McMullin v. Amsberry

[542] Argued and submitted March 14, 2019; reversed and remanded with instructions to grant relief on inadequate investigation claim, otherwise affirmed April 7, 2021 KEITH ALLEN McMULLIN, Petitioner-Appellant, v. Brigitte AMSBERRY, Superintendent, Two Rivers Correctional Institution, Defendant-Respondent. Umatilla County Circuit Court 16CV23691; A164404 485 P3d 278 Petitioner appeals a judgment denying his petition for post-conviction relief, which alleged inadequate assistance of counsel at his trial for the rape, sodomy, and sexual abuse of his adopted adolescent daughter. Petitioner claims that his lawyer’s failure to adequately investigate whether to dispute the testimony of the state’s witness concerning the significance of the absence of physical evidence of abuse violated his guarantee to adequate representation, in violation of both the state and federal constitutions. Held: Petitioner’s trial lawyer’s strategic decision was not based on a sufficient investigation of the state’s witness testimony and the significance of the absence of physical evidence of abuse. That decision therefore violated petitioner’s right to adequate counsel. As a result, petitioner suffered more than a mere possibility that adequate representation would have had a tendency to affect the result of the trial. Reversed and remanded with instructions to grant relief on inadequate investigation claim; otherwise affirmed. Dale Penn, Senior Judge. Ryan Scott argued the cause and filed the briefs for appellant. Patrick M. Ebbett, Assistant Attorney General, argued the cause for respondent. Also on the brief were Ellen F. Rosenblum, Attorney General, and Benjamin Gutman, Solicitor General. Before Aoyagi, Presiding Judge, and DeVore, Judge, and Landau, Senior Judge.* ______________ * DeVore, J., vice Hadlock, J. pro tempore; Landau, S. J., vice DeHoog, J. Cite as 310 Or App 542 (2021) 543 LANDAU, S. J. Reversed and remanded with instructions to grant relief on inadequate investigation claim; otherwise affirmed. 544 McMullin v. Amsberry LANDAU, S. J. Petitioner was charged with numerous counts of first-degree rape, first-degree sodomy, and first-degree sexual abuse arising from his alleged sexual abuse of his adopted adolescent daughter. The nurse who examined the complainant found no physical evidence of sexual activity, but she testified at trial that it was not unusual for an adolescent victim of sexual abuse not to show physical evidence of abuse. Petitioner’s trial counsel did not call an expert witness to rebut the nurse’s testimony in that regard. Nor did counsel cross-examine her about her assertion that the absence of physical signs of abuse was not unusual. Petitioner ultimately was found guilty of most of the charges. Petitioner sought post-conviction relief. He claimed that his criminal trial counsel was constitutionally inadequate, in violation of Article I, section 11, of the Oregon Constitution and the Sixth and Fourteenth Amendments to the United States Constitution, because she failed to adequately investigate whether to dispute the testimony of the state’s witness concerning the significance of the absence of physical evidence of abuse. The state1 contended that trial counsel’s investigation amounted to a reasonable strategic decision under the circumstances that, in any event, did not prejudice petitioner. The post-conviction court agreed with the state and entered judgment denying petitioner’s claim for relief. Petitioner appeals, reprising his claims that his criminal trial counsel was constitutionally inadequate. We conclude that the post-conviction court erred and that petitioner is correct that trial counsel was constitutionally inadequate in failing to adequately investigate the significance of the absence of physical evidence of sexual abuse. We therefore reverse and remand with instructions to enter judgment ordering a new trial. [1] Respondent is the Superintendent of the Two Rivers Correctional Institution. We refer to the superintendent as “the state” in this opinion. See, e.g., Richardson v. Belleque, 362 Or 236 , 238 n 1, 406 P3d 1074 (2017) (referring to the superintendent as “the state”). Cite as 310 Or App 542 (2021) 545 I. RELEVANT FACTS AND PROCEEDINGS BELOW A. Before the Criminal Trial Petitioner’s 13-year-old adopted daughter reported that petitioner had been repeatedly sexually abusing her since she was five years old, starting with sexual touching and evolving to oral sodomy and, eventually, to repeated intercourse over several years. Following the report, Smith, a nurse practitioner at the Children’s Center who specialized in child sexual assault, conducted a medical examination of the complainant. Smith observed that the child had “some decreased hymenal tissue” in one place but that it was a “normal variation” for a girl her age. Although there was no physical evidence of sexual abuse, Smith did not “expect” to find any. The state ultimately charged petitioner with 10 counts of first-degree rape, five counts of first-degree sodomy, and 10 counts of first-degree sexual abuse. Karabeika was appointed to represent petitioner. Karabeika anticipated that Smith would be called to testify for the state and that the nurse practitioner would state that it was not unusual for an adolescent victim of sexual abuse to show no physical symptoms of such abuse. Karabeika “struggled” with that idea and