly toward him around his coworkers lest she suffer cruel reprisals from him later for failing to convey the impression of a close relationship.2 Had counsel put on the coworker evidence, therefore, the government would still have had a useful answer.
Despite trial counsel’s having some information that Cosio’s coworkers had seen Cosio and A.A. together on occasion, we shall assume for the sake of argument — in line with the position of Cosio’s counsel on appeal — that trial counsel did not make a conscious, tactical choice not to interview Cosio’s coworkers for evidence that might refute the government’s “fear” argument. Thus, we shall assume that counsel did not, in fact, exercise the kind of conscious, professional judgment that a reviewing court, applying hindsight, should not second-guess. Mayo v. Henderson, 13 F.3d 528, 533 (2d Cir.), cert. denied, 513 U.S. 820, 115 S.Ct. 81, 130 L.Ed.2d 35 (1994). Nonetheless, given the inherent tension between (1) an effort to show a close relationship between A.A. and Cosio to rebut her alleged fear of him; arid (2) an effort *174to prove A.A.’s fabrication of the abuse charges based on substantial resentment of Cosío, particularly because of his efforts to keep her away from her school tutors, we can say that trial counsel made a “reasonable decision” to offer a fabrication defense that made “unnecessary” the “particular investigation[ ]” counsel on appeal now cites as an omission of constitutional dimension. Strickland, 466 U.S. at 691, 104 S.Ct. 2052. Or to explain the point in another way, suppose that trial counsel had, in fact, conducted a more thorough investigation of Cosio’s coworkers, obtained all the information presented at the § 28-110 hearing, and then consciously decided not to use it out of a concern that the “close relationship” testimony might undermine the fabrication defense based on sustained resentment. Strickland, we believe, would have required us to defer to that judgment. If that is so, then the fact that trial counsel presented a fabrication defense without actually coming to grips with the available, additional evidence on the fear issue offers no principled basis for revising that conclusion. In sum, we cannot say that the performance of Cosio’s trial counsel was so deficient that he “was not functioning as the ‘counsel’ guaranteed by the Sixth Amendment.” Strickland, 466 U.S. at 687, 104 S.Ct. 2052.
From the foregoing analysis, it is apparent that trial counsel’s alleged deficiency is premised on the claimed importance of the missing coworker evidence for the defense case. We are not ultimately persuaded of that importance. While not of merely marginal value, this missing evidence, all things considered, would not assuredly have been of major value "to the defense. And as already noted, for tactical reasons counsel might well have withheld that evidence as counterproductive for a fabrication defense. In this case, therefore, the “deficiency” analysis blends with Strickland’s “prejudice” analysis.
Of course, once we have concluded that there was no constitutional deficiency in performance, there is no basis for considering whether counsel’s performance prejudiced the defense. Id. at 697, 104 S.Ct. 2052 (the court need not “address both components of the inquiry if the defendant makes an insufficient showing on one”). Nonetheless, we do not hesitate to add that Cosío has not met Strickland’s standard for reversible prejudice. While there may have been a reasonable “possibility” that the result here would have been different if counsel had skillfully employed coworker testimony to rebut the government’s “fear” • theory while presenting, somewhat inconsistently, a fabrication defense, our review of the record convinces us that there was no reasonable “probability” of a different result, as Strickland requires for reversal. Id. at 694, 104 S.Ct. 2052 (emphasis added).
Trial counsel had a coherent defense: A.A. lied — a fabrication that counsel sought to prove not only through cross-examination of A.A. but also through admission of a 1993 medical record, buttressed by testimony from an expert witness, Dr. John Adams, tending to prove that A.A. had not been raped in 1990 (as she had claimed), and thus that she was a demonstrable liar. We cannot say that trial counsel’s failure to add the coworker testimony, which might have undermined the “resentment” defense, created a trial deficiency prejudicial to Cosio’s defense in violation of the Sixth Amendment.
D.
Cosio’s counsel on appeal asks us not to rest our decision on a perceived inconsistency between the coworker evidence that would tend to show a close, friendly relationship between Cosío and A.A. and a fabrication defense predicated on A.A.’s *175resentment of Cosio. Appellate counsel puts a high premium on Cosio’s lost opportunity to impeach A.A.’s testimony that she feared Cosio with highly credible testimony that the two were close. That rebuttal testimony, says appellate counsel, would have had a devastating impact on the government’s case once the jury heard it. And, he argues, trial counsel could then have finessed the government’s follow-up argument that Cosio was taking inconsistent positions by simply reminding the jury that anyone, on occasion, can resent a person she loves.
