SHORR, J.

Morgen E. Daniels, Deputy Public Defender, argued the cause for appellant. With her on the brief was Ernest G. Lannet, Chief Defender, Criminal Appellate Section, Office of Public Defense Services.

Peenesh Shah, Assistant Attorney General, argued the cause for respondent. With him on the brief were Ellen F. Rosenblum, Attorney General, and Benjamin Gutman, Solicitor General.

Before Armstrong, Presiding Judge, and Tookey, Judge, and Shorr, Judge.

SHORR, J.

Defendant appeals a judgment of conviction, entered after a combined trial with codefendant Mendoza–Lopez, his son, for two counts of first-degree unlawful sexual penetration, ORS 163.411, and three counts of first-degree sexual abuse, ORS 163.427. See State v. Mendoza–Lopez , 291 Or. App. 292, 419 P.3d 761 (2018). He challenges the trial court's exclusion of expert testimony from a child psychologist about memory source confusion, which defendant claims was relevant to his theory of defense. Specifically, defendant contends that the excluded evidence was relevant to demonstrate that the complainants' memories had been altered over time, and that they may have been confused about the identity of their abusers. For the reasons explained below, we conclude that the trial court erred in excluding the expert testimony on memory source confusion, and that the error was not harmless. We therefore reverse and remand defendant's convictions for first-degree unlawful sexual penetration and first-degree sexual abuse.

Defendant and codefendant were each charged with sexually abusing two young children who were their relatives, MS and SM. The complainants did not disclose the abuse until around 10 years later and were legally adults when they were interviewed at Liberty House, an agency that conducts forensic examinations and interviews of children who may have been abused. Before trial, the state moved in limine to disallow the defense from calling as a witness Dr. Bourg, a psychologist who specializes in child forensic interviewing. In support of that motion, the state asserted that it did not intend to offer "child" hearsay by calling any Liberty House staff, interviewers, or doctors in its case-in-chief, because the complaining witnesses were adults at the time of their Liberty House interviews. As a result, the state argued, any expert testimony about child forensic interviewing would be irrelevant to the case.

The court did not make a pretrial ruling on the motion and the case proceeded to a combined jury trial against defendant and codefendant, where the following testimony was elicited. MS and her younger sister SM moved from Mexico to Oregon in 2001, when they were about five and six years old. They lived with their mother and grandfather in a trailer. At some point, their grandfather's brother (defendant) and his son (codefendant) came to live with them in the trailer. Defendants lived with the family for about a year and a half before moving into the house next door, which was occupied by several other men. According to the complainants' mother, the men next door "didn't seem like good people"—they drank alcohol a lot and she often saw sex workers entering the house.

MS testified that, when defendants were living in the trailer with her family, they began to sexually abuse her. She recounted that codefendant abused her routinely when she would arrive home after school but before her grandfather and mother came home from work, and that defendant would sometimes be in the room while this abuse took place. Defendant appeared to be aware of the abuse because he "would look over and he would see [codefendant] and he would laugh." On one occasion, MS remembered her sister being in the room during the abuse and trying to hit codefendant to get him to stop; other times, she was alone with codefendant because defendant would take her sister to the house next door to "separate" the girls. MS also testified that the men who lived in the house next door would take her hand and lead her and SM into the house. There, they would play a "magic" game:

"The game is that they would take off a piece of my hair and they would put it in a paper and they would roll it up and with the paper they would have to try and pull the hair and if they were able to pull the hair, then that was magic. And each time they did that, they would have the right to fondle me or give me a kiss."

MS testified that the men next door never touched her breasts or genitals, but that she remembers them touching her hair and face and giving her kisses. She recalled one occasion inside the trailer when codefendant forced her to kiss her sister, and that the men from next door were there

watching. Both defendants threatened to hurt her family if she disclosed the abuse, so she never discussed it with her mother or grandfather.

SM testified that both defendants sexually abused her when they were living in the trailer with her and her family. She stated that codefendant abused her "many times," and that defendant and her sister were sometimes in the room when it happened. She testified that defendant also sexually abused her on numerous occasions, and that "it gradually got worse." When SM saw codefendant abusing her sister, she would attempt to hit him to make him stop touching MS, but defendant would hold her back or take her to the house next door so that she could not interfere. SM testified that the men living next door were all Mexican and "had more or less the same profile as [defendant]"—"dark hair, black hair with the same kind of body type." The men tried to get her to drink beer and watched as codefendant abused her; on several occasions, one of the other men in the house next door also touched her genitals. Sometimes, the men from next door would come to the trailer and watch while the abuse occurred. SM testified that defendants also took her to a nearby apartment where they abused her in front of other men; on one occasion, she was touched by one of the other men. SM did not tell her mother or grandfather about the abuse she endured because she was scared that defendants were going to hurt her family.

