The PEOPLE of the State of Colorado v. Kerry Lee COOPER

496 P.3d 430 The PEOPLE of the State of Colorado, Petitioner, v.Kerry Lee COOPER, Respondent. Supreme Court Case No. 19SC249 Supreme Court of Colorado. September 27, 2021 Rehearing Denied October 25, 2021 Attorneys for Petitioner: Philip J. Weiser, Attorney General, Katharine Gillespie, Senior Assistant Attorney General, Denver, Colorado Attorneys for Respondent: Megan A. Ring, Public Defender, Tracy C. Renner, Deputy Public Defender, Denver, Colorado En Banc JUSTICE SAMOUR delivered the Opinion of the Court. ¶ 1 Before expert testimony may be presented to a jury, it must pass through the gate of admissibility — a gate trial courts have been entrusted with vigilantly guarding. As gatekeepers, trial courts play an important role in ensuring that expert testimony is not admitted unless it "both rests on a reliable foundation and is relevant to the task at hand." Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 597, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). Two decades ago, we ushered in a four-part test to channel a trial court's discretion in determining the admissibility of expert testimony under CRE 702 : (1) the scientific principles underlying the testimony must be reasonably reliable; (2) the expert must be qualified to offer the testimony; (3) the testimony must be helpful to the jury; and (4) the evidence must satisfy CRE 403. People v. Shreck, 22 P.3d 68, 77-79 (Colo. 2001). The questions we agreed to review in this case and the companion case of People v. Coons, 2021 CO 70, 495 P.3d 961, focus our attention on the third factor—helpfulness to the jury— in the context of generalized expert testimony (i.e., testimony aimed at educating the jury about general concepts or principles without attempting to discuss the particular facts of the case).1 ¶ 2 We have previously explained that helpfulness to the jury hinges on "fit"—the expert testimony must fit the case. The overriding question squarely before us today is just how close a fit is required before generalized expert testimony may be admitted. ¶ 3 We now hold that generalized expert testimony fits a case if it has a sufficient logical connection to the factual issues to be helpful to the jury while still clearing the ever-present CRE 403 admissibility bar. In evaluating the fit of generalized expert testimony, a trial court must be mindful of the purposes for which such testimony is offered —that is, the reasons why the proponent of the evidence has asked the expert to educate the jury about certain concepts or principles. ¶ 4 The trial court in this case restricted the generalized expert testimony offered by the People regarding the dynamics of domestic violence. After considering the victim's testimony, the proposed expert opinions, and the reasons why the People were seeking to educate the jury, the trial court allowed only those opinions which it viewed as having a sufficient logical relation to the facts to be helpful to the jury without violating CRE 403. Because the trial court properly exercised its discretion in carrying out its gatekeeping function, and because the standard of review pertaining to the admission of expert testimony is highly deferential, the court of appeals erred in reversing the judgment of conviction. In our view, the court of appeals adopted a fit standard that is inflexible and overly exacting. ¶ 5 While generalized expert testimony must fit the case, the fit need not be perfect. In other words, each aspect of such testimony need not match a factual issue. Since generalized expert testimony, by definition, seeks to inform the jury about generic concepts or principles without knowledge of the facts, it is almost inevitable that parts of such testimony will not be logically connected to the case. For that reason, the fit inquiry must be flexible. A trial court should certainly not be expected to parse the proposed testimony and determine whether each statement the expert intends to utter is logically connected to a fact in the case. If the generalized expert testimony's logical connection to the factual issues is sufficient to be helpful to the jury without running afoul of CRE 403, the testimony fits the case. ¶ 6 Still, attorneys and trial courts should do their best to avoid introducing generalized expert testimony that has no logical connection to the facts of the case. As relevant here, prosecutors should take care to endorse generalized expert testimony about domestic violence only in appropriate cases; and, when they do so, they should endeavor to present only testimony that is logically connected to the factual issues. Trial courts, in turn, should exercise their discretion in deciding whether to permit all, some, or none of the proffered testimony under the fit standard we articulate today. In doing so, trial courts should consider the feasibility and propriety of admitting only a portion of the proposed generalized expert testimony on a particular subject. ¶ 7 Because the court of appeals incorrectly overturned the judgment of conviction, we reverse. We remand for further proceedings consistent with this opinion. I. Facts and Procedural History ¶ 8 The defendant, Kerry Lee Cooper, met L.K. through a mutual friend when Cooper was in his mid-fifties and L.K. was in her mid-twenties. L.K. was living on the streets at the time, and Cooper invited her to stay at his house. She accepted his invitation and moved in with him. About a month later, they became romantically involved. Approximately a year and a half into their intimate relationship, Cooper and L.K. had a physical altercation that resulted in the charges brought against Cooper in this case. A. The Incident ¶ 9 During a hot summer night, L.K. woke up in a panic because of the heat. Cooper's house lacked air conditioning and had many windows that were boarded up. L.K. felt like she couldn't breathe, so she asked Cooper to plug in a nearby fan. Cooper obliged but left the fan facing the floor. When L.K. asked h im to reposition the fan toward her, he told her to leave it alone. What happened next is less certain because Cooper and L.K. provided different accounts of those events at trial. 1. L.K.'s Version of Events ¶ 10 L.K. tried to reposition the fan herself, but Cooper became angry and shoved the running fan into her face, cutting her with the fan blades. She grabbed a flashlight from the nightstand and hit Cooper on the head with it to take the fan away from him. At that point, Cooper dropped the fan and began punching L.K. in the face and ribs. After L.K. crawled across the bed, made her way to an open window, and screamed for help, Cooper told her to shut up, grabbed her by the jaw, and inserted his fingers into her mouth. She proceeded to bite down on one or more of his fingers. Once he was able to free his fingers from her mouth, he snatched a tire iron that was just outside the bedroom and threatened to hit her with it if she didn't stop screaming. L.K. cowered on the bed but continued screaming. Cooper then hit her twice in the head with the tire iron. During the attack, L.K.'s dog, Buddy, came into the room and started growling at Cooper. In response, Cooper struck Buddy twice in the face.2 2. Cooper's Version of Events ¶ 11 L.K. was unhappy with the location of the fan and started yelling at Cooper to reposition it. He unplugged it, threw it on the end of the bed, and told her to plug it in herself. As he was attempting to go back to sleep, she hit him on the head with a flashlight. He tried to take the flashlight away from her to protect himself, but she bit his finger hard enough to draw blood. To get her to open her mouth and release his finger, he pushed her in the forehead. ¶ 12 He did not shove the fan into L.K.'s face, punch her, or grab her jaw. Further, he did not threaten her —let alone hit her —with the tire iron; in fact, he never even picked up the tire iron. And he never hit Buddy. 3. The Aftermath ¶ 13 Cooper's daughter, who lived two houses away and heard L.K. screaming, called 911. When officers arrived, they observed that L.K. had swollen lips, dried blood around her mouth, a bruised nose, bruising around her eyes, bruising and marks on her ribs, and red marks on her scalp. In addition, they noticed a bump on her head. The officers also documented that Cooper had a bloody finger. ¶ 14 Cooper received medical treatment for his bloody finger at the scene. L.K., on the other hand, refused medical treatment for her injuries. A few days later, she apparently saw a doctor, who informed her that she had sustained a broken nose.3 ¶ 15 L.K. moved out of the house immediately after the incident and never moved back. But she saw Cooper once more for approximately a half hour, and she had "more than a few" additional phone calls with him. During those conversations, she told him that she still loved him. B. Arrest, Charges, and Relevant Pretrial Proceedings ¶ 16 Cooper was arrested on scene. He was later charged with one felony (menacing with a deadly weapon) and three misdemeanors (third degree assault, harassment, and cruelty to animals). The People alleged that the act underlying each offense was an act of "domestic violence," as that term is defined by Colorado law. See § 18-6-800.3(1), C.R.S. (2020) (" ‘Domestic violence’ means an act or threatened act of violence upon a person with whom the actor is or has been involved in an intimate relationship," as well as "any other crime against a person, or against property, including an animal, ... when used as a method of coercion, control, punishment, intimidation, or revenge directed against a person with whom the actor is or has been involved in an intimate relationship.").4 ¶ 17 On two separate occasions, the People sought and obtained a postponement of the trial because L.K. was being uncooperative. During one of the pretrial hearings, Cooper spontaneously told the court that the People "don't have a witness" because L.K. "don't want to testify against me." Cooper added that L.K. had already "told the DA that she wasn't going to be there in the first place because she wasn't going to testify against me." At a subsequent pretrial hearing, the People informed the court that Cooper's daughters had threatened L.K. And the People later submitted a police report documenting that allegation. ¶ 18 Months before trial, the People endorsed Janet Kerr, a licensed professional counselor and the executive director of a domestic violence and sexual assault agency. The People informed the court and Cooper that Kerr intended to offer generalized expert testimony on the dynamics of domestic violence. Cooper responded by filing a motion in limine related to that proposed testimony. The court considered Kerr's endorsement and Cooper's corresponding motion in two separate hearings held during the trial. C. Jury Trial and Sentencing Proceedings ¶ 19 The court initially took up Kerr's proffered testimony on the first day of trial. Kerr had admittedly never met L.K. or Cooper, much less analyzed their relationship, and had no plans to opine about whether L.K. was telling the truth or whether Cooper was guilty of any charge. The purpose of her endorsement, as the court later described it, was simply to educate the jury "about the dynamics of abusive relationships so that the jury could more accurately assess the credibility of the witnesses given some counterintuitive dynamics that are commonly seen in