This argument would have some force if A.A. — after testifying that she feared Co-sio — denied on cross-examination that she had a close relationship with Cosio, and that denial was then rebutted with the substantial co-worker testimony to the contrary. But if, on cross-examination, A.A. were to have admitted feeling close to Cosio (certainly a real possibility), that testimony — while casting doubt on her fear of him — would also have tended to undermine Cosio’s fundamental defense that A.A. had resented him so much that she lied about the charges of sexual abuse, putting his life in serious jeopardy under the criminal law as a way of getting him out of her life (or at least away from interfering with her relationship with her tutors). While it may be true that one can resent, on occasion, a person she basically loves, it seems more solidly true that one who feels close, indeed loving, toward another person will not carry her occasional resentment to the point of putting the loved, though resented, person’s life in serious jeopardy. Resentment surely has to be pervasive to do that, and that kind of resentment is hard to prove to a jury when one also is trying to rebut the government’s “fear” argument with protestations of sustained closeness to the person resented. We are not convinced that a defense counsel — who of necessity had to demonstrate fabrication — can be held deficient for failing to present, and try to reconcile, both tactics.
Our doubts are reinforced by that fact that A.A. and Dr. Lindsay offered plausible reasons, other than fear, as to why A.A. did not report the sexual abuse earlier — reasons the government could stress to the jury even if its “fear” argument were contested. Thus, Cosio’s argument that his coworkers’ testimony would entirely undermine A.A.’s credibility by erasing her alleged fear of him rests on a premise too narrow to support reversal; that testimony would not have touched all the reasons offered for A.A.’s silence.
Appellate counsel reinforces his position, however, with trial counsel’s own testimony at the § 23-110 hearing. There, trial counsel answered “yes” when the prosecutor asked: would proof of a cordial relationship between Cosio and A.A. “have gone some distance in rebutting the picture of physical abuse and fear that [A.A.] had portrayed”? Would such proof have been helpful for use at closing argument “on the issue of whether or not there was a fear factor between [A.A.] and Mr. Cosio”? And would such proof “not only [have] throw[n] into question the whole issue of her delay in disclosure, but also throw[n] into question her credibility” in testifying that “she was afraid of him because of the beatings”?
It is not clear how trial counsel’s responses should be taken. Some defense counsel, we assume, will do everything possible to avoid the taint of “ineffective assistance” and therefore — admitting nothing — would fight any notion that an alleged deficiency, if corrected, would have helped the defense. Other defense counsel, we also assume, will become introspective, second-guess a losing strategy, and even fall on their proverbial swords to *176help, in their view, a deserving client. It is impossible to tell from trial counsel’s responses whether he is actually admitting constitutional deficiency or, we think more likely, is either acknowledging, upon reflection, that he could have done a better job or, even more likely, simply answering the questions asked without having an opportunity to place them in the larger context we have discussed above. In fairness to trial counsel, moreover, it is important to add, before finally evaluating his responses at the § 23-110 hearing, that in his affidavit submitted in response to the § 23-110 motion he noted that he “had performed a substantial amount of pretrial investigation”; that he and his investigator had interviewed Cosio’s mother, sisters, A.A.’s friends, and school teachers, and “asked them what they knew about the relationship between defendant and A.A.”; and that he had believed prior to trial that he “had investigated the relationship between Mr. ..Cosío and [A.A.] as fully as possible given the general constraints of time and resources.”
All things considered — indeed, in the absence of any questioning at the § 23-110 hearing that would have allowed trial counsel to discuss how (if at all) he would have used the coworker evidence in conjunction with a fabrication defense — we believe that his responses can have no major, let alone conclusive, significance as to his alleged deficiency. And in any event, they can have no bearing on the prejudice analysis, which his testimony never addressed. Case law applying Strickland not only warns against using “hindsight to second-guess [counsel’s] strategy choices,” Mayo, 13 F.3d at 533, but also teaches that when counsel reasonably pursues one line of defense, the pursuit of another line is not required. Provenzano v. Singletary, 148 F.3d 1327, 1330 (11th Cir.1998). That would be all the more true when a second line of defense would create tension with the first. In the end, the relevant inquiry is not what defense counsel could have done but, rather, whether the choices that defense counsel made were reasonable. Babbitt v. Calderon, 151 F.3d 1170, 1173 (9th Cir.1998). We are satisfied that defense counsel’s investigation, and the choices he made as a result, were constitutionally sufficient and, in any event, did not rise to the level of prejudice required by the Constitution for reversal of Cosio’s convictions.