After defendants moved out of the trailer and into the house next door, MS and SM accompanied their mother to Mexico for an extended visit to see their aunt and other family members. At some point, their aunt became suspicious that something bad had happened to the girls in Oregon and she took them to see a psychologist. They eventually disclosed the abuse to their aunt, but asked her not tell their mother specifically what had happened. The complainants' aunt told their mother that something bad had happened in Oregon, but she did not provide any further details. After about a year in Mexico, MS, SM, and their mother moved back to Oregon and got a place of their own.

The complainants did not report the abuse until roughly 10 years later, when MS saw codefendant standing

at a bus stop holding a little girl and was reminded of the abuse she had endured when she was a little girl. After MS told her mother that she had been sexually abused as a child, she was taken to Liberty House. SM accompanied her sister and mother to Liberty House, but did not want to speak to anyone initially. The following year, SM reported that she had also been sexually abused by defendants.

After the state rested its case, the trial court revisited the state's motion to exclude Bourg's expert testimony. The state reiterated its argument that Bourg's expertise in child psychology and interviewing was not relevant because the complaining witnesses were adults by the time they were interviewed at Liberty House and because the state had not presented any testimony from the Liberty House interviewers. Although the state acknowledged the relevance of general testimony about "how memory works," it argued that any expert testimony about memory source confusion or suggestibility would be irrelevant because there had been no evidence presented to support a defense theory that the complainants were confused about the identity of their abusers or that their memories had been tainted.

Defendant called Bourg to the stand for an offer of proof. After discussing her training and experience in the field of child psychology and child interviewing, Bourg gave her opinion about the Liberty House interviews conducted in this case. Among other things, she testified about the concept of memory source confusion, explaining that children under 12 generally lack the ability to identify "how they know what they know. *** [I]f a child learns of something from their direct experience versus if they learn about it from their mother," for example, "they're not going to have the ability to tell you how they gained that knowledge." As a result, it is also possible for young children to confuse the source of their abuse by blaming someone who did not actually perpetrate the abuse:

"BOURG: There's research showing that you can confuse who did something. Um, nothing that would relate specifically to that particular sequence of events being especially relevant, simply that it is possible to confuse who did something to you as well as what they did and as well as

how you know it. So, for example, research has shown that people can confuse whether a doctor or a nurse gave them a shot. People can confuse, research has shown that people can confuse who sexual[ly] abused them. There's research documenting that.

"[DEFENSE COUNSEL]: So, *** hypothetically then, if you knew about somebody in a trailer who was also abusing these young girls, is that a possible source of confusion they could blame somebody else for that abuse? Is that possible or not possible?

"BOURG: It is possible to confuse who did it.

"[DEFENSE COUNSEL]: And why? Why is that possible?

"BOURG: *** [R]esearch studies show that it can happen *** and they show different ways it can happen. ***

"*****

"[THE COURT]: What about this case gives you the opinion that source confusion is relevant or helpful to this jury in this case?

"BOURG: Source confusion about the individual.

"*****

"BOURG: *** [T]here was a lack of corroboration with regard to one of the accused parties with the two children who were present not corroborating one another's stories about [defendant's] involvement, for example. That's where it first arose for me.

"[THE COURT]: Okay. Alright.

"[DEFENSE COUNSEL]: Is *** there any traction to what I'm proposing, at least from a theoretical, scientific standpoint that it's possible these girls are blaming

our clients for something that occurred when actually the perpetrators were the men, the other men in the trailer. [Because] we know, you may not know, but there's been testimony about men in another trailer abusing these girls as well. ***

"BOURG: I think the thread that I can follow on that, would be this—memories are fallible, they fade over time and people can get confused about who did things just as

easily as they can get confused about other factors. There's nothing special about who did something that protects it from the impact of forgetting or source confusion; that you can confuse the source on who, what and how I know it."

The state argued that Bourg's testimony was unnecessary, because the jurors could resolve questions concerning the vagaries of memory using their own common sense. The state argued further that there had been no evidence adduced to support a theory of memory source confusion in this case.