abusive relationships." ¶ 20 Cooper asked the court to defer ruling on his motion until after L.K. testified so that the court could better assess whether Kerr's proffered testimony was relevant. Although the court originally ruled that Kerr could testify, it later agreed to table the issue until after L.K.'s testimony. It reasoned that doing so would allow it to evaluate whether Kerr's opinions were relevant in the context of the People's allegations. ¶ 21 Following L.K.'s testimony, which was largely consistent with her previous statements, the court revisited Cooper's motion outside the presence of the jury. At the outset, the court announced that it had "serious concerns" about admitting Kerr's opinions in this case. The court was familiar with generalized expert testimony related to the dynamics of abusive intimate relationships because it had allowed such testimony in other domestic violence trials to address a victim's recantation, minimization, or lack of cooperation. But the court was worried about whether such testimony fit this particular case because L.K. had responded to her subpoena and had been cooperative with the prosecutor during the trial: The most common trait that's counterintuitive that we see in the courtroom are victims of domestic violence either recanting their original statement to authorities or minimizing their statement to authorities or failing to obey a subpoena, and I think it's very important for the record to reflect that this case is not one of those. The court did not hear anything in [L.K.'s] testimony that could fairly be characterized as a minimization or a recantation in any way. Given L.K.'s testimony, the court asked the People to identify "the counterintuitive fact in this case" about which Kerr needed to educate the jury. ¶ 22 The People responded that recantation, minimization, and lack of cooperation are not defining characteristics of domestic violence; they are merely "offshoots" of it. Instead, maintained the People, "power and control" are the defining characteristics of domestic violence, and Kerr was prepared to educate the jury about those overarching concepts through the "Power and Control Wheel" (a tool developed by social scientists to explain the common dynamics of domestic violence). ¶ 23 The Power and Control Wheel, which we reproduce below, prominently features "Power and Control" in the hub, refers to physical and sexual violence along the top and bottom peripheries, and includes eight spokes representing the nonphysical forms of abuse through which power and control are typically exerted in an intimate relationship: (1) coercion and threats; (2) intimidation (including through pet abuse); (3) emotional abuse; (4) isolation; (5) making light of the abuse, denying it happened, or blaming the victim for it; (6) using children; (7) male privilege; and (8) economic abuse.5 ? ¶ 24 The People further asserted that there were some behaviors by L.K. which, while not among the most commonly seen in domestic violence cases, would nevertheless seem counterintuitive to jurors. Specifically, the People discussed: (1) L.K.'s refusal to be transported to a hospital after the assault, something that she and a responding officer had been questioned about on cross-examination; and (2) L.K.'s decision to continue to have in-person and phone contact with Cooper after the incident to express her love for him, which, according to the People, reflected the presence of the cycle of violence. The People also maintained that Kerr's testimony was necessary to educate the jury about several other matters: (1) the kindness-violence dichotomy reflected in Cooper's offer to provide L.K. a warm place to live and his subsequent attack against her — contrasting behaviors that the People feared could lead some jurors to question the credibility of the allegations; (2) "the age differential" of approximately thirty years between Cooper and L.K., which the People believed contributed to the power and control he exerted over her; (3) the assault on Buddy, which the People considered a form of intimidation and thus relevant to Cooper's power and control over L.K.; and (4) L.K.'s financial dependence on Cooper, which the People urged was part and parcel of the power-and-control dynamic. ¶ 25 Picking up on the court's initial concerns, Cooper countered that Kerr's testimony might have been admissible if L.K. had recanted or minimized her allegations to the police. But, insisted Cooper, since L.K. had neither recanted nor minimized her previous statements, Kerr's opinions were irrelevant. Further, argued Cooper, one isolated incident of alleged domestic violence does not a cycle of violence make. And, even assuming the prosecutor wasn't intending to ask Kerr questions related to the cycle of violence, Cooper deemed the proposed expert testimony problematic because, in his view, it incorrectly assumed that L.K. was a victim of domestic abuse. Thus, Cooper asked the court to exclude Kerr's opinions on the ground that they were irrelevant and constituted improper bolstering of the People's theory of the case without enough connection to the facts. ¶ 26 The court was partially persuaded by the People. Hence, although it prohibited Kerr from testifying about recantation, minimization, and lack of cooperation by victims of domestic abuse, it allowed the proffered opinions that had a sufficient logical connection to the facts of this case to be helpful to the jury without violating CRE 403. ¶ 27 First, the court permitted Kerr to educate the jury generally about domestic violence and the dynamic of power and control in an abusive intimate relationship. Relatedly, the court allowed Kerr to rely on the Power and Control Wheel. ¶ 28 Second, the court ruled that Kerr could discuss a domestic violence victim's refusal to receive medical treatment immediately after an incident. In so doing, the court agreed with the People that, given L.K.'s injuries, her response to the paramedics' offer to transport her to a hospital would likely seem "counterintuitive to a jury." ¶ 29 Third, the court acknowledged that "some jurors may need help in understanding why an act of abuse could have occurred but a person still communicates with the abuser after the fact," including to tell him that she continues to love him. Accordingly, it granted the People's request to have Kerr offer an expert opinion on that subject. ¶ 30 Fourth, the court found that Kerr's testimony about a significant age difference in a couple was admissible. The court viewed this evidence as relevant to issues of power and control in an abusive intimate relationship. ¶ 31 And fifth, the court concluded that Kerr could opine about abuse of a pet as a form of intimidation in an intimate relationship. The court believed that this evidence, too, was relevant given Cooper's alleged assault against Buddy. ¶ 32 But the court set boundaries around the generalized expert testimony it was admitting. It told the parties that Kerr was "not going to be allowed to give an opinion" about whether she thought L.K. "was telling the truth" or whether Cooper was "guilty." Instead, Kerr would simply be permitted to provide "general information to educate the jury" about the concepts and ideas the court had identified. ¶ 33 Immediately after the court's ruling, the People made a last-ditch push to introduce Kerr's opinion on the pressure a victim in an abusive intimate relationship may feel to continue the relationship because she is financially dependent on her partner. But the court wouldn't budge. It observed that, while L.K. may have felt some financial pressure to remain in the relationship with Cooper, the proposed opinion was typically used as "a plausible explanation" for a victim's recantation or minimization: "[S]omeone would not want their significant other jailed for the offense because they need the [financial] security ... that comes with having a place to live, etc." Here, explained the court, there was a real "danger for misuse" of this testimony by the jury both because L.K. was homeless when she met Cooper and because L.K. had not resumed her intimate relationship with him after the incident.6 Therefore, determined the court, any opinion about a victim's financial pressure to remain in an abusive relationship, while arguably relevant, was "unfairly prejudicial" and inadmissible under CRE 403.7 ¶ 34 The comments that followed reflect that the court was doing its utmost to allow only the generalized expert testimony that fit the facts of this case: The more we go on about these things, the more the subtle implication is ... that this information can be used to diagnose whether, in fact, an act of abuse occurred. And ... the status of the science is that there are no traits that are usable as diagnostic tools to profile who's an abuser and who is not. The Court is concerned that this testimony, if allowed to go too far and just allow [Kerr] to say everything she usually says even when it doesn't uniquely fit this case, that it's a danger to misuse by the jury . (Emphasis added.) ¶ 35 Pursuant to the court's ruling, Kerr was qualified as an expert in the area of domestic violence and, more specifically, power-and-control issues. After discussing what social science generally teaches about domestic violence, Kerr spoke about the kindness-violence dichotomy present in most abusive intimate relationships. But Kerr devoted most of her time to the power-and-control dynamic, which she said is at the heart of an abusive intimate relationship. Using the Power and Control Wheel, she talked about both physical violence (including of a sexual nature) and nonphysical violence. As to the latter, she provided examples of the behaviors listed within each spoke of the Power and Control Wheel. ¶ 36 Kerr also addressed the common types of counterintuitive behaviors in which domestic violence victims engage. As well, she talked about how the existence of a large age difference between intimate partners and the use of pet cruelty can affect the dynamic of power and control in an abusive intimate relationship. Though Kerr's testimony included some subjects not specifically related to any aspect of this litigation —including certain background information on domestic violence and examples of behaviors frequently seen in victims and abusers — the prosecutor focused her questions on the areas that fit the facts of this case.8 ¶ 37 At one point during Kerr's direct examination, the prosecutor asked about the "cycle of violence" and what it means. However, the court sustained Cooper's timely objection and did not allow Kerr to opine about that concept. It ruled that, since L.K. had not rekindled her romance with Cooper after the incident, any discussion about the cycle of violence was irrelevant.9 ¶ 38 The jury ultimately returned mixed verdicts. It found Cooper guilty of third degree assault and harassment but not guilty of menacing and cruelty to animals. During the sentencing hearing, the trial court found that the acts underlying both convictions constituted acts of domestic violence under section 18-6-800.3(1). D. The Appeal and the Petition for Certiorari ¶ 39 Cooper appealed his convictions, and a division of the court of appeals reversed in a split, published opinion. People v. Cooper , 2019 COA 21, 490 P.3d 420. The division recognized that the value of generalized expert testimony in appropriate cases is well established in Colorado. Id. at ¶ 3, 490 P.3d at 422. In fact, the division acknowledged that generalized expert testimony is "crucial" in some cases to educate the jury on what social science has learned about the counterintuitive behaviors in which domestic violence victims engage. Id. Absent such testimony, noted the division, jurors may well reach erroneous decisions. Id. But the division concluded that "virtually all" of the generalized expert testimony provided by Kerr "was wholly irrelevant to the issues properly before the jury." Id. at ¶ 5, 490 P.3d at 423. ¶ 40 As part of its analysis, the division used a two-column chart that catalogued, side-by-side, each generalized expert opinion and the evidence presented to the jury that had any relationship to it. Id. at ¶ 65 app., 490 P.3d at 432-39 app. Finding that there was no record evidence that related to the vast majority of the generalized expert opinions introduced, the division determined that Kerr's testimony did not fit the factual issues. Id. at ¶¶ 25-32, 490 P.3d at 425-26. The division added that, even if some of Kerr's testimony was arguably supported by the record, any fit was "very minor" and could not "excuse the admission of extensive irrelevant evidence of this type." Id. at ¶ 31, 490 P.3d at 426. Accordingly, the division held that the trial court had abused its discretion in allowing Kerr's generalized expert testimony. Id. at ¶ 32, 490 P.3d at 426. And, because the division ruled that this evidence was highly prejudicial, it reversed and remanded for a new trial. Id. at ¶¶ 33-44, 490 P.3d at 426-28. ¶ 41 Having concluded that a new trial was necessary, the division proceeded to consider and reject Cooper's assertion that he was entitled to a unanimity instruction. Id. at ¶¶ 45-48, 490 P.3d at 428-29. The division didn't reach Cooper's contention that the prosecutor had engaged in misconduct during closing argument because there was no way of knowing whether the same statements would be uttered at the retrial. Id. at ¶¶ 49-50, 490 P.3d at 429. ¶ 42 Judge Roman concurred in part and dissented in part. He agreed that the trial court had abused its discretion in admitting what he viewed as largely irrelevant expert testimony, but he believed that the error was harmless. Id. at ¶ 52, 490 P.3d at 429 (Roman, J., concurring in part and dissenting in part). He thus would have affirmed. Id. at ¶ 53, 490 P.3d at 429. ¶ 43 The People then filed a petition for certiorari. We granted the People's petition in part.10 II. Analysis A. Standard of Review ¶ 44 "We review a trial court's admission of expert testimony for an abuse of discretion and will reverse only when that decision is manifestly erroneous." People v. Rector, 248 P.3d 1196, 1200 (Colo. 2011). This is a deferential standard that reflects the superior opportunity a trial court has to assess both the competence of an expert witness and whether that witness's anticipated opinions would be helpful to the jury. Id. B. Admissibility of Generalized Expert Testimony ¶ 45 Except where "otherwise provided by constitution, statute, or rule, all relevant evidence is admissible." People v. Rath, 44 P.3d 1033, 1038 (Colo. 2002) (citing CRE 402 ). Evidence is relevant if it has "any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence." Id. (quoting CRE 401 ). Under CRE 403, relevant evidence "may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence." ¶ 46 It is now canonical that CRE 702 governs the admissibility of scientific or other expert testimony in Colorado. See Shreck, 22 P.3d at 70. CRE 702 provides that "a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify ... in the form of an opinion or otherwise" when her "scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue." The two focal points of the CRE 702 inquiry are the "reliability and relevance of the proffered evidence." Shreck, 22 P.3d at 70. ¶ 47 To determine whether expert testimony is reliable, a trial court should consider both whether the scientific principles as to which the witness is testifying are reasonably reliable and whether the witness is qualified to opine on the matters in question. Id. at 77. To determine whether expert testimony is relevant , a trial court should consider whether it would be useful to the jury. Id. Of course, a trial court must always exercise its discretion to ensure that the probative value of relevant evidence is not substantially outweighed by any of the dangers identified in CRE 403. Id. at 70. In short, then, in ascertaining the admissibility of expert testimony, a trial court must employ a four-part test: (1) whether the scientific principles underlying the testimony are reasonably reliable; (2) whether the expert is qualified to offer the testimony; (3) whether the testimony would be helpful to the jury; and (4) whether the testimony satisfies CRE 403. Id. at 77-79. ¶ 48 The third prong, helpfulness to the jury, which belongs on the relevance side of the admissibility ledger, is front and center in this appeal. Proffered expert testimony is helpful if it "will assist the fact finder to either understand other evidence or to determine a fact in issue." Lanari v. People, 827 P.2d 495, 502 (Colo. 1992) ; accord CRE 702 (indicating that expert testimony is admissible under certain circumstances if it "will assist the trier of fact to understand the evidence or to determine a fact in issue"). "The critical question trial courts must answer" to discern helpfulness to the jury is: " ‘On this subject can a jury from this person receive appreciable help?’ " People v. Fasy, 829 P.2d 1314, 1316 (Colo. 1992) (quoting People v. Williams, 790 P.2d 796, 798 (Colo. 1990) ). The test calls for "a common sense inquiry: whether an untrained lay person would be qualified to determine a particular issue ‘intelligently and to the best possible degree without enlightenment from those having a specialized understanding of the subject involved in the dispute.’ " Lanari, 827 P.2d at 502 (quoting Fed. R. Evid. 702 Advisory Committee's Note). ¶ 49 Helpfulness turns on whether the testimony fits the particular case. People v. Martinez, 74 P.3d 316, 323 (Colo. 2003). "Fit demands more than simple relevance; it requires that there be a logical relation between the proffered testimony and the factual issues involved in the litigation." Id. Analyzing the fit between expert testimony and the factual issues in a case necessarily requires consideration of the purposes for which the evidence is being offered. Id. ¶ 50 Courts sometimes permit generalized expert testimony to assist the jury: [I]t might ... be important in some cases for an expert to educate the factfinder about general principles, without ever attempting to apply these principles to the specific facts of the case. For example, experts might instruct the factfinder on the principles of thermodynamics, or bloodclotting, or on how financial markets respond to corporate reports, without ever knowing about or trying to tie their testimony into the facts of the case. Fed. R. Evid. 702, Advisory Committee's Note, 2000 Amendments. The Advisory Committee's Note goes on to refer to "the venerable practice of using expert testimony to educate the factfinder on general principles." Id. We need not undertake a far-flung expedition to find evidence of this practice: Our own court of appeals has acknowledged on more than one occasion the admissibility of generalized expert testimony to assist the jury in appropriate cases. See People v. Relaford, 2016 COA 99, ¶¶ 26-30, 409 P.3d 490, 495-96 (collecting cases). ¶ 51 This is not our first generalized-expert-testimony rodeo. Martinez , a shaken-impact syndrome case, involved such testimony. 74 P.3d at 323. A physician there provided a "description of accident scenarios which result in subdural hematomas," the type of injury that also occurs in shaken-impact syndrome. Id. Although we made clear that generalized expert testimony cannot be helpful to jurors if it doesn't fit the facts of the case, we didn't have occasion there to analyze just how close a fit is required. Id. at 323-25. We do here. ¶ 52 We hold that generalized expert testimony fits a case if it has a sufficient logical connection to the factual issues to be helpful to the jury while still clearing the ever-present CRE 403 admissibility bar. In evaluating the fit of generalized expert testimony, a trial court must be mindful of the purposes for which such testimony is offered — that is, the reasons why the proponent of the evidence has asked the expert to educate the jury about certain concepts or principles. ¶ 53 While generalized expert testimony must fit the case, the fit need not be perfect. In other words, each aspect of such testimony need not match a factual issue. Since generalized expert testimony, by definition, seeks to inform the jury about generic concepts or principles without knowledge of the facts, it is almost inevitable that parts of such testimony will not be logically connected to the case. For that reason, the fit inquiry must be flexible. A trial court should certainly not be expected to parse the proposed testimony and determine whether each statement the expert intends to utter is logically connected to a fact in the case. If the generalized expert testimony's logical connection to the factual issues is sufficient to be helpful to the jury without running afoul of CRE 403, the testimony fits the case. ¶ 54 Still, attorneys and trial courts should do their best to avoid introducing generalized expert testimony that has no logical connection to the facts of the case. As relevant here, prosecutors should take care to endorse generalized expert testimony about domestic violence only in appropriate cases; and, when they do so, they should endeavor to present only testimony that is logically connected to the factual issues. Trial courts, in turn, should exercise their discretion in deciding whether to permit all, some, or none of the proffered testimony under the fit standard we articulate today. In doing so, trial courts should consider the feasibility and propriety of admitting only a portion of the proposed generalized expert testimony on a particular subject. ¶ 55 So, did the trial court abuse its discretion in allowing Kerr to provide generalized expert testimony related to abusive intimate relationships? The division held that it did. But we beg to differ. Although the division correctly concluded that generalized expert testimony on the dynamics of abusive intimate relationships is not automatically admissible in every domestic violence case — the testimony must fit the facts of the case — it applied an overly arduous fit standard that lacks pliability. The division thus incorrectly determined that the fit between Kerr's opinions and the facts of this case was largely nonexistent or, at best, "very minor." Cooper , ¶ 31, 490 P.3d at 426. Applying the correct standard, we rule that the trial court did not abuse its discretion in finding that Kerr's opinions fit the facts of this case. C. The Trial Court Did Not Abuse Its Discretion ¶ 56 The trial court shrewdly deferred ruling on Cooper's motion regarding Kerr until after L.K. testified so that it could better evaluate whether the generalized expert testimony under challenge fit the facts of this case. Following L.K.'s testimony and argument by counsel, and considering the purposes for Kerr's endorsement, the court exercised its discretion and allowed some, but not all, of Kerr's