III.
Cosío also argues that the trial court abused its discretion when it excluded a defense-proffered 1993 medical record of A.A.’s physical examination at Children’s Hospital when she was ten years old. Trial counsel sought admission of that record, as interpreted by the defense expert, Dr. John Adams, to impeach A.A.’s credibility. What was counsel’s theory? If Cosio had committed an act of carnal knowledge (rape) of A.A. between 1990 and 1991, the record of the 1993 examination would have shown injury to her hymen — but in fact did not. Thus, according to counsel, the 1993 medical record tended to show that A.A.’s claim of carnal knowledge in those early years was a lie.
The 1993 medical record did reflect that a doctor had examined A.A.’s genitalia and found that all was “normal.” .But there was nothing in the medical record itself to verify that A.A.’s hymen had been visually inspected, as it had been during the examination conducted by the government’s expert, Dr. Lindsay, years later in 1997. Furthermore, the examining physician who created the medical record was not available to testify. Thus, in addition to moving the 1993 medical record in evidence, the defense sought to introduce expert testimony by Dr. Adams to prove there had *177been no carnal knowledge by confirming that the examining physician would have inspected A.A.’s hymen before recording the genitalia as “normal.”
A.
In Dyas v. United States, 376 A.2d 827 (D.C.), cert. denied, 434 U.S. 973, 98 S.Ct. 529, 54 L.Ed.2d 464 (1977), we set forth a three-part test for the admission of expert testimony: (1) “the subject matter ‘must be so distinctively related to some science, profession, business or occupation as to be beyond the ken of the average layman’; (2) ‘the witness must have sufficient skill, knowledge, or experience in that field or calling as to make it appear that his opinion or inference will probably aid the trier in his search for truth’; and (3) expert testimony is inadmissible if ‘the state of the pertinent art or scientific knowledge does not permit a reasonable opinion to be asserted even by an expert.’ ” Dyas, 376 A.2d at 832 (emphasis omitted) (quoting McCORMICK ON EVIDENCE § 13, at 29-31 (E. Cleary, 2d ed.1972)).
In addition to satisfying the Dyas criteria, the “expert witness opinion must be based on fact or adequate data. It is properly received so long as it is not a mere guess or conjecture.” Sponaugle v. Pre-Term, Inc., 411 A.2d 366, 367 (D.C. 1980). “Expert opinion evidence generally should be admitted whenever it will not mislead the jury and will prove useful in understanding the facts in issue.” Clifford v. United States, 532 A.2d 628, 632 (D.C. 1987). A physician, in particular, does not lack competence merely because he or she is not a specialist in the particular field at issue; training and specialization go to the weight of the evidence, not to its admissibility. Baerman v. Reisinger, 124 U.S.App. D.C. 180, 181, 363 F.2d 309, 310 (1966). Medical experts thus may rely on tests and reports prepared by others. Clifford, 532 A.2d at 632.
The proffered expert testimony about the medical record in this case, of course, must be relevant; that is, it must tend “ ‘to make the existence or nonexistence of a fact more or less probable than would be the case without that evidence.’ ” Winfield v. United States, 676 A.2d 1, 2 (D.C.1996) (en banc) (quoting Punch v. United States, 377 A.2d 1353, 1358 (D.C. 1977), cert. denied, 435 U.S. 955, 98 S.Ct. 1586, 55 L.Ed.2d 806 (1978)). Once the trial court determines that the evidence is relevant, it must then weigh the probative value of that evidence against the risk of unfair prejudice and the likelihood that the evidence would mislead or confuse the jury. Id. at 5.
If, after applying all the foregoing criteria, the trial court excludes the proffered expert testimony and related evidence, this court on appeal will review that decision for abuse of discretion, and the trial court’s decision will not be reversed unless “manifestly erroneous.” Coates v. United States, 558 A.2d 1148, 1152 (D.C. 1989). This means that even if the trial court exercises its discretion in error, we must review the impact of that error in order to determine whether it is of a sufficient magnitude to require reversal. Johnson v. United States, 398 A.2d 354, 366 (D.C.1979).