The trial court granted the state's motion in part, ruling that Bourg would be permitted to testify generally about memory and child interviewing and to give her opinion about the quality of the Liberty House interviews conducted in this case, but that she would not be permitted to present evidence on the issue of memory source confusion. The court explained:

"I have heard nothing in this case that would allow any testimony concerning source confusion. Now, my interpretation of source confusion is more focused on the issue of *** whether or not these two defendants or some men in another building did this and that's the way [defense counsel] was asking it. [The prosecutor] argued it a little bit differently from that, but, um, when you went in to your opinion, and this was near the end of your testimony about source confusion, I'm disallowing that. So, but I would allow the critique of the Liberty House interviews, your opinions about that [, and] the areas that you think could have been explored better."

Bourg then testified at trial on behalf of the defense. She stated that, although the Liberty House interviews in this case were generally good, the interviewers "did not ask very many questions about prior conversations or sources of influence on these girls' statements." She testified that such

questions were important because memory fades over time and, "[a]s our memory fades, we become more open to influence from other people."

Bourg also testified that, in her opinion, the inconsistencies between the complainants' accounts and the level of detail in those accounts suggested that their memories may have been influenced by outside sources:

"Because we have two people who were present at the same event and they are reporting different things and it raises the question of how they came *** to be reporting different things. What happened to their memory that allowed them to report different things about the event that they were both present at."

Bourg also expressed concern about the level of detail of the complainants' disclosures roughly 10 years after the alleged abuse. She explained that

"one of the things we know is that memories don't grow over time, they shrink. And so, when you see a memory becoming more and more detailed, it just raises a concern about what are the sources of that additional information that's coming in."

Bourg testified that leading questions from adults such as parents can have an effect on a child's memory. "Children trust adults to define reality for them and if they receive information from a leading or suggestive question from an adult that they trust, they are likely to incorporate that information in to their memory." She also discussed how bias can affect memory: "[Y]ou're told that a person is a bad person and it colors your perception of your current experience with that person where you're more likely to remember negative things about them and it colors your memory of them to where you can redefine prior acts as being negative when they weren't at the time of the original event." Consistent with the trial court's ruling excluding such evidence, Bourg did not offer testimony about the concept of memory source confusion and whether it was possible that the complainants were confused about the identity of their abusers.

During closing arguments, defendant highlighted that the alleged abuse happened a long time ago, arguing

that the complainants' memories could have been altered over time. After discussing numerous inconsistencies in their accounts, he opined that altered memory could account for "why we've gotten these disparate stories, why we've got strange stories, because maybe they didn't really happen." Defendant also discussed the men who lived in the house next to the trailer, suggesting that the complainants might have misidentified their abusers:

"They were *** drinking, they were having prostitutes over there[.] *** These are bad people. What if they did it? And some weird transference of faces, they blame our guys. Could that happen? It's possible. Is it possible the memory could *** be that badly altered? It's possible. Um, is it possible that perhaps they blame the only people they recognize because maybe they, maybe they think they took them over to the apartments, and they blame them for not protecting them. I don't know; that's a possibility."

The jury convicted defendant of two counts of first-degree unlawful sexual penetration and three counts of first-degree sexual abuse. On appeal, defendant challenges the trial court's exclusion of Bourg's proffered testimony on memory source confusion. Defendant argues that the identity of the complainants' abusers was "the central issue in this case." That is, "defendant wanted to present [evidence] that the complaining witnesses had been abused by other people—by one or more of the numerous men who lived next door—not by him or by codefendant, and [that the complainants] had confused memories of the events from a decade before." He argues that Bourg's testimony—which he characterizes as "beyond [a] simple common-sense understanding of how memory works"—was critical to his theory of defense because it would have helped explain to the jury the psychological mechanisms by which memory source confusion occurs, and that "it was reasonably possible that the abuse [in this case] was committed by other individuals."

The state asserts that defendant's claim is unpreserved, meritless, and that any error in excluding a portion of Bourg's testimony was harmless. We begin with the state's preservation argument, which we reject. The state's contention is that, although defendant made an offer of proof on the admissibility of memory source confusion testimony,

he "abandoned his attempt to pursue its admissibility" after the trial court made a preliminary ruling excluding a portion of Bourg's testimony. That preliminary ruling regarded Bourg's opinion that the complainants' aunt may have been a source of influence over them, based on a report by one of the complaining witnesses to Liberty House that her aunt had been "hurt by the same people." The court excluded any expert evidence regarding the aunt after determining that no evidence had been presented that the aunt had, in fact, been abused or hurt by defendants. After that narrow ruling, defense counsel stated:

"I don't know that I disagree with that actually, Your Honor.