proffered opinions. And the court appeared to be guided by the analytical framework we embrace today: It admitted only opinions which, based on their logical connection to the facts, would be helpful to the jury while still passing muster under CRE 403. ¶ 57 On the one hand, the court barred Kerr from testifying about recantation, minimization, and lack of cooperation by domestic violence victims. Because there was no evidence that L.K. had recanted, minimized, or failed to cooperate, the court aptly found that such testimony had no logical connection to the facts of this case and would not be helpful to the jury. Even when Kerr's proffered opinions had some logical connection to the facts of this case, the court appropriately excluded them if they were not up to snuff under CRE 403. That was the court's approach with respect to Kerr's proffered opinion on the financial pressure that may cause a victim of domestic violence to remain in an abusive relationship.11 ¶ 58 On the other hand, the court permitted Kerr to opine about: (1) the kindness-violence dichotomy present in most abusive intimate relationships; (2) the power-and-control dynamic at the core of domestic violence and, relatedly, the Power and Control Wheel that she used to help the jury visualize the substance of much of her testimony; (3) how the power-and-control dynamic often manifests itself in a domestic violence victim's counterintuitive behaviors; (4) the role that a substantial age difference in a couple can play in the power-and-control dynamic; and (5) how a domestic violence perpetrator can exert power and control over an intimate partner through pet abuse. To assess whether the court abused its discretion in allowing these opinions, we take a deeper dive into them. 1. The Kindness-Violence Dichotomy and the Power-and-Control Dynamic ¶ 59 Kerr started by discussing what social science has learned about domestic violence. She then addressed the kindness-violence dichotomy that exists in most abusive intimate relationships. She shared with the jury that perpetrators "are not abusive 100% of the time" and are often "quite kind and loving and charming and caring and ... really good partner[s]." Elaborating, she explained that they sometimes have "a Jekyll and Hyde personality"—they can be "really great [and] wonderful" and then have a "horrible ugly side"—which lures victims into thinking that the relationship is worth fighting for because there are "enough good things" to outweigh "the bad things. ¶ 60 According to Kerr, while everyone has good days and bad days, domestic violence abusers are driven by the need for "power and control." Kerr added that it's when abusers feel they are losing power and control that they use physical and nonphysical violence to regain the upper hand — "[t]hat's what drives their behavior." ¶ 61 Kerr opined that the dynamic of power and control is at the root of domestic violence. In the process of offering this testimony, she turned to the Power and Control Wheel for assistance. ¶ 62 The Power and Control Wheel is a "widely accepted tool for understanding the dynamics of domestic violence." Jane K. Stoever, Transforming Domestic Violence Representation, 101 Ky. L.J. 483, 511 (2013). It was developed by the Domestic Abuse Intervention Project in Duluth, Minnesota, based on descriptions that battered women shared about their abuse. Id. at 511-12. The location of the words "Power and Control" at the center of the tool is meant to convey that domestic violence "is based on systemic power and control issues," not "incidental issues such as drugs, alcohol, or the environment." Id. at 512. ¶ 63 While physical and sexual violence are conspicuously displayed on the upper and lower margins of the Power and Control Wheel, the instrument's eight spokes reflect a "range of abusive actions" beyond physical and sexual violence. Id. The spokes introduce categories of "interrelated dimensions of the abusive partner's exercise of power and control." Id. By highlighting insidious acts — such as intimidating the victim, using children, and controlling finances —which are more subtle and difficult to conceptualize as abuse than physical and sexual assault, the Power and Control Wheel "redefined domestic violence." Id. Today, the Power and Control Wheel is one of the most prevalent models used for domestic violence education: It is featured in some law school textbooks; it is employed in training judges, law enforcement personnel, attorneys, social workers, victim advocates, and other professionals; and it is used to provide domestic violence treatment, counsel battered women, and conduct community outreach. Id. at 512-13. ¶ 64 Neither the court of appeals nor our court has had occasion to review the admissibility of the Power and Control Wheel. However, a division of the court of appeals has upheld the admission of generalized expert testimony regarding what is considered to be the nucleus of the tool: the power-and-control dynamic at the crux of domestic violence. People v. Ruibal, 2015 COA 55, ¶¶ 11-14, 434 P.3d 606, 609-10, aff'd on other grounds, 2018 CO 93, 432 P.3d 590. And the U.S. Court of Appeals for the Eighth Circuit recently found that the Power and Control Wheel was helpful to the jury in a domestic violence case involving sexual abuse. See United States v. Johnson, 860 F.3d 1133, 1138-41 (8th Cir. 2017). ¶ 65 We conclude that the trial court did not abuse its discretion in finding that Kerr's opinions regarding (1) the kindness-violence dichotomy and (2) the power-and-control dynamic (presented with the aid of the Power and Control Wheel) satisfied the fit standard because the testimony's logical relation to the facts of this case was sufficient to be helpful to the jury without contravening CRE 403. The first opinion informed the jury that a kindness-violence paradox exists in most domestic violence cases.12 The second opinion set the stage for the overarching theme of her testimony: the dynamic of power and control as the defining characteristic of an abusive intimate relationship. Importantly, the opinion on power and control served as the underpinning for the rest of Kerr's testimony, which we explore next. 2. Counterintuitive Behaviors ¶ 66 At trial, the prosecution asserted that Cooper's power and control over L.K. helped explain why L.K. continued speaking with him and telling him she loved him after the incident. Additionally, the prosecution contended that the power-and-control dynamic in the relationship prompted L.K. to decline medical treatment immediately after the incident. Given the concern that the jury would view these behaviors as counterintuitive, draw inaccurate and unfair inferences about L.K., and ultimately discredit her testimony, the trial court permitted Kerr to educate the jury about such behaviors. ¶ 67 The trial court was spot on. Where, as here, certain behaviors by a domestic violence victim are in conflict with what an ordinary juror might intuitively expect, generalized expert testimony is helpful to educate the jury on what social science teaches about those behaviors. Cf Fasy, 829 P.2d at 1317-18 (reaching a similar conclusion with respect to post-traumatic stress disorder ); People v. Hampton, 746 P.2d 947, 951-52 (Colo. 1987) (reaching a similar conclusion with respect to rape trauma syndrome ), abrogated on other grounds by Shreck, 22 P.3d 68. ¶ 68 In Fasy , a sex assault on a child case, the People called an expert in child psychology to opine, among other things, that the victim's behaviors were consistent with a child suffering from post-traumatic stress disorder.13 829 P.2d at 1317. We determined that such testimony "clearly assisted the jury in understanding the victim's behavior after the incident." Id. The expert had "compared the characteristics of ... post-traumatic stress disorder with the victim's characteristics and concluded that she was suffering from the disorder." Id. at 1318. His testimony had thus helped explain to the jury "why the victim acted the way she did" after the incident. Id. at 1317. ¶ 69 In Hampton , another sex assault case, the People's expert opined regarding rape trauma syndrome. 746 P.2d at 948. We noted that this evidence had provided the jury with an explanation for the victim's delay in reporting the crime: Under these circumstances, such expert testimony assists the jury within the meaning of CRE 702 because it is helpful to the jury in determining what effect should be given to the victim's delay in reporting the crime. The lay notion of what behavior logically follows the experience of being raped may not be consistent with the actual behavior which social scientists have observed from studying rape victims. For many years, the law has assumed that a prompt report of a sexual assault renders it more likely that the crime was committed. Likewise, the failure to make a prompt report in certain cases is admissible as evidence that a sexual assault did not occur. Expert testimony that challenges or explains these assumptions is valuable information which the jury should hear and consider in its search for the truth. Id. at 952 (emphasis added) (citations omitted). ¶ 70 The rationale we espoused in Fasy and Hampton applies with equal force in this case. What ordinary jurors might intuitively expect about a victim's behaviors following a domestic violence incident may not be in line with what social scientists have discovered after years of research and study. ¶ 71 The fit of the expert testimony regarding counterintuitive behaviors was magnified here because, during the cross-examinations of L.K. and a responding officer, Cooper attacked the credibility of L.K.'s allegation regarding the severity of her injuries. He asked why, if her injuries were as serious as she claimed, she had failed to seek immediate medical attention. Kerr's expert testimony put Cooper's criticism in context. She told the jury that domestic violence victims often do not seek medical treatment in the defendant's presence for reasons that have nothing to do with their general credibility or the veracity of their allegations. As she elucidated, a victim may be: feeling scared (including as a result of a threat by the defendant); seeking to minimize what happened to show solidarity with the defendant; pursuing the option that will give her the best chance at safety; or attempting to prove to herself that she can handle what just occurred on her own. ¶ 72 This was "valuable information" that the jury was entitled to hear and consider as part of the truth-seeking process. See Hampton , 746 P.2d at 952 ; see also Relaford , ¶ 28, 409 P.3d at 496 (generalized expert testimony regarding " ‘the typical demeanor and behavioral traits displayed by a sexually abused child’ " is usually admissible in a child sexual assault case "because it assists the jury in understanding the victim's behavior after the incident —why the victim acted the way he or she did" (quoting People v. Mintz, 165 P.3d 829, 831 (Colo. App. 2007) )). Excluding it from the trial would have risked deceiving the jury and may have resulted in a miscarriage of justice. ¶ 73 Though we have never addressed the helpfulness of generalized expert testimony regarding behaviors by domestic violence victims, divisions of the court of appeals have repeatedly held that such testimony is useful to explain a victim's recantation. See People v. Wallin, 167 P.3d 183, 187 (Colo. App. 2007) ("In cases involving domestic violence, expert testimony concerning the reasons for victims' recantations is admissible."); People v. Johnson, 74 P.3d 349, 352-53 (Colo. App. 2002) (indicating that the victim had "changed her story" and thus the expert testimony regarding recantations, including a "list [of] several factors that can lead to such recantations," was admissible because it was "relevant to the victim's credibility, a key issue in the case"); People v. Lafferty, 9 P.3d 1132, 1135 (Colo. App. 1999) (noting that the expert's testimony concerning the cycle of violence and how it relates to recantation was admissible in light of the victim's testimony). ¶ 74 The principle behind these recantation cases applies with full vigor here. We see no reason to differentiate between generalized expert testimony related to recantation and generalized expert testimony related to other counterintuitive behaviors often exhibited by domestic violence victims. In both contexts, the expert testimony helps jurors understand why the victim engaged in behaviors that may seem unconventional, or even perplexing, to them. ¶ 75 The Eighth Circuit in United States v. Johnson affirmed the admission of generalized expert testimony about a domestic violence victim's counterintuitive behaviors. 