*219**286In this criminal case arising out of allegations of child sexual abuse, the issue is whether the expert testimony that the trial court allowed about "grooming" children for later sexual activity is "scientific" evidence that requires a foundational showing of scientific validity under OEC 702. At trial, over defendant's objection, the trial court permitted a forensic interviewer to testify about defendant's behavior that may have constituted "grooming" of the victim for sexual abuse if defendant had the requisite intent, without the state first establishing that the testimony about grooming was scientifically valid and reliable.
Defendant was convicted of first-degree sexual abuse and attempted first-degree sodomy. On defendant's appeal, the Court of Appeals held that the testimony was not scientific evidence for which a foundation was required. State v. Henley , 281 Or. App. 825, 386 P.3d 126 (2016). For the reasons that follow, we conclude that the testimony was scientific evidence and that the trial court erred in admitting it without a proper foundation. Given the record, we decline to decide the validity and reliability of the expert testimony on review. We also conclude that the admission of the testimony was not harmless. Therefore, we reverse the decision of the Court of Appeals and the judgment of the trial court and remand to the trial court for further proceedings.
I. FACTS
A. The Alleged Abuse
Defendant's trial centered on whether he had sexually abused his stepdaughter, M, during a family camping trip. As to that specific event, the state presented the following evidence.
At the time of the alleged abuse, defendant had been married to M's mother for nine years. M, who was then 11 years old, lived with her mother and defendant in Idaho. One summer weekend, defendant, M, her mother, two of M's siblings, and an adult male friend of the family went camping in eastern Oregon. They all slept in a camper-a pop-up tent trailer with fold-out beds on either end of a middle living space. Defendant and M's mother slept in one fold-out **287bed and M's siblings slept in the other. M slept on a mattress in the middle area, next to her mother and defendant. The family friend slept on another mattress in the middle area.
Defendant and the friend stayed up late and went into the camper after the others were already in bed. M's mother woke up and tried to persuade defendant to come into bed with her. Instead, defendant sat down next to M on M's bed, and M's mother went back to sleep. M awoke briefly while defendant was lying on her bed, and then she fell back to sleep.
M was awakened again early the next morning, when defendant, who was lying beside her, pulled down his own pants and pulled M's sweatpants and underwear to her ankles. He inserted his fingers into her vagina. In an effort to stop him from touching her, M rolled over onto her stomach and then onto her side. Defendant put his hands on M's sides, attempted to spread her buttocks with his thumbs, and put his penis in her "butt crack." Defendant rubbed against her, ejaculated, and said, "Ahh."
M then sat up. Defendant also sat up and asked her if she was okay. M's mother then woke up, and defendant lay back down in M's bed. M's mother asked M if anything was wrong, but M answered no. M's mother asked M to come up into her bed, and M complied and fell asleep. When M woke up, defendant was not in the camper.
Later that day, M told her mother what had happened. M's mother replied that she did not know what to say, but that she would arrange a mattress to make a barrier to prevent defendant from getting into bed with her again that night. Sometime later, M's mother told defendant that M had said he had touched her and rubbed his penis against *220her buttocks. Defendant responded that he did not know what had happened, as he had been asleep.
B. The Investigation and Charges
The police became involved later that month. M's father and his fiancée also lived in Idaho. When the camping trip was over, M, as previously planned, went to stay with her father and his fiancée for a month. M did not immediately tell her father what had happened on the camping **288trip. M eventually told her father's fiancée about massages that defendant had given her, and she relayed that information to M's father. M's father asked M to tell him if any adult had ever touched her inappropriately. Later that day, M told her father and his fiancée about the abuse in the camper. M's father called the local police in his community, where the investigation began.
Courtney Palfreyman, a forensic interviewer for Children at Risk Evaluation Services (CARES) at St. Luke's Hospital in Boise, Idaho, interviewed M. Palfreyman's interview of M was video-recorded. During the interview, M told Palfreyman about the recent camping incident. In addition, she described massages that defendant had given her that made her uncomfortable. M also said that, when she was five or six years old, defendant had crawled into bed with her one night and asked her to touch his penis. M explained that she did not comply and instead told defendant that she needed to go to the bathroom and went to find her mother.
Because the camping incident occurred in Oregon, the case was transferred to Malheur County, Oregon. Defendant was charged with one count of first-degree sexual abuse and one count of first-degree sodomy based on his conduct in the camper.
C. Trial
Citing OEC 403 and OEC 404(3), defendant filed a motion in limine seeking to exclude evidence from his impending jury trial, including testimony that defendant had massaged M inappropriately. The state responded that evidence of inappropriate massages was admissible to demonstrate defendant's "grooming behavior as part of [his] planning or preparation for the later sexual assault" of M. The trial court denied defendant's motion.
At trial, M testified in conformance with the recorded CARES interview. M testified regarding the camping incident, the incident from years earlier,1 and the massages **289defendant had given her. The massages were the subject of the expert's grooming testimony at issue. During M's direct examination, she testified that defendant had been giving her massages from time to time, at her request. M explained that defendant had massaged her shoulders but also "down [her] legs and up by [her] chest." M did not like when defendant massaged her chest, though, because she "thought he was going too far into [her] other areas." M also testified that she told her mother that defendant was massaging "too close into other areas [she] didn't like."
Palfreyman also testified for the state. She told the jury that she has a bachelor's and a master's degree in social work and received specialized training in forensic interviewing, both basic and advanced, with the American Professional Society on the Abuse of Children. She stated that she had over 10 years of experience working in child welfare and protection and as a forensic interviewer. She had worked for the State of Idaho from 1999 and then had been a forensic interviewer with CARES since 2005. At the time of the trial in 2009, she had completed over 600 forensic interviews.