"*****

"My, my main focus is I want the jury to have information about memory, how it fades; uh, that's not something they would know if they didn't have an expert talking to them about it and all the other testimony, I think, she gave is relevant to this case."

Contrary to the state's assertion, defendant did not abandon his contention about the relevance of expert testimony concerning memory source confusion or invite the error that he challenges on appeal. Rather, defendant conceded that testimony about the potential influence of the aunt on the complainants' testimony was irrelevant because there had been no evidence presented about the alleged abuse of the aunt at the hands of defendants. That testimony was separate from the proffered testimony about memory source confusion. Indeed, when the trial court issued its ultimate ruling on the state's motion in limine , it issued two separate determinations about the admissibility of Bourg's expert testimony: First, the court ruled that it would "not allow any discussion about the statement by the aunt or involvement of the aunt *** because it *** ha[d] not been part of the trial"; and, second, it ruled that it had "heard nothing in this case that would allow any testimony concerning source confusion[,]" which the court characterized as "whether or not these two defendants or some men in another building did this." Because it was clear to the court that defendant's comments concerning the aunt did not relate to defendant's

arguments about memory source confusion, we conclude that defendant did not abandon his position on the relevance of Bourg's testimony on memory source confusion, and that his current challenge to the exclusion of that testimony is preserved for appellate review.

We turn to the trial court's ruling on admissibility. The court excluded Bourg's testimony about memory source confusion on the basis that it did not sufficiently relate to facts in evidence. We take that to be a ruling that the expert testimony was not helpful under OEC 702, and we review the ruling for legal error. See State v. Jesse , 360 Or. 584, 385 P.3d 1063 (2016) (reviewing for legal error whether trial court correctly excluded expert testimony under OEC 702 on the basis that it did not sufficiently relate to the factual issue for which the defendant had proffered it).

OEC 702 provides that, "[i]f scientific, technical or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training or education may testify thereto in the form of an opinion or otherwise." "To be helpful, the subject of the testimony must be within the expert's field, the witness must be qualified, and the foundation for the opinion must intelligibly relate the testimony to the facts." State v. Brown , 297 Or. 404, 409, 687 P.2d 751 (1984).

Defendant argues that Bourg's testimony on memory source confusion was relevant to his theory of defense that the complainants had "misattributed their abuse at the hands of men who lived in a neighboring residence to defendant and co-defendant." He contends that his theory was supported by evidence in the record—namely, that the men next door had been present during and participated in the abuse, and that they shared at least some physical features with defendant. Defendant also points to inconsistencies in the complainants' testimony as supporting the theory that their memories were altered over time. He argues that the testimony about memory source confusion would have explained that it was reasonably possible that the abuse was committed by other individuals and the "psychological mechanisms by which such confusion occurs."

Relying on Jesse , the state responds that the trial court properly excluded the memory source confusion testimony because, during defendant's offer of proof, Bourg "failed to identify any non-speculative basis for inferring that the victims suffered from source confusion." That is, Bourg failed to identify any specific admissible evidence to support a source confusion theory of defense. Rather, the state argues, Bourg only testified as to a "possibility" that source confusion caused the complainants to mistakenly blame defendants for their abuse, and thus her testimony did not sufficiently relate to the facts of this case.

We agree with defendant that there was enough evidence elicited to support a defense theory that the complainants may have confused defendant with men living next door, and that Bourg's testimony on memory source confusion could have assisted the jury in understanding how their memories may have been affected to allow for such confusion. The evidence showed that there were multiple men living next door who shared some physical features with defendant and who had been present during the abuse, both at the house next door and at the trailer where the complainants and defendants lived. According to the complainants, the men often watched as defendants sexually abused them. Testimony was elicited that those men took the complainants by their hands and led them into the house next door, where they would offer the girls alcohol and play "magic" games that led to the men fondling and kissing them. There was also testimony that at least one of the neighbors had touched SM's genitals.