860 F.3d at 1140-41. That court drew guidance from cases upholding the admission of expert testimony regarding the general characteristics that sexually abused children exhibit, including the "emotional and psychological traits ... that often account for behavior such as delay in reporting the abuse or failure to ‘escape’ the abusive situation." Id. at 1140 (quoting United States v. Johns , 15 F.3d 740, 743 (8th Cir. 1994) ). It then determined that, "[r]egardless of the victim's age, expert testimony about how individuals generally react to sexual abuse —such as not reporting the abuse and not attempting to escape from the abuser-helps jurors evaluate the alleged victim's behavior." Id. ; see also Harris v. State, 84 P.3d 731, 747 (Okla. Crim. App. 2004) (declining to find plain error in an expert's use of the Power and Control Wheel to illustrate "common behavioral patterns observed in abusive relationships"). ¶ 76 In short, we conclude that the trial court did not abuse its discretion in ruling that Kerr's generalized expert testimony about counterintuitive behaviors was admissible. This evidence had a sufficient logical connection to the facts of this case to be helpful to the jury while still passing muster under CRE 403. 3. Age Differential and Pet Abuse ¶ 77 We similarly perceive no abuse of discretion in the trial court's determinations regarding Kerr's opinions about a significant age difference in a couple and pet abuse. These opinions met the fit standard because their logical relation to the facts of this case was sufficient to be helpful to the jury without flouting CRE 403. ¶ 78 As to age differential, we note that Cooper is more than twice L.K.'s age. Kerr opined that such a difference in age can affect the power-and-control dynamic in an abusive intimate relationship. ¶ 79 As to pet abuse, we observe that L.K. testified that Cooper struck Buddy during the incident, and cruelty to pets is one of the examples of abusive behaviors listed under the intimidation spoke of the Power and Control Wheel. Kerr stated that "it can be a really intimidating gesture to take, you know, a little cat or dog ... something that the offender knows the victim really loves and cares for and hurt it." According to Kerr, such abuse "sends a really strong message" that the abuser is "dangerous" and the victim should be afraid: "Look at the lengths I'm willing to go to." Consequently, explained Kerr, if a domestic violence victim's pet has been threatened by her abuser, she may decline to be transported for immediate medical treatment after an assault, reasoning that "I need to stay here to make sure nothing horrible happens while I'm gone." Not only was there evidence in this case that Cooper abused L.K.'s pet—indeed, Cooper was charged with cruelty to animals — there was also evidence that when the paramedics arrived and offered medical treatment to both Cooper and L.K., L.K. declined treatment. D. The Division Erred ¶ 80 The division saw things differently. But its analysis cannot withstand scrutiny because it is premised on an unduly demanding and rigid fit standard. ¶ 81 In ruling that there was an insufficient fit between Kerr's generalized expert testimony and the facts of this case, the division commented that there was no evidence that Cooper had assaulted or otherwise abused L.K. before the charged incident. Cooper, ¶ 26, 490 P.3d at 425. However, proof of a prior act of domestic violence is not the sine qua non of the power-and-control dynamic inherent in abusive intimate relationships. Nor is it a condition precedent to satisfy the fit requirement when the People seek to present testimony like Kerr's in a domestic violence case. A single "act or threatened act" of violence in an intimate relationship suffices to satisfy the statutory definition of "domestic violence" in Colorado. § 18-6-800.3(1). There is no statutory requirement related to a pattern of violence. See id. Thus, contrary to the division's suggestion, generalized expert testimony on domestic violence is not automatically rendered inadmissible simply because the charged act is the first act of violence (physical or otherwise) in an intimate relationship. ¶ 82 The division nevertheless determined that the fit requirement was not met here because, in its view, almost all of the generalized expert testimony provided by Kerr had no logical relation to the factual issues. As we have demonstrated, however, using a more elastic fit standard paints a different picture. ¶ 83 There was evidence that Cooper and L.K. were involved in an intimate relationship and that Cooper, who was much older than L.K., exerted power and control over L.K. by physically and emotionally abusing her, as well as by physically abusing Buddy. And the People established that, given the facts of this case, it was necessary to educate the jury about the kindness-violence dichotomy and counterintuitive behaviors. ¶ 84 True, the division correctly pointed out that some aspects of Kerr's testimony — specifically with respect to multiple spokes in the Power and Control Wheel—had no logical connection to this case. By way of example, while discussing the Power and Control Wheel, Kerr testified that victims of domestic violence often stay in abusive relationships and endure additional abuse. Yet, no evidence supporting that factual scenario was presented to the jury. ¶ 85 It is almost inevitable, however, that some generalized expert testimony about the dynamics of abusive intimate relationships will fall short of matching a fact in a particular domestic violence case. Indeed, the point of generalized expert testimony is to educate the jury about generic concepts and principles without regard for the specific facts of the case. It follows that there will almost always be some testimony that has no logical relation to the facts of the case. If we embraced the division's standard, it would risk rendering generalized expert testimony largely, if not wholly, inadmissible. ¶ 86 Here, it was impossible for Kerr to discuss the Power and Control Wheel without referring to information that wasn't logically related to the facts of this case. The division's approach would have required the prosecutor to tell Kerr which spokes of the Power and Control Wheel matched the facts of this case. It then would have required Kerr to recast, on the fly, the Power and Control Wheel into something that social science has neither studied nor approved. And it ultimately would have required Kerr to share with the jury an incomplete (and arguably inaccurate and misleading) version of the Power and Control Wheel. ¶ 87 The division's analytical framework, in addition to transforming Kerr's testimony from generalized to case-specific, may well have made her modified opinions inadmissible under CRE 403. While it was feasible and proper for the trial court to limit Kerr's testimony on counterintuitive behaviors by excluding her opinions on recantation, minimization, and lack of cooperation, it would have been infeasible and improper to similarly restrict her testimony on the Power and Control Wheel by forcing her to testify about only some of its spokes. ¶ 88 We do not share the division's concern that the jury may have been misled into inferring "that historical facts existed based solely on [Kerr's] testimony" and that, therefore, the trial process may have been corrupted. Cooper, ¶ 21, 490 P.3d at 425. According to the division, any aspect of Kerr's testimony that didn't match one of the facts of this case —for example, her opinion that victims of domestic violence often stay in abusive relationships and endure additional abuse —was unfairly prejudicial to Cooper because it may have led the jury to incorrectly believe that such a fact existed. Id. at ¶¶ 33-44, 490 P.3d at 426-28. For five reasons, we are unpersuaded. ¶ 89 First, the division's reliance on Castillo v. People, 2018 CO 62, 421 P.3d 1141, is misplaced. There, we dealt with apples; here, we deal with oranges. The question we confronted in Castillo was whether an error in giving a jury instruction that was not supported by the evidence was harmless. Id. at ¶¶ 56-61, 421 P.3d at 1150-51. In answering in the negative, we cautioned that jurors may try to fit facts into an erroneously given instruction because they understandably assume that the instruction would not have been given if it had no application in the case. Id. at ¶¶ 58-59, 421 P.3d at 1150-51. Analogizing to Castillo , the division surmised that the jury may have forced the evidence to fit into Kerr's expert testimony. Cooper, ¶¶ 21-24, 490 P.3d at 425. But the danger we spoke of in Castillo had nothing to do with a jury's reliance on nonexistent historical facts; it dealt with the possibility of a jury taking facts actually introduced into evidence and pounding them into an erroneously provided instruction. Castillo, ¶¶ 56-67, 421 P.3d at 1150-52. There is a world of difference between assuming that facts exist and assuming that an instruction applies. ¶ 90 Second, the prosecutor made crystal clear that Kerr was simply sharing general background information about certain concepts and principles related to domestic violence. Kerr herself told the jury, in no uncertain terms, that she had no knowledge of any of the facts of this case, had not interviewed Cooper or L.K., and was not opining about the veracity of either version of events or Cooper's guilt or innocence. There is no basis in the record to believe that the jury was confused and unable to distinguish between the case-specific factual testimony provided by L.K., Cooper, and others, on the one hand, and Kerr's generalized expert testimony, on the other. ¶ 91 Third, Cooper had ample opportunity to address any concerns related to the jury being misled about nonexistent historical facts. He was free to conduct (and did conduct) a vigorous cross-examination of Kerr, and he had the option to present contrary evidence. ¶ 92 Fourth, at the end of the trial, the court instructed the jurors that they were not bound by Kerr's testimony and that it was up to them to decide whether to believe all of that testimony, some of it, or none of it. Thus, having