Justice MARTINEZ
delivered the Opinion of the Court.
A jury convicted Stephen Martinez of murder in the first degree for knowingly killing Heather Mares, a child under the age of twelve in his care. Martinez violently shook four month-old Heather and slammed her head on the edge of her crib, resulting in her death. Two experts for the prosecution, Dr. Ann Martin and Dr. Patty Rosquist, explained Heather died from blunt force trauma to the head resulting from shaken-impact syndrome
-a vigorous shaking and sudden deceleration of the brain as it strikes an object. The syndrome is characterized by a constellation of injuries, including a subdural hematoma, a collection of blood between the brain and skull. Both experts testified that massive, violent force causes subdural hema-tomas, and the defense did not contest that testimony.
However, over the defense's objections on the basis of relevance, Dr. Rosquist further testified that babies only suffer subdural he-matomas in limited situations such as shaken-impact syndrome or in specific accident scenarios: falls from several-story buildings or in high-speed automobile accidents. The prosecution introduced the accident seenario evidence for two purposes: (1) to give a frame of reference to the impact required to cause a subdural hematoma and; (2) to provide a basis of the expert opinion that massive, violent force causes a subdural he-matoma. In rebuttal closing argument, the prosecution, citing Dr. Rosquist's testimony, claimed that the force Martinez used in shaking and slamming Heather was equivalent to that of the accident seenarios.
The court of appeals reversed Martinez's conviction because it found the disputed evi-denee inadmissible as it was not probative of the minimum foree Martinez may have used for purposes of proving that he knowingly killed Heather. People v. Martinez, 51 P.3d 1046, 1051 (Colo.App.2001). The court of appeals did not evaluate whether the evidence was admissible as the basis of Dr. Rosquist's opinion that subdural hematomas result from massive, violent force.
We granted certiorari to consider the prosecution's argument that the court of appeals misapplied this court's decision in People v. Shreck, 22 P.3d 68 (Colo.2001): Dr. Rosquist need not describe the minimum force necessary to cause subdural hematomas in shaken-impact syndrome for testimony regarding accident scenarios to be admissible for the purpose of proving Martinez's mens rea.
We agree that the standard of admissibility for CRE 702 is reliability and relevance, not certainty.
We evaluate the admissibility of accident scenario evidence under CRE 702 and CRE 403 and in the context for which it is offered. Like the court of appeals, we also find Dr.
Rosquist's accident seenario testimony irrelevant to the question of Martinez's mens rea, but not for the reason the court of appeals articulated; rather we find it inadmissible as a frame of reference because Dr. Rosquist did not show how the accident seenarios relate to shaken-impact syndrome.
Nevertheless, the testimony of accident scenarios is admissible under CRE 702 and CRE 403 as the basis of Dr. Rosquist's opinion that a subdural hematoma results from massive, violent force. Assuming that Dr. Rosquist was qualified and her opinion was reliable, we find her testimony was helpful to the jury's understanding of subdural hemato-mas. Finally, we do not find an abuse of discretion in the trial court's weighing of the probative and prejudicial value of the evidence in dispute. Accordingly, we reverse the court of appeals.
I. FACTS AND PROCEDURE
A jury convicted Stephen Martinez of murder in the first degree for knowingly causing the death of a child who had not yet attained twelve years of age and the person committing the offense was in a position of trust with respect to the child, pursuant to section 18-6-401(7)(c), 6 C.R.S. (2002). While under Martinez's care, Heather Lynn Mares, a four-month old baby, suffered injuries that ultimately led to her death. Martinez is not Heather's father, but was the live-in partner of Heather's mother, Kim Velverde.
In October 1998, while Velverde left Heather with Martinez for fifteen minutes to run a quick errand, Heather began to ery. In response to her incessant crying, Martinez shook Heather hard, and while shaking her, hit the back of her head on her crib. Vel-verde returned to find Heather being carried to an ambulance parked outside of her home. She was limp and purple. Later that evening, Heather died. Her injuries included a complex skull fracture and two linear bruises on the back of her head; subarachnoid bleeding, bleeding on the surface of the brain; a subdural hematoma, a localized three-dimensional mass of blood between the brain and the skull; and bilateral retinal hemorrhages.
Martinez maintained, in a 911 emergency phone call, in a written statement to an investigating officer, and in a videotaped interrogation, that Heather had suddenly started to choke. In the videotaped interrogation, Martinez explained that Heather's head injuries could have resulted from an accident two weeks earlier in which both he and Heather had fallen together onto the kitchen floor after he tripped on a telephone cord. Prodding by the interrogating officer led Martinez to finally confess that he had shaken Heather hard and hit her head on her crib. In demonstrating to the interrogating officer how he shook Heather, using a doll, Martinez depicted only a light shaking. He also stated that he had never heard of shaken baby syndrome, but responded affirmatively to the interrogating officer's questions, "Do you think it's dangerous to shake a little tiny baby? ... So your action was dangerous in and of itself?"