Bourg's testimony about memory source confusion—in particular, that it is possible for young children who have been sexually abused to confuse the identity of the perpetrator of that abuse—intelligibly related to the facts elicited about the men next door. And her explanation of the psychological phenomenon of memory source confusion may have aided the jury in understanding how a victim of sexual abuse could confuse the identity of her abuser. See State v. O'Key , 321 Or. 285, 298, 899 P.2d 663 (1995) (explaining that OEC 702 is intended to serve multiple functions, "such as: (1) supplying general propositions which will permit

inferences from data which the trier of fact would otherwise be forced to find meaningless; (2) applying general propositions to data so as to generate inferences where the complexity of the body of propositions applied, the difficulty of the application, or other factors make the expert's conclusion probably more accurate or precise than that of the trier of fact; (3) modifying, qualifying, and refining general propositions which the trier of fact may reasonably be expected to use; and (4) adding specialized confirmation and, thus, confidence to general propositions otherwise likely to be assumed more tentatively by the trier" (internal quotation marks omitted) ).

The state's reliance on the Supreme Court's decision in Jesse is misplaced. That case involved the exclusion of expert testimony from a psychologist about the defendant's adjustment disorder. 360 Or. 584, 385 P.3d 1063. The defendant sought to introduce the testimony to support his theory of defense, which was that his adjustment disorder caused him to make a false confession to the police. During an offer of proof, the expert testified that the defendant's adjustment disorder caused him to ruminate and "become obsessive," and that he lacked adequate coping skills and self-esteem. Id . at 588, 385 P.3d 1063. She testified further that,

"[i]n regards to [the defendant's] confession, I would factor my knowledge about him, about his adjustment, you know his high degree of distress and so that—that would be where with my medical certainty would be he was very distressed. And whether he's confessing to something out of distress, that's within the realm of possibility."

Id . at 589, 385 P.3d 1063.

The state moved to exclude the expert testimony on the basis that it was unhelpful to the jury because there had been no evidence linking the formal diagnosis to " ‘the material events that occurred in this case.’ " Id . at 587, 385 P.3d 1063. The Supreme Court agreed that the evidence was not helpful under OEC 702. The court explained that "[s]ome expert testimony, like other forms of evidence, only conditionally relates to a fact in issue, meaning that it is relevant only if another fact—the conditional fact—is first proved." Id . at 597, 385 P.3d 1063. The court then observed that the proffered expert testimony

about the defendant's adjustment disorder was only relevant if there was evidence to demonstrate that the adjustment disorder actually influenced him to falsely confess. However, the defendant failed to present any evidence to support that conditional fact. Id . at 601, 385 P.3d 1063. The court noted that, while the expert testified that "it was within the realm of possibility" that the defendant's adjustment disorder influenced his confession, she did not testify that the "defendant's disorder can produce confessions that are not genuine, nor did she provide any indicia that the jury could use to draw a reasonable inference that his admissions were not actual confessions." Id . Furthermore, the defendant "did not [otherwise] seek to prove the conditional fact" by, for instance, testifying that "he admitted the facts he did because he was obsessively ruminating or was unable to cope." Id .

This case differs from Jesse in two respects. First, whereas the expert in Jesse failed to provide testimony that linked a diagnosis of adjustment disorder to false confessions, Bourg did offer testimony linking memory source confusion to defendant's theory that the complainants had misidentified their abusers. She explained that memory is fallible, and that a possible effect of altered memory is that a person can confuse who did something to them. For instance, she explained how a victim of sexual abuse could confuse the identity of the perpetrator of that abuse. Bourg's testimony could have been useful in aiding the jury in understanding an unfamiliar concept or in confirming a general proposition likely to be assumed more tentatively by the jury, O'Key , 321 Or. at 298, 899 P.2d 663, and would have permitted an inference in this case that the complainants had misidentified their abusers.

Second, and significantly, while there was no evidence offered in Jesse to support the defense theory that the defendant had actually confessed falsely due to his adjustment disorder, there was evidence presented in this case to support the defense theory that the complainants had been abused by the men next door and were confused as to the identity of their abusers. Because there was evidence to support that defense theory, Bourg's testimony about memory source confusion would have assisted the jury in

understanding how such confusion is possible. For instance, when the trial court asked Bourg what made source confusion relevant in this case, Bourg noted for the court that there were discrepancies in the complainants' accounts regarding defendant's involvement in the abuse. Based, in part, on those discrepancies, Bourg explained how memory source confusion was relevant to the case and could be helpful to the jury in understanding the evidence. See State v. Gherasim , 329 Or. 188, 985 P.2d 1267 (1999) (expert testimony that the victim suffered dissociative amnesia was helpful to the jury where the disputed issue at trial was whether the defendant or someone else assaulted the victim, and the victim's statements to the police and testimony at trial contained discrepancies).