decided to consult with Dr. Brady, a former medical examiner. She sent Brady the examination report, but he responded that he would not be a helpful witness for the defense. Brady had reviewed the report but concluded that “he can’t really help us.” As Karabeika later recalled, Brady told her, “I’ve been retired for a long time. I haven’t performed an exam in a while and I don’t think I can help you.” Karabeika then located a second expert, Fitzgerald, a nurse practitioner whom she thought might be helpful. But after consulting with Fitzgerald, she concluded that the expert “had some really wacky ideas that weren’t necessarily supported by the information I wanted to talk about.” As she explained to petitioner, Fitzgerald was “an odd duck and not terribly likeable and frankly she goes off on tangents and I don’t think she’d help us.” Karabeika also consulted with two other attorneys. As she later recalled, both lawyers told her that “[y]ou’re 546 McMullin v. Amsberry not expected to see damage unless there’s an acute incident. There’s really not going to be—she’s a menstruating teenager and she’s of a certain age. There is not going to be findings one way or the other.” One of the lawyers, Cohen, gave what Karabeika regarded as a “glib” response that “you can’t make nothing out of no hymen anymore.” The other lawyer, Maxfield, told Karabeika that “[i]f you plan to aggressively cross the expert on the ‘normal’ exam, we should talk about which studies she is likely to hide behind and what weaknesses there are in those studies.” Maxfield noted that “[t]here is plenty of literature to suggest that it would be highly unusual for a girl to suffer repeated ‘blunt force trauma’, especially recently and not have physical evidence that the hymen has been torn.” Maxfield cautioned, however, that such cross-examination “needs to be done surgically or it can blow up.” Maxfield suggested a different approach, one that appealed to the jurors’ common sense: “Invite them to use their common sense. Does it make sense that a very thin, taut, nonelastic hymen would not tear when a hard penis that is two or three times the size of the opening is thrust into the vagina again and again? There is a reason married women no longer have hymens. Penetration tears the hymen. You know that. Yet the state wants you to believe that an adult erect penis was thrust into [the victim] again and again, twice a week for years and it never tore.” Karabeika ultimately decided not to retain an expert witness to testify for the defense and further decided not to challenge on cross-examination Smith’s expected testimony concerning the insignificance of any physical evidence of sexual abuse. Based on her conversation with Maxfield, she concluded that the strategy of aggressively cross-examining Smith “could blow up in your face. It could backfire and you could basically give [the state] another platform to say the same thing that they’ve already said but in a new way with new information.” B. The Criminal Trial Proceedings At trial, Karabeika’s strategy was to challenge the complainant’s credibility and, without expert testimony, Cite as 310 Or App 542 (2021) 547 appeal to the jurors’ “common sense,” asserting the unreasonableness of the idea that an adolescent girl could be repeatedly sexually abused without any resulting physical evidence of abuse. In her opening statement, Karabeika told the jury that people working at the Children’s Center “diagnose or at least look for injuries to support allegations of child abuse.” Nevertheless, she continued, “despite the fact that [the complainant] claimed to have been abused for nine years and having intercourse for two-plus years, there isn’t one shred of proof during that physical exam that she has been sexually abused.” During the trial itself, Karabeika elicited from one of the state’s witnesses that the child had a reputation for being untruthful, and from the complainant’s pediatrician’s testimony that the child had reported to him that she had never had any sexual experiences nor ever been forced to do something sexual that she did not want to do. Petitioner also testified, denying ever having any sort of sexual contact with the child. As Karabeika expected, Smith testified that, although the complainant had reported multiple instances of sexual intercourse with petitioner, the child’s physical examination was normal. Smith explained that a “Kellogg” study2 of 36 pregnant adolescents established that “you can have a penetrating trauma and have a completely normal exam.” Karabeika did not object to Smith’s testimony and conducted no cross-examination about the Kellogg study or the witness’s assertion that she had “expected” a normal physical examination despite the allegations of repeated penetrative trauma. In closing argument, Karabeika returned to her chosen strategy: “The physical piece of this doesn’t add up. [The complainant] is about 5 feet tall, maybe shorter than that. She alleges nine years of abuse, two and a half years of being raped on a regular basis daily by her father, who is a large man. They say there are no physical findings that she was raped. No physical findings that there are any injuries to [2] Nancy D. Kellogg et al, Genital Anatomy in Pregnant Adolescents: “Normal” Does Not Mean “Nothing Happened,” 113 Pediatrics, no 1 at e67 (Jan 2004), https://pediatrics.aappublications.org/content/pediatrics/113/1/e67.full.pdf (accessed Mar 18, 2021). 