During Palfreyman's testimony, the prosecutor played for the jury the video recording of her interview of M at CARES, pausing it at various points to ask Palfreyman clarifying questions. In the recorded interview, M said *221that she had told her father's fiancée about several massages from defendant and described the massages to Palfreyman:
"When I'm sore, he'll massage me, but he won't get where I need him to. He'll go all over the place [gesturing and putting her hands palms down in the middle of her chest]. And, like last month, I asked him to massage my neck 'cause I had a neck ache. I'm never gonna do that again because of what happened, because I'm so scared of him-with the camping. But, I was asking him to massage my neck. He was massaging it, and he was going down here [putting her hands in the middle of her chest], and then I asked him to **290stop going down here [putting her hands to her chest]. And he was all, okay; then he went down here [gesturing toward her lower back]."
In the recording, Palfreyman confirmed that M was pointing to her lower back, and M described that defendant had also massaged her on the back of her upper legs that time.
After the conclusion of the video, the prosecutor asked Palfreyman the following question about her training on grooming: "In terms of this particular interview or in general, have you had any training regarding a concept called grooming?" Palfreyman answered in the affirmative. Defendant objected to Palfreyman's qualification to testify as an expert on grooming behavior. In response, the prosecutor requested an opportunity to qualify Palfreyman as an expert and asked her several other questions about her training on grooming:
"[PROSECUTOR]: So you've actually had training in the area of grooming is that correct?
"A: Yes.
"[PROSECUTOR]: Okay. But you're-you're not a psychologist or anything like that?
"A: No.
"[PROSECUTOR]: Okay. What sort of training have you had?
"A: Um, just due to my forensic interview training we talk about grooming (INAUDIBLE) leading up to offending, and through my college courses, different trainings that I've had."
When the prosecutor proceeded to ask Palfreyman to describe what she had been taught to look for in terms of grooming behavior, defendant again objected, on the ground that the prosecutor had not laid a proper foundation for that testimony-that is, that Palfreyman was not an expert and had had no special training with regard to grooming. The prosecutor responded that he was asking Palfreyman about her "training and experience" with regard to grooming. At that point, defense counsel suggested that the matter should be argued without the jury present.
**291Outside the presence of the jury, the court asked the prosecutor whether he would go beyond asking Palfreyman to define grooming and her training, by asking whether she saw signs of grooming. The prosecutor told the court that Palfreyman had "been taught to recognize certain behaviors which could be considered grooming" and argued that, although he would ask if she saw any signs of grooming in this case, he was not asking for a scientific, medical, or expert opinion. Defendant argued that testimony about grooming was similar to testimony about syndrome evidence or a diagnosis of sexual abuse and that, like evidence on those topics, "there has to be some sort of scientific validity" to it, which had not been shown in this case.
The trial court overruled the objection. It permitted Palfreyman to define what grooming behavior is and to describe behaviors by defendant that concerned her. The court also granted defendant a continuing objection to evidence concerning sexual grooming.
Once the jury was recalled to the courtroom, the prosecutor continued to question Palfreyman about grooming behavior in general:
"[PROSECUTOR]: So in terms of grooming can you just briefly describe some of the activities that you're familiar with that might be considered grooming?
"A: Some of the activities would include spending time together, sometimes allowing the child to do things the parents *222wouldn't allow like video games when it's not allowed or alcohol use. It would include things like giving them money for things, tickling, massaging, that sort of thing.
"[PROSECUTOR]: So the grooming's done by the offender to sort of lay the groundwork for later abuse?
"A: Yeah, to build trust and weaken defense[s] of the child. "
(Emphasis added.)
The prosecutor then moved to a question about defendant's behavior, asking whether Palfreyman had seen "anything in this case which you might consider or might be considered grooming?" Defense counsel objected, and the court directed the prosecutor to rephrase the question. The prosecutor resumed:
**292"[PROSECUTOR]: In this particular interview was there any behavior which could be considered grooming?
"A: Um, when she just talked about the massaging where she wanted it on her neck but he would go lower into her chest area."
The prosecutor ended his examination there.
On cross-examination, Palfreyman acknowledged that an adult who spent time with a child was not necessarily grooming the child for sexual abuse and that she had not spoken with defendant about his intentions. Defendant elicited the following testimony:
"[DEFENSE COUNSEL]: Now you mentioned spending time with-spending time together.
"A: Mmm hmm (INDICATING YES).
"[DEFENSE COUNSEL]: Surely you're not saying that anytime an adult spends time with a child that they're grooming them for sexual abuse.
"A: No, not every time.
"[DEFENSE COUNSEL]: So I'm reading to my child at night, in bed, I'm grooming them for sexual abuse in the future? That not true is it?
"A: No.
"[DEFENSE COUNSEL]: Okay. If I decide to take my child out for an ice cream cone I'm grooming them to abuse them sexually later?
"A: It depends on your motives. I mean if, as a parent, um, that's not what I'm saying. I'm saying as an outsider looking to build trust and weaken defenses; if that's your motive for getting into this child's circle of trust, then that could be potential grooming.
"[DEFENSE COUNSEL]: Okay, so if you do anything as a parent with an evil intent that is what you call grooming?
"A: Potential.
"[DEFENSE COUNSEL]: Okay, but you don't know anybody's intents. You didn't talk to [defendant] did you?
"A: No."