been told that Kerr knew nothing about this case and was simply providing general background information about domestic violence, the jurors were then advised that, in any event, it was their prerogative to decide whether to accept or disregard her opinions in whole or in part. "We presume that a jury follows the trial court's instructions and would acquit ... if the prosecution did not prove all of the elements of the ... charge beyond a reasonable doubt." Galvan v. People, 2020 CO 82, ¶ 29, 476 P.3d 746, 755 (omissions in original) (quoting People v. Trujillo, 83 P.3d 642, 648 (Colo. 2004) ). Notably, though Kerr's generalized expert testimony related to all four charges, the jury convicted Cooper of two of them but acquitted him of the other two. If, as the division feared, Kerr's testimony was unfairly prejudicial to Cooper because it invited the jury to assume facts not in evidence, there presumably would have been four guilty verdicts. ¶ 93 Lastly, the division seemed to second-guess the trial court's finding that the probative value of Kerr's testimony was not substantially outweighed by the risk of confusing or misleading the jury. While we understand the need for trial courts to guard against the potential for expert evidence confusing or misleading the jury, we conclude that the division in this case ultimately failed to afford the trial court's CRE 403 determination the deference it deserved. When reviewing a ruling regarding the admission of expert testimony, our task as appellate courts is to check for an abuse of discretion, not to substitute our judgment for the trial court's. ¶ 94 One final point in the division's analysis bears discussion. The division considered the risk that Kerr's testimony improperly bolstered L.K.'s credibility. Cooper, ¶¶ 19-20, 490 P.3d at 424-25. Though this, too, is something trial courts must look out for, it does not change our decision. ¶ 95 Generalized expert testimony may not bolster the credibility of a victim by impermissibly implying that she is telling, or has previously told, the truth about the charged incident. See Venalonzo v. People, 2017 CO 9, ¶¶ 32-34, 388 P.3d 868, 877-78 ; CRE 608(a). CRE 608(a) prohibits the admission of testimony to establish that another witness testified truthfully or was truthful on a particular occasion prior to trial. See CRE 608(a) (providing that the credibility of a witness may be attacked or supported only by opinion or reputation evidence and only by referring to character for truthfulness or untruthfulness); see also People v. Wittrein, 221 P.3d 1076, 1081 (Colo. 2009) ("In Colorado, neither lay nor expert witnesses may give opinion testimony that another witness was telling the truth on a specific occasion."). ¶ 96 In Wittrein , we acknowledged that an expert may not offer generalized expert testimony that children usually do not fabricate sexual abuse allegations, as such testimony is "tantamount" to informing the jury that "the child victim was telling the truth about her allegations." 221 P.3d at 1082. Along the same lines, the Eighth Circuit warned in Johnson that an expert may not impermissibly vouch for the victim by, for example, diagnosing her with sexual abuse or opining that she was sexually abused, for such testimony invades the jury's exclusive responsibility to assess the credibility of the witnesses. 860 F.3d at 1140-41. ¶ 97 But here, Kerr's generalized expert testimony about domestic violence did not touch upon the truthfulness of L.K.'s version of events. Kerr did not opine about the veracity of L.K.'s testimony. Nor did she testify that L.K.'s post-incident behaviors demonstrated, or even suggested, that she had been abused by Cooper. Rather, Kerr's expert testimony educated the jurors regarding matters about which they likely lacked knowledge or insight, and any bolstering effect it may have had on L.K.'s credibility was incidental and acceptable. Cf. Relaford, ¶ 30, 409 P.3d at 496 (explaining that generalized expert testimony "may incidentally give rise to an inference that a victim is or is not telling the truth about the specific incident," but that "this fact alone is insufficient to deny admission of the evidence, because expert testimony generally tends to bolster or attack the credibility of another witness" (quoting People v. Koon, 724 P.2d 1367, 1370 (Colo. App. 1986) )). III. Conclusion ¶ 98 We reiterate that generalized expert testimony, like that provided by Kerr in this case, is not automatically admissible in domestic violence trials —the testimony must fit the facts of the case. Today we clarify that the fit requirement means that generalized expert testimony must have a sufficient logical connection to the facts of the case to be helpful to the jury while still clearing the ever-present CRE 403 bar. Prosecutors should take care to endorse generalized expert testimony about domestic violence only in appropriate cases; and, when they do so, they should endeavor to present only testimony that is logically connected to the factual issues. Trial courts, in turn, should exercise their discretion in deciding whether to permit all, some, or none of the proffered testimony under the fit standard we articulate today. In doing so, trial courts should consider the feasibility and propriety of admitting only a portion of the proposed generalized expert testimony on a particular subject. ¶ 99 Inasmuch as the division mistakenly concluded that the trial court abused its discretion in allowing Kerr to provide generalized expert testimony on domestic violence, we reverse. On remand, the division should consider the merit, if any, of Cooper's prosecutorial misconduct argument. I am authorized to state that JUSTICE HOOD and JUSTICE HART join in this dissent. JUSTICE GABRIEL, dissenting. ¶ 100 In this case, we granted the People's petition for certiorari to consider whether the division below erred in concluding that (1) generalized expert testimony on domestic violence must be limited to those facets of a subject that are specifically tied to the particular facts of the case; (2) generalized expert testimony on domestic violence was inadmissible because the charged act was the first act of domestic violence in the relationship; and (3) the admission of the expert testimony was not harmless. The first two issues, however, misstate what the division below actually did. Specifically, contrary to the People's assertions, with which the majority appears to agree, see maj. op. ¶¶ 4, 55, 80-81, the division concluded only that expert testimony, including generalized expert testimony, must "fit" the factual issues, as we defined "fit" in People v. Martinez, 74 P.3d 316, 323 (Colo. 2003), and that there was no such "fit" here, People v. Cooper, 2019 COA 21, ¶¶ 18-32, 490 P.3d 420, 424-26. ¶ 101 Because I believe, contrary to the majority's holding today, that the division's conclusions in this regard were correct, and because I also agree with the division that the error in admitting the expert's testimony was not harmless, I respectfully dissent. I. Factual Background ¶ 102 To avoid unnecessary repetition of the majority's factual recitation, I only briefly summarize the background facts that are pertinent to my analysis. I will, however, discuss at some length the expert testimony at issue because the details of that testimony are central to my disagreement with my colleagues. ¶ 103 This case did not involve a history of domestic violence between defendant Kerry Cooper and the victim. Rather, it appears to have arisen from a first incident of violence between the two. ¶ 104 Moreover, the case was something of a he said-she said scenario. The victim claimed that after asking Cooper to move a fan so that it was pointing at her and then requesting that he adjust it, Cooper pushed the fan, with its moving blades, into the victim's face. The victim further alleged that Cooper punched her in the face and body, grabbed her jaw, punched her dog when the dog came in, and threatened to and then did hit her in the head with a tire iron. The victim asserts that she attempted to defend herself by hitting Cooper in the head with a flashlight and biting down hard on his fingers. ¶ 105 For his part, Cooper claimed that after the victim complained about his placement of the fan, he threw the fan on the bed and told the victim to plug it in herself. He says that he then went back to sleep, but the victim proceeded to hit him in the head with a flashlight and then bit his finger hard when he tried to wrest the flashlight from her. Cooper denied pushing the fan into the victim's face or ever picking up a tire iron. ¶ 106 As a result of the foregoing events, Cooper was eventually charged with felony menacing (for threatening the victim with a tire iron), third degree assault (for pushing a fan into the victim's face and hitting her), harassment (for striking the victim with intent to harass, annoy, or alarm her), and cruelty to animals (for punching the dog in the midst of the altercation). ¶ 107 The case proceeded to trial, and as pertinent here, the prosecution indicated its intent to call as a general expert Janet Kerr, a licensed professional counselor who has worked principally in the fields of domestic violence and sexual assault and who had given expert testimony in these areas 159 times prior to the present case. Before the prosecution called Kerr to the stand, however, the trial court stated, "I believe we need to talk about Ms. Kerr because I have serious concerns about whether she will be able to testify." ¶ 108 At that point, the prosecution presented its arguments as to why Kerr's testimony was admissible. Among other things, the prosecution asserted that a defining characteristic of domestic violence is "power and control" and that in a domestic violence relationship, some of the victim's actions might be counterintuitive. For example, the victim might recant her allegations, continue speaking with her abuser after an assault, and even refuse medical treatment. The prosecution contended that all of these actions reflect the abuser's power and control over the victim because the victim thinks that if she reacts in a certain way, then it would be safer for her. The prosecution added that economic entrapment (i.e., when a victim feels that her economic position requires her to remain in an abusive relationship) and an age differential between a victim and her abuser also tend to show such power and control. And the prosecution posited that the existence of such factors tends to suggest an initial cycle in the so-called cycle of violence. ¶ 109 Cooper responded by moving to preclude Kerr's testimony. In support of this request, he argued that (1) this case involved "one isolated incident of alleged domestic violence"; (2) the facts of this case were insufficient to make Kerr's proffered expert testimony useful; and (3) Kerr's testimony was irrelevant and amounted to improper bolstering of the prosecution's power and control theory without a sufficient connection to the underlying facts. ¶ 110 After hearing the arguments of counsel, the court precluded Kerr's testimony, in part. Specifically, on the one hand, the court ruled that Kerr would not be permitted to discuss either recantation or economic entrapment because the evidence did not support either of those factors. On the other hand, the court ruled that it would allow Kerr to testify to power and control generally and to