548 McMullin v. Amsberry her. How is it possible that this small child could be forced to be raped this many times and for there not to be one shred of proof that she was abused in this action? Now the Children’s Center people said, ‘Oh well, that’s what we expected.’ Well, that’s not true, because that’s why they do this exam. They are looking for evidence of injuries or tearing or scarring or something. She had nothing—nothing. So that gives you nothing to go on except, wow, is she telling us the truth. Why is she saying she is being raped this often? Wouldn’t there be some evidence of it? Nothing.” The jury returned verdicts of guilty on 10 counts of first-degree rape, two counts of first-degree sodomy, and 10 counts of first-degree sexual abuse. C. The Post-Conviction Hearing Petitioner filed a petition for post-conviction relief. He alleged, among other things, five claims for relief related to Karabeika’s handling of Smith’s testimony: (1) Trial counsel “failed to retain a qualified expert witness to help her prepare to cross-examine the state’s expert on the ‘Kellogg Study’ ”; (2) counsel “failed adequately to cross-examine * * * Smith to show that her understanding of the Kellogg study was incorrect”; (3) counsel “failed to prepare and investigate the scientific and medical issues in the state’s case”; (4) counsel “failed to discover and understand the relevant studies that undercut the Kellogg study”; and (5) counsel “failed to retain and call a qualified expert to testify that there was a high likelihood that evidence of abuse would have been present because of the number of reported instances of intercourse.” The state argued that Karabeika did consult with at least two experts, as well as two lawyers with experience trying sexual abuse cases. In consultation with those resources, she decided not to attack Smith’s testimony and its reliance on the Kellogg study. That, the state argued, was a strategic call that was at least reasonable under the circumstances. In support of his post-conviction claims, petitioner offered the testimony of Dr. Guertin, a child abuse physician; medical director of the children’s center and director of the pediatric care unit at Sparrow Hospital in Lansing, Cite as 310 Or App 542 (2021) 549 Michigan; and associate professor of pediatrics at Michigan State University School of Medicine. Guertin testified that, had Karabeika retained him in this case, he could have suggested “many ways [counsel] could have cross-examined Ms. Smith on the limitations” of the Kellogg study. First, Guertin noted that the Kellogg study was flawed in that it relied solely on a retrospective review of photographs of 36 adolescents, which he explained is “not the same as being able to literally go all the way around the hymen and then record what you’re seeing.” He explained that the study was further flawed in that the authors failed to recognize that, of the 36 adolescents, eight had inconclusive exams, and five had abnormal exams. Finally, he explained that the study was also flawed because the authors had adopted an “extremely restrictive” definition of what constituted physical evidence of abuse. Second, Guertin testified that Smith had relied on only a single study, which he characterized as “a disservice to the jury, because you’re not giving them proper perspective.” He explained that “the literature actually is all over the place in terms of what the frequency is of seeing” physical evidence of sexual abuse. He cited as an example a second Kellogg study—by the same lead author of the study Smith relied on—involving adolescent girls who had admitted to having vaginal intercourse. He said that the study showed that 63 percent of the older subjects had the sort of hymenal “notches” that indicated intercourse, which he said casts doubt on the other study: “So in the same year, 2004, the same author showed that there are findings that are—that are—that actually are typical for intercourse and showed that in the most experienced population, that is the older kids, incidence of abnormality was 63 percent. “So to be fair to a jury, if you look at the same author in the same year, you have one study that says 18 percent abnormal in a fairly flawed study and you have one study that says 63 percent abnormal in a far less flawed study.” Moreover, he said, the second study concluded “that those particular physical findings are really important and 550 McMullin v. Amsberry they are indicators of sexual trauma, of penetrative sexual trauma.” Guertin also cited as another example a study cited in the second Kellogg study, which reported that of a group of children who had reported penile-vaginal intercourse, 74 percent had hymenal injuries. Guertin continued: “So it’s really interesting that you have a lot of consistency between those two studies and they are a far cry from the single study that was quoted by [Smith]. “So I think to be fair, if you’re in front of a jury, if you’re asking me what I would have told the attorney, I would have told the attorney that there’s perspective that needs to be created here and that that perspective is found in the literature and it wasn’t presented.” Guertin concluded that, in his opinion, had Karabeika “been properly prepared, she could have substantially undermined Smith’s conclusion that in the circumstances of this case, a normal physical