B.
Before ruling on the proffered evidence, the trial court received conflicting testimony about whether a Children’s Hospital doctor would have examined the hymen of a ten-year-old girl during a standard “school physical.” Dr. Adams testified that he “assume[d],” based upon his “knowledge of the standard practice and that [1993] record,” that the examining *178doctor had visually inspected A.A.’s hymen, but he acknowledged it was possible that the finding of “normal” contained in the record resulted from an examination of only the external area of the genitalia.
Dissatisfied with Dr. Adams’s testimony, the trial court commented:
[M]y problem is as a woman, as a mother of two daughters, and watching the faces of the women in this courtroom who are observing, the notion of your expert who doesn’t do these routine examinations that in a routine examination of a young girl the labia is pulled aside and the hymen is observed ... just flies in the face of every experienced — [ ]... I won’t make a ruling based on that knowledge of mine but I certainly think that anything that flies in the face of common sense, directly needs to be explored further.”
The court asked the prosecution to call as a rebuttal witness Dr. Muriel Wolf, a pediatrician at the Children’s Hospital. Dr. Wolf testified that her personal practice was to visually inspect the hymen. She further testified, however — over vigorous defense objection — that based on her informal survey of eleven other physicians there was no uniform practice as to whether pediatricians at Children’s Hospital would visually examine the hymen during routine physical examinations.
The trial court then excluded the medical record on three interrelated grounds. First, the court noted that the record’s “arguable relevance” was “outweighed by prejudice here,” namely “the confusion and the suggestion that the premise on which relevancy depends has been established and it has not,” that is, “that the hymen was observed in 1993.” Second, the court found in the alternative that Cosio’s expert was not “qualified to examine routine child records” and to opine on whether the hymen would have been visually examined during A.A.’s 1993 physical. Third, the court added that “whether or not the hymen was observed in 1993” was “not really a subject of opinion testimony anyway. It’s fact.” Thus, “the factual predicate” for relevance was “not established.”
The latter two grounds for exclusion are problematic. As to Dr. Adams’s qualifications, the trial court record establishes that he was an experienced pathologist who regularly interpreted medical records in his work, including records resulting from the examination of children. The court accordingly had found Dr. Adams qualified to talk about medical records generally. Thus, the later finding that he was not qualified to testify about “routine” children’s medical records is perplexing— especially because a medical witness’s “training and specialization,” as well as the “degree of certainty” expressed, goes to the weight, not admissibility, of evidence. Baerman, 124 U.S.App. D.C. at 181, 363 F.2d at 310; Coates, 558 A.2d at 1152.
We also are doubtful about the trial court’s finding characterizing as “fact,” not as “opinion testimony,” the question whether A.A.’s “normal” examination of her genitalia in 1993 would have included an inspection of the hymen. There might well have been a standard practice that an expert could have verified. We think, however, that in offering this third ground for exclusion the trial court, in effect, was repeating the concern that she expressed in her first ground: that however relevant the 1993 record might have seemed in the abstract, there was too much uncertainty expressed by both Dr. Adams and Dr. Wolf about the likelihood that the hymen had in fact been examined — an uncertainty that would have led to prejudicial confusion and thus jury speculation. We turn, therefore, to this remaining ground for exclusion.
*179Dr. Adams was prepared to testify that observation of the hymen was a standard practice, and that the 1993 medical record showed no indication of any injury. Furthermore, the court’s own expert witness, Dr. Wolf, acknowledged that she personally would have examined the hymen. She also testified, however, that this practice was not universal; the decision by others to do so would depend on the practice of the individual doctor. Cross-examination of Dr. Adams, moreover, would have revealed that he had no personal knowledge of whether the examining physician had observed the hymen and that, as Dr. Wolf had testified, there was an unmeasurable possibility that A.A.’s hymen had not been examined.