**293In addition to M and Palfreyman, among other witnesses, the state called M's father to testify. He explained that, while he was at the police station talking with the police, defendant called him on his cell phone and left a voicemail message, which the police listened to and then recorded. The recording was received in evidence and played in the courtroom. In it, defendant told M's father:
"I really wish you would talk to me about it. Because if I did do something, I didn't mean anything-I don't even know what I did. I know for a fact that I was not naked in bed. I, I wasn't. Please call me and talk to me about this because it-yeah, this is really, really messing with me, man, and I know it's messing with you too. Thanks, bye."
In closing argument, the prosecutor highlighted Palfreyman's testimony about grooming behavior. The prosecutor argued that defendant's massages of M constituted "classic grooming behavior" that makes it easier for the offender to commit a future "offense":
"The massaging is important because it's grooming behavior. Courtney Palfreyman talked to you about offenders trying to get children accustom[ed] to touching. The reason they do that, folks, is to get them to a point where when later touching comes about they don't get the kind of reaction where the child walks out of the room, after he asks to touch his thing, and they go tell mom. The child gets accustomed to the behavior, the barriers get broken *223down, and it makes it easier for an offense in the future. It's classic grooming behavior."
In his closing argument, defense counsel argued that the state's grooming argument should raise concerns with the jury, because "common sense would tell you that just about anything that the parent does for or with the child can be construed as grooming." He provided examples: "If you take your child out for an ice cream cone by themselves are you grooming them for something later? If you decide you're going to go play catch with them, are you grooming them to let their sexual guard down? *** If you tickle *** your child are you grooming them so that they can *** not be sensitive sexually to you later?" But in rebuttal argument, the prosecutor reprised his argument that defendant's massages of M mattered because defendant "was grooming [M]."
**294The jury convicted defendant of first-degree sexual abuse. And although it acquitted him on the charge of first-degree sodomy, the jury convicted defendant of the lesser included offense of attempted first-degree sodomy.
D. Appeal
Defendant appealed his convictions to the Court of Appeals, arguing, among other things, that the trial court erred in allowing Palfreyman's testimony about grooming techniques used on children. Defendant asserted that the testimony was scientific evidence that required a foundational showing of scientific validity, which the state had failed to establish. Defendant contended that the testimony concerned scientific knowledge because (1) the concept of grooming was grounded in behavioral science and (2) a jury, informed of Palfreyman's experience and her training on grooming, would assume that grooming was well-established in the social sciences.
As noted, the Court of Appeals rejected defendant's argument and affirmed defendant's convictions. The court explained that Palfreyman's testimony "did not purport to draw its convincing force from principles of science," Henley , 281 Or. App. at 833, 386 P.3d 126, which the court viewed as the key factor in determining whether evidence is "scientific," id. at 832, 386 P.3d 126. The court concluded that Palfreyman had specialized knowledge about grooming, a subject area outside "the average person's common knowledge," id. at 831, 386 P.3d 126, and that her expert testimony was admissible on that basis.
II. ANALYSIS
A. "Scientific" Evidence Under OEC 702
We accepted review to address whether the evidence presented concerning the grooming of children for sexual abuse is "scientific" evidence under OEC 702. As relevant, OEC 702 provides that, if "scientific" 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."
**295And as this court explained in State v. O'Key , 321 Or. 285, 293, 899 P.2d 663 (1995), scientific knowledge cannot assist the trier of fact if it is not "scientifically valid." Thus, if the evidence was "scientific," the state was required to comply with the standards for admission of scientific evidence set out in O'Key and in State v. Brown , 297 Or. 404, 687 P.2d 751 (1984).
The standard by which courts determine whether evidence is properly considered as "scientific" is not found in the Oregon Evidence Code. This court has observed that the term "scientific" is not defined in the Oregon Evidence Code or its commentary, and, instead, the legislature left the standard to be determined by judicial decision. O'Key , 321 Or. at 290, 899 P.2d 663. Although this court has not "precisely defined what makes evidence 'scientific,' " State v. Marrington , 335 Or. 555, 561, 73 P.3d 911 (2003), through Brown and O'Key , the court has identified some markers for determining whether evidence is "scientific" and therefore requires a foundational showing of scientific validity before it may be admitted.
*224In Brown , which involved whether polygraph test results were admissible in the defendant's rape trial, this court stated that evidence is "scientific" under OEC 702 if it "draws its convincing force from some principle of science, mathematics and the like." Brown , 297 Or. at 407, 687 P.2d 751. The court observed that, "[t]ypically, but not necessarily, scientific evidence is presented by an expert witness who can explain data or test results and, if necessary, explain the scientific principles which are said to give the evidence its reliability or accuracy." Id. at 407-08, 687 P.2d 751.
The court in Brown appears to have addressed how polygraph test results draw their convincing force from a principle of science by describing the premises of the polygraph technique: " 'an individual's conscious attempt to deceive engenders various involuntary physiological changes due to an acute reaction in the sympathetic parts of the autonomic nervous system' " and the " 'polygraph machine is an electromechanical instrument which measures and records these physiological fluctuations.' " Id. at 419, 687 P.2d 751 (quoting with approval United States v. Alexander , 526 F.2d 161, 163 (8th Cir. 1975) ). That the Brown court implicitly concluded that **296the polygraph evidence was scientific is apparent in that it proceeded to develop and to apply a multifactor test-factors that are not exclusive-for determining whether scientific evidence is admissible based on its probative value and reliability. 297 Or. at 417-18, 687 P.2d 751. In developing that test, the court declined to adopt the test for admissibility of scientific evidence derived from Frye v. United States , 293 F. 1013, 1014 (D.C. Cir. 1923) (holding that scientific evidence is admissible only if the scientific principle is "sufficiently established to have gained general acceptance in the particular field in which it belongs"). 297 Or. at 408, 687 P.2d 751.