examination of [the victim] was to be expected.” On cross-examination, Guertin acknowledged that normal genital findings are not, in general, uncommon for sexually abused children. He explained, however, that the key to such findings is that they are based on populations of both pre-pubertal and post-pubertal girls. He said that studies of older adolescents, particularly those older adolescents who have reported penile-vaginal intercourse, there is commonly a high rate of abnormal physical findings. The state countered with an affidavit of Smith, who asserted that some of the studies on which Guertin relied were dated. She said that, had she been cross-examined, she would have cited other studies than the Kellogg study that she mentioned at trial, which support the view that a normal physical examination of sexual abuse victims is to be expected. The post-conviction court concluded that petitioner failed to prove the allegations of his petition, noting that it agreed with and adopted the state’s arguments. Cite as 310 Or App 542 (2021) 551 II. ANALYSIS A. Standards of Review Article I, section 11, of the Oregon Constitution guarantees a criminal defendant the right to the assistance of adequate counsel. Farmer v. Premo, 363 Or 679, 690 , 427 P3d 170 (2018). To establish a violation of that right, a petitioner must prove two things—first, that trial counsel “failed to exercise reasonable professional skill and judgment,” and second, that petitioner “suffered prejudice as a result of counsel’s inadequacy.” Johnson v. Premo, 361 Or 688, 699 , 399 P3d 431 (2017). Whether a petitioner proved those two things presents a question of law. Green v. Franke, 357 Or 301, 312 , 350 P3d 188 (2015). The Sixth Amendment likewise guarantees the right to the assistance of adequate—or, in federal parlance, “effective”—counsel. To establish a violation of that right, a petitioner must also prove both that trial counsel was deficient and that the deficiency caused prejudice. Strickland v. Washington, 466 US 668, 687 , 104 S Ct 2052 , 80 L Ed 2d 674 (1984). The Oregon Supreme Court has concluded that the federal Sixth Amendment standard of proof is “functionally equivalent” to the state standard under Article I, section 11. Montez v. Czerniak, 355 Or 1, 6-7 , 322 P3d 487 , adh’d to as modified, 355 Or 598 , 330 P3d 595 (2014). In this case, however, we find that state law is adequate to dispose of the issues on appeal. B. Arguments on Appeal Petitioner argues that Karabeika failed to exercise reasonable skill and judgment in failing to retain an expert and more thoroughly investigate the strategy of challenging Smith’s reliance on the Kellogg study for the proposition that a normal physical examination is to be expected from an adolescent girl who has been repeatedly vaginally raped. Petitioner argues that such a strategic decision must be based on a reasonable evaluation of the costs and benefits of pursuing such an investigation. He argues that, without consulting a sufficiently knowledgeable expert, Karabeika could not evaluate those costs and benefits. 552 McMullin v. Amsberry According to petitioner, counsel’s evaluation of the matter consisted of briefly discussing the case with two experts. The first, Brady, explained that he had been retired a long time and, without further explanation, said that he could not provide any help. The second, nurse practitioner Fitzgerald, counsel immediately dismissed as “wacky” and unreliable. Petitioner notes that in neither case did Karabeika actually discuss the possibility of challenging Smith’s testimony. Petitioner acknowledges that Karabeika also discussed the matter with two lawyers who had experience defending sexual abuse cases. The first of the lawyers, he argues, simply responded that “no hymen” is not a fruitful strategy “anymore,” without providing further detail. And petitioner notes that the second, Maxfield, actually told Karabeika that “[t]here is plenty of literature to suggest that it would be highly unusual for a girl to suffer repeated blunt force trauma especially recently and not have physical evidence that the hymen has been torn.” Petitioner concedes that Maxfield did caution Karabeika that challenging Smith’s testimony “needs to be done surgically or it can blow up.” But, he insists, Karabeika never looked for the literature that Maxfield mentioned; without even knowing what that literature said, she simply concluded that she would pursue a different strategy, one that did not involve directly challenging Smith on cross-examination. That decision, petitioner argues, fell below the requirements of constitutionally adequate assistance. Moreover, petitioner argues, Karabeika’s failure to investigate prejudiced him. Had she followed up on Maxfield’s suggestion that there is “plenty of literature” to suggest that it is actually unusual for an adolescent child victim of rape not to show physical evidence of that abuse, she would have been in a position to undercut Smith’s reliance on a single study for a contrary conclusion. And offering the testimony of an expert witness such as Guertin could have been especially significant, petitioner argues. He argues that Guertin’s critique of the Kellogg study and his reference to other studies showing that between 63 and 74 percent of adolescent girls reporting sexual activity showed physical symptoms strongly suggests that it is significantly less Cite