In reviewing this evidentiary issue on appeal, we must give Dr. Adams’s proposed testimony “its maximum reasonable probative force and its minimum reasonable prejudicial value.” Ibn-Tamas v. United States, 407 A.2d 626, 639 n. 27 (D.C.1979) (quoting WEINSTEIN’S EVIDENCE ¶ 403[03] (1978)). There is some force to the argument that Dr. Adams’s bottom-line opinion — that the examining physician would have inspected A.A.’s hymen — was for the jury to consider under proper instructions; it was not for the judge to reject simply because Dr. Adams had acknowledged the possibility of no such inspection, and because another expert cast further doubt on the Adams opinion. The jury’s role, after all, is to weigh conflicting testimony and decide which to credit. Furthermore, although defense counsel did not object to the court’s calling its own expert witness (Dr. Wolf), counsel did object strenuously to the court’s admission of Dr. Wolfs particular testimony that relied on hearsay — contrary to her own personal medical practice — derived from an informal poll of unidentified doctors whose credentials were not established.3 That objection seems persuasive, given the lack of any discernible basis for saying that Dr. Wolf folded her informal survey into an expert opinion. It was never clear what Dr. Wolfs own bottom line opinion was: the same as Dr. Adams’s, based on Dr. Wolfs own experience? Or a contrary opinion, based on her informal survey, that a Children’s Hospital physician would not have examined the hymen? If the Adams and Wolf testimony were the entire universe of evidence on which ota’ ruling should be based, therefore, we would conclude that the trial court erred in allowing Dr. Wolfs testimony to scuttle Dr. Adams’s testimony and the accompanying medical record, and that this error, all things considered, would have been of a magnitude requiring reversal — and thus an abuse of discretion. Johnson, 398 A.2d at 367.
But the Adams and Wolf testimonies were not the only expert evidence that had relevance to the 1993 medical record. Even if the trial court erred in limiting Dr. Adams’s testimony and excluding that record, we believe that Cosio proceeds here from a flawed premise and suffered no *180significant prejudice from these trial court rulings.
We shall accept for purposes of discussion, as Cosio would have it, that the doctor had visually inspected A.A.’s hymen during her 1993 physical exam when he determined that the genitalia were “normal.” If A.A. had been raped in 1990 or 1991, the assailant — to quote A.A. — would have “put his penis in [her] vagina.” Otherwise, her hymen would probably not have been injured. But relying on both her “own experience and studies that she ha[d] read,” the government’s expert, Dr. Lindsay, testified — and defense counsel called on neither Dr. Adams nor anyone else to rebut her observation — that when “adult men rape young girls” (la, in statutory language have “carnal knowledge”), the penis typically “lie[s] in between the lips,” that is, “the labia majora,” while not “penetrating into the vagina,” in which case there would be “no” injury to the hymen. In other words, as the statute defines carnal knowledge, the assault could have occurred through such partial penetration, permitting the hymen itself to remain “normal.”4
Furthermore, Dr. Lindsay noted that even when there is enough penetration to injure the hymen, that kind of injury can “heal quickly” and “may or may not leave a sign or a scar.” Cosio offered no testimony from Dr. Adams (or anyone else) to rebut that expert assessment. Finally, although A.A. herself testified that Cosio had put his penis in her vagina when she was seven, Dr. Lindsay noted, without contradiction, that many times children believe the penis has entered the “vagina” when in fact it has penetrated only the labial lips.
Accordingly, a physical examination of A.A. in 1993 finding her hymen “normal,” would have been only marginally, if at all, probative of whether there had been a rape (carnal knowledge) of this young girl in 1990 or 1991. Dr. Lindsay’s uncontra-dicted testimony demonstrated that even if A.A.’s examining physician had observed her hymen in 1993 and found it “normal,” it would require pure speculation to conclude that no rape had occurred.