certain alleged facts in this case that the jurors might deem counterintuitive, including the victim's arguably refusing medical treatment, her continuing to speak with Cooper after the incident at issue, and the age differential between the victim and Cooper. In so ruling, the court again expressed its hesitancy regarding Kerr's proffered testimony, noting, "We don't have a cycle going on in this relationship because [the victim] left [Cooper] right after this abusive act." The court also expressed concern that the jury could misuse Kerr's testimony to "diagnose" whether an act of abuse occurred: "The Court is concerned that this testimony, if allowed to go too far and just allow her to say everything she usually says even when it doesn't uniquely fit this case, that it's a danger to misuse by the jury." ¶ 111 The prosecution then called Kerr to the stand, and after qualifying her as an expert in the area of domestic violence (specifically relating to matters of "power and control and issues that fall under that"), the prosecution proceeded to have Kerr testify regarding the so-called "Power and Control Wheel" that is at issue in this case. ¶ 112 Kerr began by describing the wheel as "one of the core standards that we use in the domestic violence field to talk about this issue." She explained that the wheel is "a visual that really describes what we talk about when we talk about how somebody can abuse a person in a relationship without ever physically touching them .... So these are the nonphysical ways that a person can do that." Kerr then went through each of these different nonphysical methods, which are depicted as spokes on the graphical wheel. These methods include (1) emotional abuse (e.g., name calling and put-downs); (2) financial or economic abuse (e.g., when a victim is financially dependent on a partner and a partner uses that as a way to control them); (3) isolation of a victim from the victim's sources of support; (4) minimizing or denying conduct while blaming the victim; (5) threatening and manipulating children to control the victim; (6) an attitude of male privilege, in which the offender behaves like "the king of the castle" who is "in charge" and makes the decisions; (7) coercion and threats (e.g., threatening harm if the victim calls the police); and (8) intimidation (e.g., taking out a gun and waving it around without ever actually making a specific threat to do anything with it). ¶ 113 Kerr next discussed certain conduct in which domestic violence victims often engage that may seem counterintuitive to lay people. In this regard, she noted that domestic violence victims frequently do not disclose what happened to them. She explained that this is because victims can be embarrassed, ashamed, and humiliated. In addition, Kerr opined that victims have often grown up in relationships in which abusive behavior was modeled for them. And she pointed out that victims frequently say that they truly love the person who abused them, and they therefore put up with the bad behavior. When the prosecution then asked how a person who is loving and kind can also be abusive, Kerr responded that "in this case, you have offenders who are driven by their need for this power and control." ¶ 114 The prosecution next asked how hurting a victim's animal plays into power and control, and Kerr responded that hurting an animal that the victim loves can be a really intimidating gesture and sends a strong message that "this person is dangerous and I should be afraid." ¶ 115 Last, the prosecution inquired why, after being physically assaulted, a victim might refuse treatment. Kerr replied that the victim might have been told that if she talked about the incident, then there would be a consequence. Or the victim might have wanted to show solidarity with the offender. Or "[i]t could be that the pet or a child or somebody else has been threatened and they think, I need to stay here to make sure nothing horrible happens while I'm gone." ¶ 116 Cooper's counsel then proceeded to cross-examine Kerr, and as pertinent here, Kerr clarified that she was not saying that every dispute between couples signifies domestic violence. Instead, she opined that "it becomes domestic violence when there's this patterned use of these tactics. And you don't have to have all of the tactics, but a patterned use of tactics that are designed to coerce, manipulate, control another person." ¶ 117 The prosecution picked up on many of these themes in its rebuttal closing argument (notably, the prosecution did not say much about Kerr's testimony during its initial closing argument). In particular, the prosecutor discussed Kerr's testimony about how difficult it is for abuse victims to share their experiences because they carry the shame. She further reiterated Kerr's testimony about how (1) family dynamics when one is growing up might impact future relations and (2) an offender can be like Jekyll and Hyde, nice one minute and the opposite the next. And she argued that power and control is "the defining characteristic in a domestic abuse relationship," noting, in this regard, that harm to pets, the victim's alleged refusal of medical treatment, and her alleged refusal to leave the relationship are all relevant to determining whether a domestic violence relationship exists. ¶ 118 The jury ultimately acquitted Cooper of felony menacing with a tire iron and cruelty to animals but convicted him of third degree assault and harassment. Cooper appealed, the division below reversed his conviction based on the impropriety of Kerr's testimony, Cooper , ¶¶ 5, 17–44, 490 P.3d at 423-28, and we granted certiorari. II. Analysis ¶ 119 I begin by explaining why I believe that Kerr's testimony was inadmissible in this case. I then address why I believe the error in admitting her testimony was not harmless. A. Generalized Expert Testimony and "Fit" ¶ 120 CRE 702 provides, "If 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." The focus of a CRE 702 inquiry is on whether the proffered evidence is both reliable and relevant. People v. Shreck, 22 P.3d 68, 77 (Colo. 2001). In determining whether the expert testimony at issue is reliable, "a trial court should consider (1) whether the scientific principles as to which the witness is testifying are reasonably reliable, and (2) whether the witness is qualified to opine on such matters." Id. In deciding whether the evidence is relevant, the court should consider whether the testimony would be useful to the jury. Id. Expert testimony is useful if it "will assist the fact finder to either understand other evidence or to determine a fact in issue." People v. Ramirez, 155 P.3d 371, 379 (Colo. 2007). ¶ 121 In Martinez, 74 P.3d at 323, we expanded on the concept of usefulness of expert testimony to a jury. We explained: Helpfulness to the jury hinges on whether the proffered testimony is relevant to the particular case: whether it "fits." Fit demands more than simple relevance; it requires that there be a logical relation between the proffered testimony and the factual issues involved in the litigation. That is, even if good grounds exist for the expert's opinion, it must be validly and scientifically related to the issues in the case. That particular expert testimony fits or is valid for one facet or purpose of a proceeding does not necessarily compel the conclusion that it fits all facets. Therefore, the admissibility of evidence must be evaluated in light of its offered purpose. Id. (citations omitted). ¶ 122 In my view, admitting expert testimony without the above-described fit presents at least two significant risks. First, such testimony can improperly bolster the credibility of one or more witnesses. See Venalonzo v. People, 2017 CO 9, ¶¶ 32-36, 388 P.3d 868, 877-78. Second, a witness's status as an expert can augment improper testimony with an aura of trustworthiness and reliability, thereby magnifying the harm caused by the error in admitting such testimony. See People v. Baker, 2021 CO 29, ¶ 41, 485 P.3d 1100, 1108 ; see also People v. Koon, 724 P.2d 1367, 1371 (Colo. App. 1986) (observing that a therapist's status as an expert witness "augmented her [improper] testimony with an aura of trustworthiness and reliability"). ¶ 123 Here, the majority cites the test set forth in Martinez with approval, but it proceeds to say that we did not analyze in that case how close a fit is required before generalized expert testimony will be admitted. Maj. op. ¶ 51. The majority then goes on to adopt a new test, namely, that generalized expert testimony fits a case if it has a sufficient logical connection to the factual issues to be helpful to the jury while still clearing the ever-present CRE 403 admissibility bar. Id. at ¶¶ 3, 52, 98. ¶ 124 I respectfully disagree with the majority's view that Martinez did not discuss how close a fit is required. Nor can I subscribe to the majority's new test, which strikes me as no test at all (and certainly less of a standard than that which we articulated in Martinez ). The majority's new test appears to be circular, defining helpfulness in terms of helpfulness, and then it subjects a trial court's analysis of the helpfulness question to a CRE 403 analysis. I do not see this as an improvement on what we said in Martinez — quite the contrary — and I think that we do better by sticking with our analysis in Martinez , which I believe has served its purpose well for the better part of two decades. ¶ 125 Applying that standard, I would conclude that Kerr's testimony was not useful to the jury because there simply was little to no logical relation between her testimony and the facts of this case. For example: • Kerr testified that the Power and Control Wheel shows how someone can abuse another without touching them, but the incident in this case involved only a physical confrontation. • Kerr testified regarding each of the examples of nonphysical abuse contained in the Power and Control Wheel (e.g., emotional and financial or economic abuse; isolation, minimizing or denying conduct while blaming the victim; threatening and manipulating children to control the victim; male privilege; and coercion, threats, and intimidation), but little to no evidence supported any of these kinds of alleged abuse. Indeed, based on the lack of evidence, the court expressly precluded Kerr from testifying regarding economic abuse, but the prosecution elicited such testimony anyway. • Kerr testified that domestic violence victims frequently do not disclose the abuse, but the victim here immediately did so. • Kerr testified that domestic violence victims often have had abusive behavior modeled for them in their family histories, but no evidence indicated that the victim here had this experience (nor does the fact, alleged by the prosecution, that the victim may not have been speaking to her father at the time of the incident at issue establish any such history). • Kerr testified that hurting animals as a form of intimidation tends to demonstrate power and control, but no evidence showed that this occurred here. Specifically, although some evidence indicated that Cooper may have hit the victim's dog when the dog came at h im in the middle of the parties' altercation, no evidence showed that Cooper engaged in such conduct in order to intimidate the victim. • Kerr testified that domestic violence exists when there is a patterned use of tactics designed to coerce, manipulate, and control, but no