In O'Key , this court expounded on the defining characteristics of "scientific" evidence first stated in Brown . In O'Key , which concerned the horizontal gaze nystagmus (HGN) test used by police officers in the field to determine whether someone is driving under the influence of alcohol, the court looked at the term both in a technical way and in a practical way.
For its technical approach to understanding what "scientific" evidence is, this court found the United States Supreme Court's discussion of the term "scientific" as used in FRE 702 helpful in construing that term as it is used in OEC 702. This court observed that, in Daubert v. Merrell Dow Pharmaceuticals , 509 U.S. 579, 590, 113 S.Ct. 2786, 125 L.Ed. 2d 469 (1993), the Supreme Court had explained that "scientific" as an adjective "implies a grounding in the methods and procedures of science" and that "scientific knowledge" is " 'an inference or assertion [that] must be derived from the scientific method.' " O'Key , 321 Or. at 292, 899 P.2d 663. However, the court explained that it was not attempting "precisely to distinguish 'scientific' from other types of expert testimony under the Oregon Evidence Code." Id. at 293, 899 P.2d 663.
And from a practical standpoint, this court recognized that much expert testimony " 'rests at least partly on science,' " id. at 291, 899 P.2d 663 (quoting Christopher B. Mueller & Laird C. Kirkpatrick, Modern Evidence § 7.8, 990 (1995) ), and stated that evidence is "scientific" if it is likely to be perceived by jurors as grounded in science:
"Evidence perceived by lay jurors to be scientific in nature possesses an unusually high degree of persuasive **297power. The function of the court is to ensure that the persuasive appeal is legitimate. The value of proffered expert scientific testimony critically depends on the scientific validity of the general propositions utilized by the expert. Propositions that a court finds possess significantly increased potential to influence the trier of fact as scientific assertions, therefore, should be supported by the appropriate scientific validation. This approach ensure[s] that expert testimony does not enjoy the persuasive appeal of science without subjecting its propositions to the verification processes of science."
O'Key , 321 Or. at 291-92, 899 P.2d 663 (footnote and quotation and citations omitted). Accord Jennings v. Baxter Healthcare Corp. , 331 Or. 285, 304, 14 P.3d 596 (2000) ("A jury is likely to believe that a doctor's testimony *225about [his opinion that silicone caused the plaintiff's neurological condition] is a scientific assertion and, therefore, the proponent of the testimony must show that it is scientifically valid.").
This court's ultimate conclusion in O'Key rested primarily on the technical understanding of "scientific" knowledge. During the HGN test, which is part of a battery of field sobriety tests for determining whether a person is under the influence of alcohol, a police officer looks for nystagmus, an involuntary rapid or jerky movement of the eyeball as the eye tracks a steadily moving object, such as a finger, pencil, or pen light. 321 Or. at 287, 899 P.2d 663. The court concluded that evidence of the HGN test result was scientific because the value of the result depended on the scientific validity of the causal relationship between the result and alcohol consumption inherent in the test:
"The HGN test provides evidence that purports to draw its convincing force from a principle of science, namely, the asserted scientific proposition that there is a causal relationship between consumption of alcohol and the type of nystagmus measured by the HGN test. The value of HGN testing depends critically on the demonstrated scientific validity of that proposition."
Id . at 296, 899 P.2d 663 (citation omitted).
But this court also considered, at least implicitly, how jurors would perceive evidence concerning the result of the HGN test. The court explained that the HGN test **298evidence "rests on a manifestation of alcohol consumption not easily recognized or understood by most people"-namely, a "relationship between the effects of alcohol on the central nervous system, the nystagmus phenomenon, and the HGN test" that is "not within the realm of common knowledge of the average person." Id . at 297, 899 P.2d 663. The court contrasted the HGN test with other field sobriety tests, which it concluded were not "scientific." The one-leg stand test, the walk-and-turn test, and the finger-to-nose test, for example, are not scientific evidence, because they "obtain their legitimacy from effects of intoxication based on propositions of common knowledge" and not from scientific principles. Id .
Since O'Key , this court has continued to consider whether lay people would consider evidence to rest on science-and therefore accord it significant persuasive value-in determining whether expert testimony should be viewed as scientific evidence for purposes of OEC 702. This court's decision in Marrington illustrates the significant role that the jury's perception plays when the proponent of the expert testimony features the expert's familiarity with-and explicit reliance on-scientific research and literature to provide information to the jury.
In Marrington , the state presented the testimony of an expert, a counselor working for a nonprofit organization specializing in the treatment of sexually abused children, who stated that delay in reporting is a predominant feature in a child's disclosure of sexual abuse. 335 Or. at 563, 73 P.3d 911. The expert explained that there was a "body of literature" and a significant study conducted by a psychiatrist affiliated with UCLA supporting that assertion. Id. at 559, 73 P.3d 911.