as 310 Or App 542 (2021) 553 likely that sexual intercourse occurred in this case, where there were no physical symptoms at all. The state contends that Karabeika reasonably decided not to investigate further whether to challenge Smith’s testimony. According to the state, counsel “based her decision on a full investigation, including consultation with multiple experts, and with multiple attorneys specializing in the field.” In any event, the state argues, petitioner failed to prove that any deficiency in Karabeika’s representation at trial caused him prejudice. The state argues that Smith’s testimony was “minor” in the larger scheme of things. There was, the state contends, “plenty of far more significant evidence for the jury to evaluate” in determining whether the complainant was telling the truth. In the state’s view, that more significant evidence consisted of several witnesses about her report of the abuse, a recorded Children’s Center interview with the complainant, and the testimony of the complainant herself. The state suggests that, if Karabeika had put Guertin on the stand, his testimony would not have refuted Smith; it would simply have precipitated a battle of experts, with each citing studies to support their opinions. Moreover, the state contends, Guertin would have had to concede—as he did during the post-conviction hearing—that it would not be unusual for a child victim of sexual assault to have a normal exam. C. Discussion 1. Reasonable skill and judgment We begin with the question whether trial counsel Karabeika failed to exercise reasonable skill and judgment in failing to investigate whether to challenge the state’s expert’s assertion that a normal physical examination is to be expected from an adolescent victim of sexual abuse. In Krummacher v. Gierloff, 290 Or 867, 875 , 627 P2d 458 (1981), the Oregon Supreme Court held that the constitutional right to counsel requires that counsel investigate the relevant facts and the controlling law “to the extent 554 McMullin v. Amsberry appropriate to the nature and complexity so that [counsel] is equipped to advise [the] client, exercise professional judgment and represent the defendant in an informed manner.” The standard “is not whether counsel investigated or introduced every shred of evidence.” Montez, 355 Or at 16. It is instead whether the extent of the investigation was “based on a reasonable evaluation of the likely costs and potential benefits of pursuing the investigation.” Stevens v. State of Oregon, 322 Or 101, 109 , 902 P2d 1137 (1995). In our evaluation of trial counsel’s calculation of the costs and potential benefits in this case, we are informed by several Supreme Court decisions. First, in Stevens, the petitioner was charged with first-degree rape of a young girl. Id. at 103 . With no witnesses to the assault itself and no physical evidence, the trial was essentially a credibility contest between the victim and the petitioner. Id. Following conviction on the lesser-included offense of second-degree sexual abuse, the petitioner filed a petition for post-conviction relief. Id. at 104 . He alleged that his trial counsel was constitutionally inadequate because of a failure to investigate properly. Id. at 104-05 . Specifically, the petitioner alleged that his trial counsel failed to interview any members of the victim’s school’s staff or her classmates, relying instead on the police report to identify material witnesses. Id. at 105-06 . He also complained that his trial counsel had failed to obtain medical diagnosis that he had been impotent at the time of the assault. Id. at 104-05 . At the post-conviction hearing, the petitioner introduced the testimony of two teachers, a teacher’s assistant, and four classmates, each of which contradicted the testimony of the victim in various aspects. Id. at 105-06 . He also introduced an affidavit of a urologist, which included the opinion that the petitioner had been impotent for years and was unable to attain an erection at the time of the assault. Id. at 106 . The state relied on the testimony of the petitioner’s trial counsel, who stated that the extent of his investigation was a matter of strategy under the circumstances of that case. In particular, counsel argued that he did not obtain Cite as 310 Or App 542 (2021) 555 a medical diagnosis of his client because he believed that presenting that sort of evidence would risk the petitioner’s credibility. Id. at 106-07. The Supreme Court agreed with the petitioner. The court explained: “In investigating a case, a lawyer inevitably is faced with choices as to what avenues of investigation to pursue. A ‘tactical decision’ in the course of an investigation is a conscious choice by a lawyer either to take or to omit some action on the basis of an evaluation of the nature and complexity of the case, the likely costs and potential benefits of the contemplated action, and other factors. But the fact that a lawyer has made a ‘tactical decision’ does not mean that the lawyer’s choice meets the constitutional standard for adequate assistance of counsel.” Id. at 109. In that case, the court concluded, there was no evidence that counsel had engaged in any sort of cost-benefit determination; rather, it appeared to the court that counsel had merely relied on the list of witnesses from the police report. Id. at 109-10. Second, in Richardson v. Belleque, 362 Or 236 , 406 P3d 1074 (2017), the petitioner had been found guilty of manslaughter and assault. The state had then sought to have him sentenced as a dangerous offender who suffered from a serious personality disorder. Id. at 240. Before the presentencing hearing, the petitioner’s trial counsel did not conduct an extensive investigation of the petitioner’s background. Id. at 238. Nor did counsel consult with an expert. Id. Instead, counsel chose to rely on cross-examination of the state’s expert witness. Id. At the hearing, the state’s expert testified that the petitioner suffered from a serious personality disorder indicating a propensity toward crimes that seriously endanger others. Id. at 242. On cross-examination, the petitioner’s counsel challenged the expert’s foundation for that opinion. Id. at 243-44. The jury ultimately found that the petitioner suffered from a serious personality disorder, and the court imposed a dangerous-offender sentence. Id. at 248. The petitioner sought post-conviction relief, alleging that his trial counsel had provided inadequate assistance 556 McMullin v. Amsberry in failing to conduct an adequate investigation before making the tactical decision not to obtain a defense expert and instead to rely on cross-examination alone. Id. The petitioner introduced evidence from a clinical psychologist that the state expert’s diagnosis was incorrect and based on incomplete information. Id. at 250-51. The state argued that the petitioner’s counsel had made a reasonable strategic decision. Id. at 251. The Supreme Court disagreed. Strategic decisions, the court said, must be based on an adequate investigation of the relevant information: “[A]lthough defense counsel asserted that he had made a calculated strategic decision that this was one of the minority of cases in which it was preferable not to rely on a defense expert, he did so without adequate knowledge of the underlying facts. * * * [A]dequate counsel in this situation would have gained further information about petitioner’s psychological conditions and juvenile history and consulted with a defense expert in the field of psychology to determine how best to counter [the state’s] evidence that petitioner suffered from an antisocial personality disorder.” Id. at 262. Third, in Farmer, the petitioner was convicted of murder with a firearm. 363 Or at 681 . At trial, part of his defense had been that the weapon seized from another suspect’s residence was likely the murder weapon. Id. at 682 . The petitioner’s trial counsel had retained an expert to testify to that effect. Id. at 684 . After hearing the state’s expert testify, though, counsel decided not to call the defense expert, because she worried that the expert’s credentials were not adequate to withstand cross-examination and because she mistakenly believed that the testimony that she had just heard from the state’s witness would not differ significantly from what her own expert would say. Id. at 682-87 . Seeking post-conviction relief, the petitioner argued that he had been denied constitutionally adequate counsel when his trial lawyer decided not to call the defense expert witness. Id. at 687 . The state argued that the trial counsel had made a reasonable tactical choice. Id. at 691 . The Oregon Supreme Court disagreed. A tactical choice, the court explained, “must be grounded on a reasonable Cite as 310 Or App 542 (2021) 557 investigation.” Id. at 690. And such a reasonable investigation necessarily requires acquiring an accurate understanding of the relevant information: “We see little difference between failing to gather information and failing to understand its import. In ensuring that individuals receive the legal assistance that the constitution requires, it would make little sense to demand that attorneys gather sufficient information to make a decision, but not require them to reasonably understand and assess it. After all, the purpose of requiring attorneys to make a reasonable investigation is to enable them to reasonably consider the costs and benefits of pursuing a given action and thus permit them to make an informed decision.” Id. at 697. Applying those principles to the case at hand, the court concluded that the petitioner’s trial counsel could not make the required cost-benefit calculation based on a misunderstanding of the nature of the expert testimony. Id. at 699-700. The consistent thread that runs through the foregoing three cases is this: For purposes of adhering to state constitutional standards of adequate representation, tactical decisions cannot be made in a vacuum. They must be preceded by an investigation that, under the circumstances of the case, enables trial counsel to meaningfully evaluate the costs and potential benefits of further investigation. In this case, Karabeika’s representation of petitioner did not meet that standard. She did briefly discuss the case with two experts, Brady and nurse practitioner Fitzgerald. But her consultation with Brady was cut short by his explanation that he had been retired for quite some time and that he could provide no help. And her consultation with Fitzgerald did not go beyond her determination that the witness was “wacky” and would not be helpful. Thus, in neither case did Karabeika actually acquire information about whether the state’s expert witness correctly understood the literature concerning the incidence of normal physical examinations among