In sum, the testimony of Dr. Adams, Dr. Lindsay, and Dr. Wolf, individually and collectively, left room for substantial doubt that A.A.’s hymen had been inspected. But even if it had, there was uncontradict-ed expert testimony that, because of A.A.’s young age, her assailant would probably not have penetrated her vagina and injured her hymen. Furthermore, uncontra-dicted expert testimony showed that even if penetration had occurred, the hymen might well have healed during the two or more years between the assault and the “normal” physical exam. Finally, external inspection of the genitalia by the examining physician in 1993 would have had no probative value for a criminal offense premised on sexual penetration. Therefore, we cannot say that the trial court’s refusal to allow the jury to inspect the 1993 record, or to hear what Dr. Adams would say about it, brought any meaningful prejudice to Cosio’s defense; the probative value in his favor would have been virtually nil once the government had demonstrated from uncontradicted testimony, based on expert studies, that the presence of a “normal” hymen in 1993 was a virtual *181irrelevancy. Any use of that record in Cosio’s favor would have been speculative at best. Based on all the evidence, there was no abuse of trial court discretion in withholding the medical record and Dr. Adams’s comments about it from the jury.5
RESPONSE TO DISSENT
The dissent’s approach to the ineffectiveness-of-counsel claim requires further comment. The Supreme Court has recognized, in Strickland, that for every defense of a criminal case there is a “wide range of reasonable professional assistance” and, accordingly, that there must be a “strong presumption” that counsel has given constitutionally adequate assistance. Strickland, 466 U.S. at 689, 104 S.Ct. 2052. With that teaching in mind, we have stressed that, if counsel had learned of the coworker evidence but elected not to use it, that tactical decision would not have been constitutionally deficient; Strickland would have granted leeway for counsel’s professional judgment as to whether that coworker evidence would have undermined or aided Cosio’s fabrication defense. On the facts of this case, not even an imputed awareness of the coworker evidence would have required trial counsel — and thus does not require this court — to substitute a different trial strategy for the one counsel himself employed in ignorance of that evidence.
The case from this jurisdiction that comes closest to informing the analysis is Chatmon v. United States, 801 A.2d 92, 108 (2002). In that case, in which the defendant was convicted of first-degree felony murder and armed robbery, defense counsel, through cross-examination of a government witness, opened the door to admission of damning identification testimony that the government had agreed not to proffer because of defective identification procedures. At the hearing on collateral attack alleging ineffective assistance of counsel, trial counsel offered a highly questionable rationale for why he had invited the identification testimony — a rationale that the trial judge said made “no sense.” The government, however, offered several arguments explaining why counsel’s tactic actually did make sense. This court thus confronted the question whether Strickland permitted a reviewing court to ignore a deficient rationale for defense counsel’s performance by invoking a sound, alternative rationale that counsel had not considered and, as a result, to save the conviction. Judge Ruiz said no: “Analysis under Strickland is highly fact bound .... Therefore, once the record establishes the actual tactical explanation for counsel’s actions” (which was nonsensi*182cal in Chatmon), “the government is not free to invent a better-reasoned explanation of its own.” Id. at 108-109.
Chatmon, however, also recognized that, under Strickland, a defendant on collateral attack must “ ‘show that counsel’s representation fell below an objective standard of reasonableness.’ ” Chatmon, 801 A.2d at 108 (quoting Strickland, 466 U.S. at 688, 104 S.Ct. 2052). And, as we have noted earlier, Strickland forbids a finding of constitutional deficiency if counsel has made a “reasonable decision” to offer a defense that made “unnecessary” the particular investigation[ ]” that defendant has cited on collateral attack as an omission of constitutional dimension. Strickland, 466 U.S. at 691, 104 S.Ct. 2052. We believe these principles support the following proposition: If it is true (as we explained earlier and elaborate further below) that Cosio’s trial counsel would not have been constitutionally deficient had he known about the coworker evidence and consciously decided to withhold it, we do not inn afoul of Strickland, as interpreted in Chatmon, by saying — on the facts here— that trial counsel was not deficient in failing to discover and introduce that evidence. By relying on this hypothetical proposition to make the point, we are not widening the universe of information “to invent a better-reasoned explanation” than the record permits; we are merely explaining the “objective reasonableness” of the defense presented. Unlike the blatant sin of commission in Chatmon, which this court would not permit the government to rationalize away, we consider here an alleged sin of omission, which the enhanced record on collateral attack does not permit us to characterize as a prejudicial trial deficiency under Strickland.
In the first place, the “actual tactical explanation” for trial counsel’s actions in this ease — unlike the explanation in Chat-mon that made “no sense” — is simple, straightforward, and plausible. Counsel sought to establish a fabrication defense through (1) cross-examination of A.A. that would reveal her deep resentment of Co-sío, coupled with (2) introduction of a medical record — interpreted by expert testimony — contradicting A.A.’s assertion that she had been raped in 1990. In short, counsel sought to establish that A.A. was a liar. No one, and certainly not our dissenting colleague, has suggested that this defense was inherently deficient.6 To the contrary, the dissent argues only that counsel could have provided a better defense by adding coworker testimony to rebut A.A.’s claimed fear of Cosio.