As to the question of whether the delayed-reporting testimony was scientific evidence and therefore required an appropriate scientific foundation, the court stated that whether the evidence is scientific "depends primarily on whether the trier of fact will perceive the evidence as such," 335 Or. at 561, 73 P.3d 911, and that a court's inquiry should focus on whether the expert's assertions " 'possess significantly increased potential to influence the trier of fact as scientific assertions.' " 335 Or. at 562, 73 P.3d 911 (quoting O'Key , 321 Or. at 292, 899 P.2d 663 ). In that regard, this court observed that the witness had stated **299that she had degrees in psychology, that she was a certified counselor, that she estimated that she had spoken to over 200 children who complained about having been sexually abused, and that she had testified that she was familiar with recent literature and research in the area of sexual abuse of children. Id. at 562-63, 73 P.3d 911. Given that context, this court concluded that "[t]he reference to 'research' implies a scientific foundation with the results published in the 'literature' of the area of study." Id . at 563, 73 P.3d 911. The court also noted that, in the *226witness's testimony, she referred to "the characteristics of sexually abused children," which, the court explained, implied that there was "a well-defined, empirically verified set of characteristics that a significant percentage of sexually abused children display." Id . Additionally, the court stated, the witness's use of terms such as "study," "body of literature," and "psychiatrist" in her testimony invoked "the vocabulary of scientific research." Id . Finally, the witness testified that the theory that she had espoused had been confirmed empirically, insofar as researchers had found that delayed reporting was prevalent in cases in which sexual abuse had been substantiated by confession of the perpetrator or in some other way. Id.
The court concluded that an expert "who has a background in behavioral sciences and who claims that her knowledge is based on studies, research, and the literature in the field, announces to the factfinder that the basis of her testimony is 'scientific,' i.e. , is grounded on conclusions that have been reached through application of a scientific method to collected data." Marrington , 335 Or. at 563-64, 73 P.3d 911. And "[b]ecause that is how the factfinder would understand it, a court has a duty to ensure that such information possesses the necessary indices of scientific validity" by requiring the proponent to "show that that asserted rule of behavior was scientifically valid under the standards established in Brown and O'Key ." Id . at 564, 73 P.3d 911.
It appears that courts in other jurisdictions have largely avoided answering the question whether testimony of an expert about grooming behaviors is scientific.2 Although the parties have cited cases that involve grooming testimony from other jurisdictions, many of those cases involve other issues and are not on point. E.g. , State v. Akins , 298 Kan. 592, 315 P.3d 868 (2014) (deciding whether the prosecutor had improperly asserted personal knowledge of the evidence and argued facts not in evidence when she referred to defendant's behaviors as grooming, though she did not present any evidence-expert or otherwise-that the described behaviors were grooming); People v. Ackerman , 257 Mich. App. 434, 441, 669 N.W.2d 818 (2003), rev. den. , 469 Mich. 1012, 677 N.W.2d 24 (2004) (concerning admissibility of "other acts" evidence, specifically, lay witness testimony describing the defendant's grooming-type behaviors). We now proceed to apply Oregon law to the circumstances of this case.
B. Application
We recognize that, in this case, Palfreyman did not purport to establish that sexual grooming has been studied by social scientists or that her assertions about grooming have been scientifically verified phenomena. Instead, in contrast to Marrington , this case involves expert testimony as to which the proponent disclaimed a scientific connection.
In the trial court, the prosecutor argued that Palfreyman had been trained to recognize certain behaviors as grooming but disclaimed that he was asking her for a scientific opinion. Distinguishing Marrington , the state argued to the Court of Appeals that Palfreyman did not couch her opinion in the "vocabulary of science" and the jury would not have viewed Palfreyman as presenting a *227scientific opinion. Therefore, the state urged, her testimony did not need the support of a scientific foundation. And on review, the state again contends that nothing in Palfreyman's testimony about the nature of grooming would have suggested to the jury that "grooming" was a scientific concept and that no aspects of Palfreyman's testimony would have led the jury to believe that she was making a scientific assertion. The state points out that Palfreyman is not a psychologist and did not testify about the existence of research, studies, or literature on the topic of sexual grooming. Thus, the state argues, unlike the expert in Marrington , Palfreyman gave limited testimony that did not make any "scientific assertion" to the jury. Rather, according to the state, Palfreyman's testimony was merely descriptive of some activities that she herself was familiar with in her experience as a forensic interviewer.
But the fact that the proponent of expert evidence at trial disclaims that the evidence is scientifically grounded does not obviate the possibility that it nevertheless constitutes "scientific" evidence under OEC 702. Expert evidence is "scientific" under OEC 702 when it is expressly presented to the jury as scientifically grounded, as in Marrington . Expert evidence also is "scientific" under OEC 702 when it "draws its convincing force from some principle of science," as in Brown , 297 Or. at 407, 687 P.2d 751, or "implies a grounding in the methods and procedures of science" and would likely be perceived by the jury as imbued with the "persuasive appeal of science," O'Key , 321 Or. at 292, 899 P.2d 663. In this case, we conclude that Palfreyman's testimony about grooming, in the context of her testimony overall, was "scientific" evidence, because, as in O'Key , the evidence implied that it was grounded in science and the jury likely would have viewed the evidence that way.
First, Palfreyman's testimony implied a grounding in behavioral science. Palfreyman was presented as an expert in child sexual abuse. She had a bachelor's degree and a master's degree in social work, with over 10 years of experience in child welfare and protection and forensic interviewing with the State of Idaho and CARES. She also told the jury that she had specialized training in forensic interviewing and had training on the subject of grooming behavior by child sexual abuse offenders, both through her college coursework and in her forensic interview training. Given that context, her testimony about her training implied that its substance was authoritative and grounded in some sort of behavioral science.