evidence supported any such patterned use of tactics like these. ¶ 126 In short, virtually none of Kerr's testimony regarding power and control had anything to do with the facts of this case. Instead, the prosecution's reliance on such testimony appears to have been the same type of reflexive reliance on expert testimony that we only recently warned against in Lawrence v. People, 2021 CO 28, ¶¶ 44-48, 486 P.3d 269, 278-79 (noting and warning against prosecutors' apparent practice of reflexively designating current or former securities commissioners or deputy commissioners as experts in securities fraud prosecutions). Specifically, it seems that prosecutors routinely seek to introduce expert testimony regarding the Power and Control Wheel in cases of alleged domestic violence, whether the factors comprising the wheel are relevant or not. I, however, do not believe that this is proper. Simply stated, the reflexive reliance on this type of testimony does not render the testimony admissible irrespective of whether it fits the facts of the case. ¶ 127 Accordingly, I would conclude that Kerr's testimony was not admissible under the standard that we established almost two decades ago in Martinez, 74 P.3d at 323, and that the majority's opinion, for reasons that I cannot discern, appears to undermine today. ¶ 128 In reaching this conclusion, I disagree with the majority's assertion that "it was impossible for Kerr to discuss the Power and Control Wheel without referring to information that wasn't logically related to the facts of this case." Maj. op. ¶ 86. This assertion assumes that the Power and Control Wheel has some special importance in cases like this, such that we must begin with the presumption that testimony regarding the wheel is admissible. I do not see why that is the case. If it fully fits, then it may be admissible. If it does not, then a trial court must determine what parts of it are relevant, just as with any other proffered evidence. ¶ 129 I likewise disagree with the People's assertions, with which the majority appears to agree, that the division below (1) improperly required that a general expert's testimony must be limited to the facets of a subject that are specifically tied to the facts of a case and (2) erroneously determined that such testimony must be narrowly tailored to "fit" the case at hand. In my view, the division did no such thing. Rather, it required nothing more than what we required in Martinez , namely, that under CRE 702, a proffered expert's testimony must be reliable and relevant, with those terms requiring an assessment as to whether the expert's opinions will be helpful to the fact finder. Cooper, ¶¶ 17–18, 25-32, 490 P.3d at 424-26. ¶ 130 Nor did the division state that Kerr's testimony was inadmissible because the charged act here was the first act of domestic violence in the parties' relationship. To the contrary, although the division pointed out the absence of any prior abusive conduct, it went on to describe how no record evidence supported the vast majority of the opinions that Kerr expressed. See id. at ¶¶ 26-31, 65 app., 490 P.3d at 425-26, 432-39 app. ¶ 131 And although I acknowledge that our review here is for an abuse of discretion and that deference is therefore due the trial court, I cannot agree that we must defer and refrain from concluding that the court abused its discretion merely because the court imposed some limits on Kerr's testimony. See maj. op. ¶¶ 4, 34, 56-58, 92-93. Although I applaud the trial court for its efforts in this regard, and although I acknowledge how difficult it is for trial courts to assess this kind of evidence in advance (whereas we appellate judges have the benefit of hindsight), here, where most of the general expert's testimony had no relation at all to the facts of this case, I feel compelled to conclude that the court abused its discretion in admitting this testimony. In this regard, my conclusion is consistent with the well-settled principle that review for an abuse of discretion does not allow an appellate court to abdicate its responsibility to ensure a fair trial. See Morgan v. People, 624 P.2d 1331, 1332 (Colo. 1981) ("The placing of [great discretion regarding jury selection] in the trial judge does not, however, permit appellate courts to abdicate their responsibility to ensure that the requirements of fairness are fulfilled."). ¶ 132 For these reasons, I believe that Kerr's testimony was inadmissible, and I proceed to the question of whether the admission of that testimony was harmless. B. Harmless Error ¶ 133 We review a preserved nonconstitutional trial error like the one at issue for harmless error. Hagos v. People, 2012 CO 63, ¶ 12, 288 P.3d 116, 119. Under this standard, we will reverse only when the error affects the substantial rights of a party, that is, "if the error ‘substantially influenced the verdict or affected the fairness of the trial proceedings.’ " Id. (quoting Tevlin v. People, 715 P.2d 338, 342 (Colo. 1986) ). ¶ 134 For several reasons, I cannot say that the error in admitting Kerr's testimony in this case was harmless. ¶ 135 First, the evidence here was not overwhelming. The victim and Cooper were the only witnesses to the altercation at issue, and they vigorously disputed who initiated that altercation. Accordingly, at some level, this case came down to a credibility contest between the victim and Cooper. ¶ 136 Second, and related to my first point, I believe that Kerr's testimony substantially undermined Cooper's credibility and, conversely, tended to bolster the victim's credibility. Specifically, if the victim and Cooper were in the type of abusive relationship that Kerr described — and human nature suggests that the jury may well have drawn such an inference from her testimony — then the jury was far more likely to conclude that Cooper committed the acts alleged here. ¶ 137 Third, Kerr's testimony painted a picture of alleged domestic violence offenders as power-obsessed actors who will harm children and pets to intimidate the victim, and of alleged victims as people who have gotten caught in abusive relationships due to a history of violence in their families and fear of escape. The prosecutor then relied on this testimony to suggest that Cooper and the victim fit these profiles in numerous respects, even though the evidence in this case supported none of those arguments. In my view, the unfair prejudice resulting from such unfounded caricatures was manifest. ¶ 138 Finally, the foregoing facts lead me to conclude that Kerr's testimony likely did precisely what the trial court feared it would do: it gave the jury the means to misuse Kerr's testimony and to "diagnose" whether an act of abuse occurred, even though the facts of this case did not support employing Kerr's testimony for that purpose. Indeed, Kerr's testimony appears to have served no purpose other than to allow the prosecution to try to squeeze the facts of this case into the Power and Control Wheel in order to paint a picture of Cooper that found little to no support in the evidence presented. ¶ 139 Accordingly, I am compelled to conclude that Kerr's inadmissible testimony substantially influenced the jury's verdict and that, therefore, the error in admitting this testimony was not harmless. III. Conclusion ¶ 140 For these reasons, I believe that Kerr's testimony was inadmissible, the error in admitting her testimony was not harmless, and Cooper is thus entitled to a new trial. ¶ 141 I would therefore affirm the judgment of the division below, and, accordingly, I respectfully dissent. 1 Perhaps taking a cue from the case law, the parties use the term "blind expert testimony." We prefer "generalized expert testimony." 2 Though L.K. was cooperative with the People at trial and did not recant or minimize her prior statements to the police, some of her testimony was inconsistent with those earlier statements. For example, she told the police that Cooper had not pressed the blades of the fan into her face. Additionally, not all of the details in L.K.'s testimony were included in her statements to the police. For instance, when she spoke with the police, she did not mention that Cooper had used the tire iron as a weapon. 3 In a pretrial statement, L.K. indicated that she had also suffered broken ribs. But she didn't testify about an injury to her ribs at trial. 4 A person convicted of a crime, the underlying factual basis of which has been found to be an act of domestic violence, is subject to certain consequences. See § 18-6-801, C.R.S. (2020). 5 This graphic of the Power and Control Wheel is for illustrative purposes only. Though not a replica of the exhibit used at trial, it is substantively identical to it. 6 Cooper was in custody during all but one month of the pretrial timeframe. 7 This excluded testimony should not be confused with Kerr's brief testimony related to "economic abuse" —one of the spokes of the Power and Control Wheel. The latter relates to "preventing [a victim] from getting a job or keeping a job, making her ask for money, giving her an allowance, taking her money, [and] not letting her know about or have access to family income." 8 Whether the prosecution's questions or Kerr's answers at any point crossed the limitations set by the court is not an issue before us. 9 We need not pass judgment on the ruling excluding evidence of the cycle of violence. We note, however, that studies have explained that the cycle of violence refers to three distinct phases in an abusive intimate relationship: a tension-building phase, a violence phase, and a contrition phase. See People v. Yaklich, 833 P.2d 758, 761 (Colo. App. 1991) ; see also Hernandez v. Ashcroft, 345 F.3d 824, 836 (9th Cir. 2003) ("Abuse within intimate relationships often follows a pattern known as the cycle of violence, ‘which consists of a tension building phase, followed by acute battering of the victim, and finally by a contrite phase where the batterer's use of promises and gifts increases the battered woman's hope that violence has occurred for the last time.’ " (quoting Mary Ann Dutton, Understanding Women's Responses to Domestic Violence: A Redefinition of Battered Woman Syndrome, 21 Hofstra L. Rev. 1191, 1208 (1993) )). 10 Here are the issues we agreed to review: 1. Whether the court of appeals erred in concluding that blind expert testimony on domestic violence was inadmissible because the charged act was the first act of domestic violence in the relationship. 2. Whether the court of appeals erred in concluding that blind expert testimony on domestic violence must be limited to those facets of a subject that are specifically tied to the particular facts of the case. 3. Whether the court of appeals erred in finding that the admission of the expert testimony was not harmless. 11 As mentioned, we do not address the trial court's ruling prohibiting the prosecutor from eliciting an opinion on the cycle of violence. 12 Today's decision should not be understood as permitting generalized expert testimony on domestic violence whenever there is evidence that a defendant has been both kind and violent. As we make clear, this wasn't the only fact on which the trial court relied in finding that Kerr's testimony fit this case. 13 This was not all generalized expert testimony because the expert was the victim's treating psychologist. Fasy, 829 P.2d at 1315. However, the testimony on which we rely here was generalized expert testimony.