The question, then, is not — as in Chat-mon — whether a reviewing court can accept a government-proposed cure of an inherently deficient rationale for counsel’s performance by identifying a sound rationale that counsel failed to discover. The question, rather, is whether a reviewing court can reject an inherently sufficient rationale for counsel’s performance by faulting counsel’s failure to discover evidence that allegedly would have made the defense stronger.
The answer might be yes if counsel had made no effort to discover the kind of evidence at issue. But that was not the *183situation here; trial counsel explored A.A.’s relationship with Cosio through questioning those in the best position to know: their family members. This is not a case like Kimmelman v. Morrison, 477 U.S. 365, 106 S.Ct. 2574, 91 L.Ed.2d 305 (1986), relied on by the dissent, premising reversal for ineffective assistance of counsel on “a total failure to conduct pretrial discovery,” indeed a “complete lack of pretrial preparation.” Id. at 385, 386, 106 S.Ct. 2574.
We would agree that counsel also might be held deficient if counsel’s failure to discover particular evidence (such as the coworker testimony), if negligent, deprived the defendant of testimony powerful enough to create a reasonable probability that the outcome of the trial would have been different — an analysis effectively merging the deficiency and prejudice prongs of Strickland. But, as we have already explained, that is not the case here, either. The co-worker evidence might or might not have helped the Cosio; its admission might well have tended to undermine the fabrication defense, creating a tough judgment call for defense counsel if he had been apprised of the evidence. Accordingly, we do not believe that a reviewing court can declare counsel constitutionally deficient simply because the appellate judges, looking back over the trial record, conclude that they could have found a way — had they been trial counsel — to finesse the tension between the coworker testimony (demonstrating closeness between A.A. and Cosio) and the fabrication defense (demonstrating how much A.A. resented Cosio).
The dissent, however, offers still another, final answer to our analysis. In effect, the dissent calls the tension we perceive between the coworker testimony and the fabrication defense a red herring and, as a result, claims that trial counsel’s defense falls below Strickland’s standard of “objective reasonableness.” Why? Because the closeness the co-workers saw between A.A. and Cosio appears to have predated A.A.’s relationship with her tutors, and thus — it is claimed — predated the resentment on which her fabrication was built. In short, the dissent suggests, defense counsel could (and thus should) have rebutted A.A.’s alleged fear of Cosio with the co-worker testimony, followed by an argument that she had lied about the sexual abuse because she had come to resent Cosio’s attempted interference with her mentor relationship.
We do not believe that the defense necessarily would sort out into such neat, sequential and thus separable packages. To rebut A.A.’s alleged fear of Cosio over a long period of time, counsel would have had to use the co-worker testimony to demonstrate that A.A. was very, very close to Cosio for a long period of years extending through October 1997, when the last abusive sexual act allegedly took place. Counsel could not have stopped with showing merely moderate closeness; otherwise, A.A.’s fear might not have been erased in the jury’s mind. If counsel were to establish such a tight bond between them, however, we cannot be confident that a jury would then find it credible that a young woman, portrayed in that intimate, loving way, would have turned on her half-brother very suddenly — to the point of charging him with a multitude of very serious, abusive crimes — within days of meeting her new tutors in the fall of 1997. It is just as likely, if not more so, that the jury would have found such a defense presentation confusing, counterintuitive, and thus unpersuasive.7
*184Furthermore, the coworker evidence, if offered to rebut A.A.’s claimed fear of Cosio, would have had no bearing on the fear A.A. expressed for the first period in which she claimed Cosio had raped her (in 1990), since none of the co-workers observed her with Cosio at IDP until she had returned to D.C. from boarding school several years later. Finally, as elaborated earlier, fear was not the only reason the government proffered (with the help of expert testimony) for A.A.’s failure to tell anyone about her sexual abuse. Accordingly, the coworker testimony, even if presented at trial, would not have been a complete answer to the so-called “fear factor.” In sum, we cannot say that, if trial counsel had been aware of the coworker testimony, the only objectively reasonable trial strategy would have required counsel to introduce that evidence in the defense case.
To repeat: in pointing out the various reasons — all based in the record — why competent counsel might not have chosen to offer the coworker testimony at trial, even if aware of it, we are not providing rationales that cannot be imputed to trial counsel under the teaching of Chatmon. In that case, trial counsel had no acceptable — no objectively reasonable — -rationale for his performance, and we would not permit the government to think one up for him. Here, in contrast, trial counsel had a plausible theory of the defense. And the record reflects sound reasons why the alternative (or supplementary) theory, proffered on collateral attack, would not necessarily have been so much more effective than the defense presented that we can say counsel’s performance fell below the “objective standard of reasonableness” required by the Constitution. Strickland, 466 U.S. at 688, 104 S.Ct. 2052.