The prosecution treated Palfreyman's training on grooming as authoritative. As noted, the state sought Palfreyman's testimony on grooming in aid of establishing that, by massaging M inappropriately, defendant was grooming M and planning or preparing for his later sexual abuse of M. And as intended, the state used Palfreyman's expert testimony at trial as substantive evidence that defendant had groomed and then sexually abused M. In short, Palfreyman told the jury on direct examination that she knows from her training on grooming that an offender engages in grooming conduct, such as massaging, as the prelude to his future abuse of a child: (1) a child-abuse offender grooms a child before sexually abusing the child, (2) the offender engages in grooming to build trust and weaken the child's defenses, and (3) defendant's massaging of M could be considered a grooming behavior.3
And as Palfreyman suggested in describing her training on grooming as a social worker in the child welfare and protection field and a forensic interviewer of child sexual abuse victims, both parties agree that *228grooming behavior has been the subject of research in behavioral science, the study of human behavior. This court has frequently stated that, for the purposes of OEC 702, "scientific" evidence need not necessarily be based on the "hard" sciences in which experiments to control a host of variables can be designed and run to test hypotheses. Rather, the court has held that expert testimony in the realm of the "soft" sciences-that is, social and behavioral sciences, which rely on observation and interpretation of human behavior rather than on controlled experiments or mathematical models-also possess "the increased potential to influence the trier of fact as scientific assertion" and, therefore, must be subjected to the crucible of scientific validation. Marrington , 335 Or. at 561, 73 P.3d 911. Thus, in Jennings , this court held that a doctor's testimony based on clinical diagnoses "bear[s] the mark of science" and was subject to the foundational requirements of Brown and O'Key , even though the science involved was not "hard" science. Jennings , 331 Or. at 304, 14 P.3d 596. See also State v. Milbradt , 305 Or. 621, 630-31, 756 P.2d 620 (1988) (suggesting that an expert's testimony about "how normal children usually react to sexual abuse," if it was relevant at all, was scientific evidence that was subject to the Brown foundational requirements). In light of her credentials and training, which the prosecution highlighted, Palfreyman's expert testimony implied that the training she had received on grooming, and the information about grooming from that training that she conveyed to the jury, was accepted and grounded in behavioral science.
Second, even though the prosecution did not highlight the scientific nature of Palfreyman's testimony or focus its examination on studies, research, and literature in the field that supported her testimony, as was the case in Marrington , lay jurors likely would have accorded the testimony the persuasive value of scientific principle. A number of circumstances surrounding Palfreyman's testimony convince us that that is so.
Palfreyman appeared to the jury as an expert on identifying child sexual abuse. She described her educational background and experience in child welfare and protection and forensic interviewing, and she stated that she had conducted over 600 forensic interviews of children. As presented to the jury, Palfreyman also appeared to have specialized training about grooming by sexual offenders. She testified about her "forensic interview training" and "college courses" in which she had learned about the grooming behaviors "leading up to offending" and what "offenders" did when grooming children and why.
And as an expert, Palfreyman conveyed information about grooming of children for sexual abuse by offenders that, like the delayed-reporting evidence in Marrington , was not common knowledge. She first testified about "some of the activities that [she is] familiar with that might be considered grooming," describing massaging by an offender as one of several examples. Then, Palfreyman explained to the jury that the offender engages in grooming to "lay the ground work for later abuse" and "to build trust and weaken defense[s] of the child." Through that testimony, Palfreyman, appearing as an expert concerning child sexual abuse, conveyed that a child-abuse offender grooms a child before sexually abusing the child, that the grooming is for the purpose of building trust and weakening the child's defenses, and that massaging is an example of a grooming activity. The prosecutor then asked for Palfreyman's penultimate expert conclusion: given M's description of defendant massaging into her chest area, defendant's behavior "could be considered grooming."
For the reasons above, we conclude that Palfreyman's testimony about grooming was "scientific" evidence under OEC 702. It follows, therefore, that the trial court erred in permitting Palfreyman to define the phenomenon of grooming-and her conclusion that defendant had engaged in behavior that could be considered grooming of M for sexual abuse-without first requiring the state to establish its scientific validity.
C. Request for Judicial Notice and Disposition
According to the state, even if the trial court erred when it ruled that testimony concerning grooming behavior is not scientific *229evidence, this court may affirm the trial court's ruling on the ground that sexual grooming possesses a sufficient minimum level of scientific validity to be admissible as scientific evidence under OEC 702. That is, the state contends, the "validity of proffered scientific evidence * * * is a question of law," O'Key , 321 Or. at 309 n. 35, 899 P.2d 663, and this court may take judicial notice of facts supporting the admissibility of proffered scientific evidence, Brown , 297 Or. at 420 n. 7, 687 P.2d 751. And in this case, the state argues, the articles and other research materials cited in defendant's own briefing filed with this court readily establish that the concept of sexual grooming possesses a sufficient minimum level of scientific validity to be admissible as scientific evidence under OEC 702. Thus, the state urges, this court should resolve issues of scientific validity, even though the materials on which the court would rely in doing so were not presented to the trial court.
To decide the matter of scientific validity and reliability for the first time on review, we would be required to decide based on judicial notice of legislative facts-that is, nonadjudicative facts-used to determine the foundational basis for admission of evidence.4 Although we have taken judicial notice of settled social science data to supply legislative facts for purposes of the test for admission of eyewitness identifications in State v. Lawson/James , 352 Or. 724, 740, 291 P.3d 673 (2012), and to describe the techniques used to administer a polygraph test and their reliability and validity in Brown , 297 Or. at 420-38, 687 P.2d 751, we have two concerns with taking judicial notice in this case.
Significantly, the parties have not been given a full opportunity to adduce evidence, including expert testimony, and to provide behavioral science or other pertinent literature on that topic. Although the state notes that defendant provided citations to social science research papers, defendant did so to establish that grooming has been studied by behavioral scientists, in support of his argument that, because grooming is a "recognized concept" in behavioral science, then, as in Brown , Palfreyman's testimony was grounded on a scientific principle that required validation as a foundation before its admission.5 Defendant was not specifically attempting to undercut the foundation for Palfreyman's testimony that offenders will first groom a child by engaging in certain behaviors as a lead-in to future sexual conduct with the child.