Having decided that the trial court erred in excluding Bourg's testimony on memory source confusion, we must now decide whether that error requires reversal of defendant's convictions. Or. Const., Art. VII (Amended), § 3. An error is harmless if there is "little likelihood that the error affected the verdict." State v. Davis , 336 Or. 19, 32, 77 P.3d 1111 (2003). In making that determination, we review all pertinent portions of the record. State v. Maiden , 222 Or. App. 9, 11, 191 P.3d 803 (2008), rev. den. , 345 Or. 618, 201 P.3d 909 (2009). "With respect to the erroneous exclusion of impeachment evidence, a trial court's error is harmless if either: ‘(1) despite the exclusion, the [factfinder] nonetheless had an adequate opportunity to assess [the witness's] credibility; or (2) [the witness's] credibility was not important to the outcome of the trial.’ " State v. Jones , 274 Or. App. 723, 728, 362 P.3d 899 (2015) (quoting State v. Titus , 328 Or. 475, 482, 982 P.2d 1133 (1999) (brackets in Jones ) ).

There can be no question that MS's and SM's credibility were central to the outcome of this case. The question remains whether, despite the erroneous exclusion of Bourg's testimony about memory source confusion, the jury nevertheless had an adequate opportunity to assess their credibility. The state argues that Bourg's trial testimony about memory in general provided ample evidence from which the jury could assess whether the witnesses' memories were "clinically unreliable."

We recognize that, at trial, Bourg was permitted to testify that memories are fallible and that, over time, memories are more susceptible to influence from other people. She identified several factors that can suggest that a person's memory has been altered over time, including the consistency and level of detail of recalled events. She also talked about the effect of outside influences on a child, explaining that leading or suggestive questions from a trusted adult could affect a child's memory. With respect to MS and SM specifically, Bourg testified that the inconsistencies in their testimony raised a concern over the reliability of their accounts:

"Because we have two people who were present at the same event and they are reporting different things and it raises the question of how they came *** to be reporting different things. What happened to their memory that allowed them to report different things about the event that they were both present at."

While Bourg's trial testimony provided the jury with information about memory in general and how it can fade over time and be altered by outside influences, it did not supply the jury with an understanding of the particular concept of memory source confusion. Such testimony would have been of a different nature than the testimony that Bourg was permitted to give, and would have provided stronger and more specific support for the defense theory that the complainants' memories had been altered over time, which had caused them to misidentify their abusers. See Titus , 328 Or. at 482, 982 P.2d 1133 (error in excluding bias evidence was not harmless where the excluded evidence was qualitatively different from other bias evidence that had been admitted, and thus the jury did not have an adequate opportunity to assess the witness's credibility); State v. Crum , 287 Or. App. 541, 556, 403 P.3d 405 (2017) (error in excluding bias evidence not harmless because excluded evidence was "different, stronger, more specific, and less subject to challenge" than other evidence admitted, thereby depriving the jury of adequate opportunity to assess the credibility of the state's witnesses).

The trial court's error in excluding Bourg's testimony on memory source confusion is particularly significant

in light of the closing arguments made by defendant and the prosecution. Defendant attempted to advance his theory of misidentification during closing argument by asking the jury to consider the possibility that, through "some weird transference of faces," the complainants blamed defendant for the actions of the men who lived next door: "[I]s it possible that perhaps they blame the only people they recognize because maybe they, maybe they think they took them over to the apartments, and they blame them for not protecting them[?]" However, he was not able to draw much support for that hypothesis from his own expert's testimony. In response, the state argued that there was no evidence that the complainants' "memories actually involve sexual abuse from someone else," observing that both complainants testified that they were certain that the abuse was perpetrated by defendants. Had Bourg been permitted to explain the concept of memory source confusion and how and why a child might confuse the identity of her abuser, the jury might have had a better understanding of defendant's argument that the complainants were confused about who had abused them 10 years earlier, and may have viewed the complainants' certainty in the identity of their abusers differently.

Absent Bourg's testimony on memory source confusion, the jury did not have an adequate opportunity to fully assess the complainants' credibility on a significant trial issue—namely, whether defendants or neighbors had committed the abuse. Thus, the court's exclusion of that testimony was not harmless.

Convictions for first-degree unlawful sexual penetration and first-degree sexual abuse reversed and remanded; otherwise affirmed.