A. The Jury Trial
At trial, the only material fact at issue was Martinez's - culpability, and - specifically, whether he shook Heather knowing that his actions were practically certain to cause her death. See § 18-1-501(6), 6 C.R.S. (2002) ("A person acts 'knowingly' ... with respect to a result of his conduct, when he is aware that his conduct is practically certain to cause the result."). The defense strategy focused on the fragility of babies, that Martinez only shook Heather with minimal force, and that Heather's head accidentally hit the crib during the shaking. Since only minimal force was used and because babies are fragile, the defense maintained it was impossible for Martinez to know that his actions could lead to her death.
In contrast, the prosecution argued that Martinez, in shaking and slamming Heather, used such massive and violent force that he was practically certain his conduct would cause Heather's death. The prosecution sought to admit evidence on the amount of force used to cause shaken-impact syndrome through the testimony of a deputy coroner, Dr. Ann Martin, who performed Heather's autopsy and the testimony of a pediatrician, Dr. Patty Rosquist.
Both experts demonstrated using a doll the biomechanics and amount of foree required in shaken-impact syndromes.
In comparison to the shaking and impact that Martinez demonstrated on his videotaped interrogation, Dr. Martin found his demonstration "not sufficient ... not close" to the conduct required to cause Heather's injuries. Dr. Rosquist similarly stated that Martinez's demonstration was inconsistent with the amount of force required to cause the constellation of injuries that Heather suffered.
Both experts also testified that a subdural hematoma, one of the injuries that Heather incurred, requires massive, violent force. The defense did not contest that assertion. When the prosecutor sought further explanation of "massive, violent force" through other real-life accident seenarios, the defense objected on the basis of CRE 403 stating that it was misleading and irrelevant. The prosecutor argued that the evidence was offered to provide the jury with a frame of reference regarding the force Martinez used: It would show jurors "the sort of impact we're talking about." The trial court sustained the defendant's objection based solely on the form of the prosecutor's question, but stated that it would "let the prosecution go into the area of how much force is required to cause this injury." The prosecution was unsuccessful in admitting accident scenario evidence through Dr. Martin.
The prosecution again attempted to introduce accident scenario testimony through Dr. Rosquist. The defense objected again, and the prosecution argued that the evidence served as the basis for the doctor's opinion that subdural hematomas result from massive, violent force and that it provides the jury with a frame of reference. The trial court ultimately allowed the testimony.
Thus, Dr. Rosquist testified that a subdural hematoma rarely occurs and does so only in limited situations: where a baby falls from a several story building; in a high-speed motor vehicle accident either as a pedestrian or where the baby is unrestrained; or when a baby is violently shaken or violently shaken and slammed. However, Dr. Rosquist also testified that the amount of force necessary to cause a subdural hematoma via shaking-impact syndrome is unknown, because obviously no such experiments on babies can be conducted and perpetrators are less than honest about the force employed.
DR. ROSQUIST: In talking about how much force that it requires to cause this kind of injury, we can't take babies and shake them and do that for the obvious reason, so what we have to do is take-there have been multiple studies of series of witnessed falls or witnessed accidents where babies had similar kinds of injuries. Okay? And in those studies, babies who had similar kinds of injuries, the subdural hematoma, have been things like a fall from a multiple story building. Being in a high speed motor vehicle accident either as a pedestrian hit by a high speed motor vehicle or, for example, an unrestrained passenger in a high speed motor vehicle accident, so those are the kinds of witnessed injuries that can lead to a similar sort of injury.
Notably absent from Dr. Rosquist's testimony is whether Martinez did, or did not, use force equivalent to the described accident scenarios. Furthermore, the limit of the doctor's testimony is apparent: other than the similarity of one injury-subdural hemato-ma-she did not link, via biomechanics or the amount of foree required, shaken-impact syndrome to the accident seenarios.
In rebuttal closing argument, the prosecution concluded from Dr. Rosquist's testimony that Martinez used force equivalent to that described in the accident seenarios.
The
jury was offered instructions on first degree, knowing murder of a child by a person in a position of trust; child abuse resulting in death (reckless); child abuse resulting in death (negligence); manslaughter; and criminally negligent homicide. The jury returned a verdict of murder in the first degree, finding Martinez, in a position of trust, had knowingly killed four-month old Heather Mares.
B. The Appeal
Martinez appealed to the court of appeals, asserting in part that the district court violated Martinez's constitutional right to a fair trial by admitting expert testimony of accident scenarios as proof that Martinez used an equivalent degree of force. The court of appeals agreed, applying the test we announced in People v. Shreck, 22 P.3d 68 (Colo.2001) to determine whether the accident scenarios were admissible, and reversed the conviction. People v. Martinez, 51 P.3d 1046, 1048, 1050 (Colo.App.2001).