adolescent victims of sexual abuse. Karabeika also consulted with two lawyers with experience in trying sexual abuse cases. But once again, 558 McMullin v. Amsberry those consultations did not provide her with the information necessary to enable her to make the required cost-benefit calculation about whether to challenge Smith’s testimony. To the contrary, one of the attorneys, Cohen, merely said something oblique to the effect that such challenges are not done “anymore.” And the other, Maxfield, told her that “[t]here’s plenty of literature to suggest that it would be highly unusual for a girl to suffer repeated ‘blunt force trauma’, especially recently and not have physical evidence that the hymen has been torn.” Maxfield did caution that cross-examination on the basis of that literature “needs to be done surgically, or it can blow up.” But she also said that she and Karabeika should talk about the relevant studies before making a decision whether to go after the state’s expert. Karabeika did not do that. She did not familiarize herself with any of the literature that Maxfield had referred to, much less consult with an expert about whether Smith’s reliance on the Kellogg study was warranted. Instead, she decided that the better tactical choice was to avoid crossexamining Smith on the matter and make an appeal to the jury that it made no sense that a victim of repeated sexual assault would not show physical signs of that abuse. It was a tactical choice, to be sure. But it was not an informed one. Without acquiring an adequate understanding of the relevant scientific issues, Karabeika could not reasonably assess the costs and potential benefits of more directly challenging the state’s expert. That deprived petitioner of the reasonable professional skill and judgment that he is guaranteed by Oregon’s constitution. 2. Prejudice We turn to the question of prejudice. Under Article I, section 11, to establish inadequate assistance of counsel requires that a petitioner establish that trial counsel’s deficient representation “had a tendency to affect the result of the trial.” Burdge v. Palmateer, 338 Or 490, 492 , 112 P3d 320 (2005). To satisfy the “tendency” requirement, a petitioner must show “more than a mere possibility, but less than probability” of such an effect. Green, 357 Or at 322-23 . In a failure-to-investigate case, a petitioner must establish “that there is ‘more than a mere possibility’ that competent Cite as 310 Or App 542 (2021) 559 counsel ‘could have used’ the information that counsel failed to uncover or understand in a way that ‘could have tended to affect’ the outcome of the trial.” Farmer, 363 Or at 700 -01 (quoting Richardson, 362 Or at 266). Whether that showing has been made is a question of law. Stomps v. Persson, 305 Or App 47, 55 , 469 P3d 218 (2020), rev den, 367 Or 496 (2021). In Farmer, the Oregon Supreme Court directly addressed the application of that standard in the context of a failure-to-investigate case. As we have noted, in that case, the petitioner’s criminal trial counsel elected not to call a defense expert to testify that a gun seized from another person was likely the murder weapon. 363 Or at 687 . Trial counsel had made that decision after hearing the state’s expert testify that it could not be determined whether that weapon was the one that killed the victim. Id. at 684 . Counsel mistakenly believed that the defense expert’s testimony would not be appreciably different. Id. at 687 . The Supreme Court concluded that counsel’s mistaken understanding of the defense expert’s opinion rendered counsel’s decision not to call the expert constitutionally deficient. Id. at 699-700 . Turning to the issue of prejudice, the court rejected the state’s contention that counsel’s performance could not have had a tendency to affect the outcome of the case because the evidence against the petitioner was “overwhelming.” Id. at 701 . That, the court explained, is beside the point; it does not mean that omitted evidence would not have tended to affect the outcome of the trial. Id. That evidence still “would have challenged” the state’s expert’s testimony, a fact made all the more significant by the absence of physical evidence, leaving the trial to turn on witness credibility. Id. Similarly, in Richardson, the Oregon Supreme Court concluded that trial counsel’s failure to investigate the petitioner’s background or consult with an expert before crossexamining the state’s expert in a dangerous-offender hearing was prejudicial. 362 Or at 267. The court explained that, had trial counsel consulted with an expert, it was more than a mere possibility that he could have used that information in a way that could have tended to affect the outcome of the dangerous-offender hearing by calling the expert, or better cross-examining the state’s expert, or both. Id. That 560 McMullin v. Amsberry evidence, the court noted, could have been used to show that the state’s expert “had not been thorough in reaching his opinion.” Id. Further, the court said, trial counsel could have called a defense expert to rebut the state’s expert’s diagnosis and provide an explanation for the petitioner’s conduct that was not as damaging as the state’s expert had suggested. Id. at 268. Under the circumstances, the court concluded, it was more than a mere possibility that the jury could have rejected the state’s expert’s diagnosis of a severe personality