Second, and relatedly, we are uncertain regarding whether we have been advised of the full scope and nature of the research on sexual grooming of children and whether there is any dispute in the behavioral science disciplines regarding the validity and reliability of the assertions in Palfreyman's testimony. See John E.B. Myers et al . Expert Testimony in Child Sexual Abuse Litigation , 68 Neb. L. Rev. 1, 143 (1989) (stating that "the clinical and scientific literature does not support the existence of a profile of a 'typical' child sexual abuser" and that profile-like evidence in "the form of testimony describing the 'typical' techniques employed by child molesters to get close to their victims" is similarly suspect as a violation of the rule against character evidence). Accordingly, we decline to address, for the first time on appeal, the scientific validity and reliability of the sexual grooming evidence admitted at trial.
This court faced a similar situation in Marrington and remanded the case to the trial court. As in this case, the trial court in Marrington had overruled the defendant's objection that the proffered expert testimony was untested scientific evidence and admitted *230the evidence. This court held that the expert's testimony was scientific evidence and that the trial court had erred in admitting it without requiring the state first to establish its scientific validity. This court determined that the error was prejudicial and remanded the case to the trial court for further proceedings, without determining for the first time on appeal that delayed reporting evidence was scientifically valid. Marrington , 335 Or. at 566, 73 P.3d 911.6 As this court has stated, "in the absence of a clear case, a case for judicial notice, or a case of prima facie legislative recognition, trial courts have an obligation to ensure that proffered expert scientific testimony that a court finds possesses significantly increased potential to influence a trier of fact as 'scientific' assertions is scientifically valid." O'Key , 321 Or. at 293, 899 P.2d 663 (footnote omitted). This is not such a clear case, and the trial court is best suited for the development of the evidentiary record concerning admissibility.
D. Prejudicial Error
Having concluded that the trial court erred in failing to determine whether sexual grooming evidence possesses the requisite level of scientific validity and reliability for admissibility under OEC 702, and having declined to make that determination ourselves, we turn to the question whether the error was prejudicial and requires reversal. It is axiomatic that evidentiary error does not require reversal if it is harmless-that is, if it had little likelihood of affecting the verdict. State v. Newman , 353 Or. 632, 647, 302 P.3d 435 (2013) ; State v. Davis , 336 Or. 19, 28, 77 P.3d 1111 (2003).
In making a determination of harmlessness, the court does not ask whether the evidence of guilt is substantial or compelling, but rather whether the trial court's error was likely to have influenced the verdict. Davis , 336 Or. at 32, 77 P.3d 1111. If erroneously admitted evidence relates to a "central factual issue" in the case, it is more likely to have affected the verdict than evidence relating to a tangential issue. Marrington , 335 Or. at 566, 73 P.3d 911. And, as we have said, scientific evidence or evidence perceived by lay jurors to be scientific in nature possesses an unusually high degree of persuasive power. O'Key , 321 Or. at 291, 899 P.2d 663 ; see also State v. Willis , 348 Or. 566, 573, 236 P.3d 714 (2010) ("Any juror would wish to have the scientific answer, and could be expected to give it weight.").
The state argues that, in this case, any error in admitting evidence about sexual grooming was harmless, because other evidence tending to show defendant's sexual purpose toward M was admitted unchallenged. That is, the jury heard evidence about defendant massaging M in ways that made her uncomfortable and about the incident that occurred when M was five or six years old, when defendant asked M to touch his penis. Therefore, according to the state, the jury would have been expressly called upon by the evidentiary record to consider whether defendant's prior conduct toward M suggested a nefarious sexual purpose, even if Palfreyman's testimony about the grooming behaviors of sexual offenders had not drawn the jury's attention in that direction.
We do not think that the matter is so clear-cut. That is so for at least three reasons: (1) the credibility of M and defendant was a central issue at trial; (2) the expert testimony on grooming appeared to be grounded in science; and (3) the prosecution's use of the grooming evidence was not as limited as the state's argument suggests.
Other than M, there were no witnesses to the claimed sexual abuse or to the incident that occurred when M was five or six years old, and there was no physical evidence corroborating the alleged abuse. There was evidence that M's mother did not think that defendant had behaved inappropriately when massaging M, and there was evidence that, at one point, M had recanted her story about defendant asking her to touch his penis when she was five or six years old. Defendant's recorded voicemail message for M's father could be viewed by jurors as indicating that he had engaged in some improper conduct on *231the camping trip (defendant said, "Because if I did do something, I didn't mean anything-I don't even know what I did. I know for a fact that I was not naked in bed."), but defendant denied abusing M. Therefore, the credibility of defendant and M was a "central factual issue," and the grooming evidence bolstered M's testimony.
Further, as we have held, Palfreyman's testimony was "scientific" evidence and, as such, possessed a high degree of persuasive power. The prosecutor recognized that power and, pointing to Palfreyman's testimony about massaging as a grooming behavior, made a syllogistic argument in closing. He reminded the jury, "Courtney Palfreyman talked to you about offenders trying to get children accustom[ed] to touching" through grooming. He argued that defendant's "massaging is important because it's grooming behavior." Indeed, he said, defendant's massaging was "classic grooming behavior." The logic of the prosecutor's argument was that, because (1) offenders groom children by massaging and (2) defendant groomed M by massaging her, it follows that defendant was an offender-that he sexually abused M. Stated another way, the prosecutor was able to argue that, because "offenders" act in certain ways to groom a child as a lead-in to sexual abuse of the child, and defendant also acted in one of those ways, he groomed M and was an offender.