Because Cosio has not demonstrated constitutionally ineffective assistance of trial counsel or the abuse of trial court discretion in excluding evidence and testimony concerning A.A.’s 1993 medical examination, we must affirm all convictions.
So ordered.
. Effective May 23, 1995, the crimes of taking indecent liberties with a minor child and carnal knowledge were incorporated into the definitions of first and second degree child sexual abuse. Because the charged acts of abuse took place both before and after May 23, 1995, separate counts as to each were charged, and submitted to the jury, under the statutory provision in effect at the time of each alleged incident. .
. Contrary to the dissent at 68, this argument does not necessarily presuppose a "statutory rape” situation.
. Without question, the trial court’s decision to exclude the 1993 medical record and Dr. Adams's related expert testimony was based in part on the results of the court’s intervention into the proceedings by calling Dr. Wolf as the court’s own witness. Although "a judge is not an investigator,” Davis v. United States, 567 A.2d 36, 42 (D.C. 1989), a trial judge does have "the right to call witnesses on [her] own initiative when [she] has reason to feel the witness may contribute to a determination of the truth as to the matter in controversy.” Fortune v. Fortune, 138 A.2d 390, 391 (D.C.1958). Cosio has not challenged the conduct of the trial judge in calling a court witness as such (in contrast with challenging the admissibility of a portion of that witness’s testimony), and thus we need not inquire further into the propriety of the court’s decision to call Dr. Wolf as a witness.
. The "carnal knowledge” (rape) statute that would have been applicable to an assault in 1990 or 1991, D.C.Code § 22-2801 (1981), requires as "an essential element of the crime” the "penetration of the victim's sexual organs.” Williams v. United States, 357 A.2d 865, 867 (D.C.1976). But “the government need not prove full penetration since the offense is committed if the male organ enters only the labia of the female organs.” Id.
. Citing Scull v. United States, 564 A.2d 1161, 1166 (D.C.1989), Cosío asks us to apply not "abuse of discretion” but “constitutional harmless error” as the standard of review, meaning that reversal would be required unless the trial court's error in excluding the evidence was harmless beyond a reasonable doubt. See Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967). Scull applied the Sixth Amendment confrontation clause to exclusion of evidence during cross-examination of a witness for bias — an issue of constitutional dimension under the elaboration of Chapman which this court spelled out in Clark v. United States, 639 A.2d 76 (D.C.1993). There, we said that constitutional harmless error analysis applies where “the trial court’s evidentiary ruling wholly deprived the defendant of any opportunity to cross-examine a witness or present evidence concerning bias or a central issue in the case....” Id. at 81. In this case, however, the exclusion of the medical record and Dr. Adams's related testimony had a bearing not on bias but on A.A.’s general credibility. Although her credibility was a "central issue,” id., defense counsel was not precluded from cross-examining A.A. and questioning her credibility in other ways. Accordingly, we are not presented with an evidentiary issue of constitutional dimension.
. As indicated in Part III above, the trial court — we conclude erroneously — refused to admit the medical record and related expert testimony in evidence, although we have further concluded that this error was not of a magnitude requiring reversal and thus (in the language of our standard of review) was not an abuse of discretion. Accordingly, the failure of the defense to proceed on all cylinders with its intended defense was not enough in itself to warrant reversal — a conclusion with which the dissent does not disagree. Nor does appellant or the dissent attempt to link this failure of the proffered defense to the claim of ineffective assistance of trial counsel now under discussion.
. Nor could counsel have argued convincingly to the jury that if A.A. was really afraid of *184Cosio, she would have been less likely to bring charges against him than if she was not afraid of him, and thus that the coworker testimony would have complemented, not contradicted, the fabrication defense. That argument presupposes a profound resentment simultaneously present in a loving half-sister — a psyche not necessarily easy to grasp. Moreover, the argument misses the point, for once A.A. felt she was in the safe harbor of -her tutors, it is not particularly difficult to understand why her intense fear of Cosio, while not erased, might have been neutralized by courage, coupled with encouragement, to the point of A.A.'s willingness to press charges.