disorder. Id. More recently, in Running v. Kelly, 306 Or App 589 , 475 P3d 450 (2020), we addressed whether the trial counsel’s failure to call an expert witness in the penalty phase of an aggravated murder trial was prejudicial. At the sentencing hearing, the state had successfully argued that the petitioner’s extensive history of violent acts, especially when in custody, was indicative of his future dangerousness. Id. at 593 . Seeking post-conviction relief, the petitioner offered the testimony of an expert witness, Reidy, who testified that, had he been called during the sentencing hearing, he could have challenged the state’s assertion that the petitioner’s record of violence while in custody was an accurate predictor of future dangerousness. Id. at 599-600 . The state responded that the absence of such expert testimony was not prejudicial because it relied on statistical analysis that would have been subject to challenge on cross-examination. Id. at 604-05 . We concluded that the failure to call Reidy was prejudicial. The court explained that the jury could have credited Reidy’s opinion, which could have undercut the state’s assumption that prison behavior is an accurate predictor of future dangerousness. Id. at 605 . As for the state’s argument that such testimony would have been challenged on cross-examination, we explained: “Although Reidy’s testimony was not ‘conclusive’ regarding petitioner’s future dangerousness, and any expert called as a witness by petitioner’s trial counsel would have been subject to thorough cross-examination about the limitations of statistical analysis in predicting future dangerousness in prison, * * * there was more than mere possibility that the outcome of the penalty phase of petitioner’s criminal trial Cite as 310 Or App 542 (2021) 561 would have been different if his trial counsel presented expert testimony on the subject of future dangerousness.” Id. at 605. In this case, we agree with petitioner that Karabeika’s failure to conduct an adequate investigation caused him to suffer prejudice. Smith’s testimony that a normal physical examination was to be expected of an adolescent girl who had been repeatedly raped was based solely on the Kellogg study. Had Karabeika conducted an adequate investigation, as in Richardson, she could have retained a defense expert or better cross-examined Smith, or both. Cross-examination or the testimony of an expert such as Guertin could have brought to the jury’s attention the flaws in the Kellogg study itself and the fact that the principal author of the Kellogg study actually concluded in another study that “physical findings are really important and they are indicators of sexual trauma, of penetrative sexual trauma.” Cross-examination or testimony from an expert such as Guertin likewise could have brought to the jury’s attention other studies showing that, contrary to the Kellogg study, between 63 and 74 percent of adolescent girls reporting sexual activity showed physical signs. It is more than merely possible that such evidence could have made a difference to the outcome of the case. As in Farmer, the state’s argument here that there was a great deal of evidence of the petitioner’s guilt misses the point. Here, like Farmer, the absence of physical evidence left the case to turn on the reliability and credibility of witnesses. Under those circumstances, it is more than possible that better cross-examination or expert testimony, or both, would have made a difference. Likewise, the state’s argument that an expert like Guertin would have been subject to vigorous crossexamination is unavailing. As we said in Running, defense experts always will be subject to cross-examination, and the fact that a defense expert’s testimony may not conclusively rebut the prosecution’s case is not dispositive. As in Richardson, defense expert testimony may show that the state’s expert “had not been thorough in reaching his opinion.” 362 Or at 267. And the absence of such testimony may be 562 McMullin v. Amsberry prejudicial. Id. In this case, for instance, cross-examination or testimony of an expert such as Guertin could have shown that Smith—who relied on the single Kellogg study—had not been thorough in reaching her own opinion. It is more than possible that such information could have made a difference. As for the state’s argument that Guertin’s own testimony would have been undercut by his concession that most children who reported abuse involving vaginal penetration will have normal genital exams, the state mischaracterizes the expert’s testimony. As we noted above, Guertin agreed that, among children of all ages, a normal genital examination is common. But he also emphasized repeatedly that studies reporting such findings are based on populations of both pre-pubertal and post-pubertal girls. He said that in studies of older adolescents—particularly those older adolescents who have reported penile-vaginal intercourse— there is commonly a high rate of abnormal physical findings. The victim in this case was post-pubertal and reported that petitioner had repeatedly raped her. It is more than merely possible that testimony that a high rate of abnormal physical findings would be expected in such circumstances could have made a difference to the outcome of this case. Reversed and remanded with instructions to grant relief on inadequate investigation claim; otherwise affirmed.