Finally, the jury likely viewed Palfreyman's testimony about grooming as substantive evidence that defendant had sexually abused M in the camper as charged. The availability of Palfreyman's grooming testimony made it easy for the jury to draw the same inference that this court criticized in State v. Hansen , 304 Or. 169, 174, 743 P.2d 157 (1987) : that defendant had sexually abused M because he had engaged in massaging, an act that sexual abusers of children also engage in. Hansen concerned an objection to a police detective's experience-based testimony concerning grooming based on the considerations in OEC 403. This court explained that testimony concerning grooming techniques that a child abuser may use-such as "physical and psychological 'testing' of the child, giving gifts, showing affection, praising, making the child feel comfortable in the abuser's presence, etc."-had almost no relevance as evidence that the defendant committed sexual abuse of the child. Id. at 176, 743 P.2d 157. The court stated: "That child abusers use these techniques has no bearing on whether a person who does these things is a child abuser." Id. This court then concluded that, under the balancing test required by OEC 403, "the danger of unfair prejudice to defendant from the unwarranted inference that, because defendant engaged in acts that sexual child abusers engage in, she, too, is a sexual child abuser is simply too great." Id.
Without Palfreyman's testimony, the prosecutor would have been left with a much weaker argument to the jury that defendant had appeared sexually interested in M, based on her testimony concerning the massages and the earlier incident when she was five or six years old. The prosecution could not have argued that defendant-like other offenders, according to Palfreyman's testimony-groomed his child victim for later sexual conduct. Contrary to the dissent, we conclude that the expert testimony and the prosecutor's argument were likely influential in the jury's deliberations and verdict.
For those reasons, we cannot say that there was "little likelihood" that the jury resolved the credibility dispute between defendant and M by relying on Palfreyman's testimony that defendant's behavior-specifically, massaging M's chest-is recognized as grooming behavior in sexual offenders. It follows that this court must reverse defendant's conviction and remand the case to the trial court for further proceedings.
The decision of the Court of Appeals and the judgment of the circuit court are reversed, and the case is remanded to the circuit court for further proceedings.
Kistler, J., dissented and filed an opinion in which Balmer, J., joined.
As for the incident when she was five or six years old, M recounted that defendant wanted her to touch his "wiener" and that her mother responded to M's report of that conduct by assuring her that defendant would not go into her room again and telling M to sleep the rest of the night with her. Later during the trial, one of M's friends testified that M initially had said that defendant asked M to touch his penis, but that M later recanted, telling the friend that she had lied and made up the incident.
Although the New Mexico Supreme Court has suggested in dicta that whether a defendant's conduct constituted grooming according to a scientific or specialized definition would require an expert to explain the theory, State v. Sena , 144 N.M. 821, 827, 192 P.3d 1198 (2008), a number of other jurisdictions proceed on the assumption that grooming testimony is scientific or at least some kind of specialized expert testimony, and then move on to determine whether the testimony is reliable and helpful to the jury. See, e.g. , Morris v. State , 361 S.W.3d 649, 667 (Tex. Crim. App. 2011) (concluding that grooming as a phenomenon exists, that a law enforcement official with a significant amount of experience with child sex abuse cases may be qualified to talk about it, and that it involves matters beyond the understanding of the jury and therefore is useful to the jury); United States v. Romero , 189 F.3d 576, 584-85 (7th Cir. 1999) (noting different standards under Daubert and Kumho Tire Co, Ltd v. Carmichael , 526 U.S. 137, 119 S.Ct. 1167, 143 L.Ed. 2d 238 (1999), for scientific and nonscientific evidence, holding that FBI agent's expert testimony concerning grooming was reliable and helpful without determining whether it was scientific or nonscientific); United States v. Hayward , 359 F.3d 631, 636 (3d Cir. 2004) (holding that grooming testimony was not inadmissible under FRE 704(b), which prohibits an expert from testifying about the defendant's mental state).
To recap her specific expert testimony on grooming, Palfreyman began by giving examples of grooming conduct: "spending time together"; "sometimes allowing the child to do things the parents wouldn't allow like video games when it's not allowed, or alcohol use"; giving the child "money for things"; and "tickling, massaging, that sort of thing." In terms of defining grooming, Palfreyman agreed with the prosecutor that grooming is "done by the offender" to "lay the groundwork for later abuse." She added that an offender engages in grooming to "build trust and weaken defense[s] of the child." When asked whether any of defendant's behavior that M described in the CARES interview "could be considered grooming," Palfreyman identified defendant's massaging of M. She acknowledged on cross-examination that whether any particular behavior by a person is grooming depends on the person's intent.
Adjudicative facts are " 'the facts of the particular case.' " Chartrand v. Coos Bay Tavern , 298 Or. 689, 693, 696 P.2d 513 (1985) (quoting commentary to OEC 201(a) ). Judicially noticed adjudicative facts must be either generally known in the jurisdiction, OEC 201(b)(1), or "[c]apable of accurate and ready determination," OEC 201(b)(2). In other words, adjudicative facts must have a "high degree of indisputability." Chartrand , 298 Or. at 693, 696 P.2d 513. We do not decide whether any part of Palfreyman's testimony about grooming in this case constituted an adjudicative fact.
We do not view that statement as a concession that the clinical and scientific literature supports the existence of a profile of offenders that includes typical grooming techniques employed by sexual offenders against children.
In State v. Perry , 347 Or. 110, 218 P.3d 95 (2009), this court subsequently concluded that evidence concerning the phenomenon of delayed reporting was scientifically valid.