In Shreck, we announced the seope of the inquiry of admissibility of expert testimony on scientific evidence. Specifically, we described CRE 702's reliability and relevance prongs. Id. at T7. Expert scientific testimony is reliable if (1) the scientific principles are reliable; and (2) if the witness is qualified to opine on those scientific principles. Id. As to relevance, a trial court must evaluate whether the testimony would be helpful to the jury. Id. The liberal approach in admitting evidence is tempered by CRE 408 which gives a court discretion to exclude expert testimony where its probative value would be substantially outweighed by unfair prejudice. Id. at 79.
In this case, the court of appeals found that the accident scenarios did not meet the relevancy prong and additionally that it must be excluded under CRE 408 because, as to the announced purpose, to establish the force necessary to cause a subdural hematoma, it was highly prejudicial and likely to confuse or mislead the jury. Martinez, 51 P.3d at 1050.
The court of appeals used a syllogism to establish that through the use of these observed extreme accident seenarios, the prosecution had ignored the undistributed middle:
"(ilt is one thing clearly to state that a
certain quantum of force is necessary to produce a subdural hematoma; it is quite another to use examples of obviously extreme forces and violence that have been demonstrated to have caused subdural hematomas and then suggest that they constitute the minimum force necessary to cause such an injury in any particular case." Id. at 1051. Because reasonable inferences could not be drawn, the court of appeals stated that the observed accident scenarios testimony was not probative on the issue of the foree Martinez used and consequently the mens rea he possessed while shaking Heather. Id.
Combined with the prosecutor's allusion to the expert's testimony in rebuttal closing argument, the court of appeals held that the admission of the evidence was not harmless, reversed the conviction and remanded for a new trial Id. In summary, the court of appeals held that the accident scenario evidence did not help the jury understand the minimum force Martinez may have used to cause a subdural hematoma, and therefore, its unfair prejudicial value substantially outweighed its probative value.
II. ADMISSIBILITY OF DR. ROS-QUIST'S TESTIMONY OF OBSERVED ACCIDENT SCENARIOS
The prosecution states that the court of appeals misapplied the Shreck test for admissibility of expert evidence when it deemed Dr. Rosquist's testimony of observed accident scenarios irrelevant and lacking any probative value. We agree that the standard for admissibility is relevance and reliability, not certainty. In particular, we would find that Dr. Rosquist need not testify to the minimal force required to cause a subdural hematoma for observed accident scenarios to be admissible as the basis of her opinion that subdural hematomas result from massive, violent force. Therefore, because the disputed testimony satisfies CRE 702 and CRE 403, we find that the trial court did not abuse its discretion in admitting Dr. Rosquist's testimony.
A. Standard of Review
Trial courts are vested with broad discretion to determine the admissibility of expert testimony, and the exercise of that discretion will not be overturned unless manifestly erroneous. Masters v. People, 58 P.3d 979, 988 (Colo.2002). A trial court has a superior opportunity to determine the competence of the expert as well as assess whether the expert's opinion will be helpful to the jury. Id. However, we have channeled a trial court's discretion through the test announced in Shreck: (1) the scientific principles underlying the testimony must be reasonably reliable; (2) the expert must be qualified to opine on such matters; (8) the expert testimony must be helpful to the jury; and (4) the evidence must satisfy CRE 403. Shreck, 22 P.3d at 77, 79. Although the trial court here did not have the guidance of Shreck because that opinion was released after the jury trial in this case, we review the admissibility of Dr. Rosquist's testimony pursuant to Shreck,. Specifically, we examine whether the trial court abused its discretion under either CRE 702 or CRE 403.
B. Colorado Rules of Evidence Germane to Inquiry on Admission of Scientific Expert Testimony: CRE 401, CRE 402, CRE 403 & CRE 702
"[UJInless 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, in the logical sense, as long as it is 'evidence having 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). However, relevant evidence must be excluded, if "its probative value is substantially outweighed by the danger of unfair prejudice." Id. (quoting CRE 403). Thus, while CRE 401 and CRE 402 reflect liberal admission of evidence, CRE 403, in conjunction with CRE 702, tempers broad admissibility by giving courts discretion to exclude expert testimony unless it passes
more stringent standards of reliability and relevance.
CRE 702 and CRE 403 govern the admissibility of expert testimony. Masters v. People, 58 P.3d 979, 988 (Colo.2002) (citing People v. Shreck, 22 P.3d 68, 70 (Colo.2001)). CRE 702 addresses relevancy and reliability issues and states:
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.
In applying Shreck, we assume, as it is not in dispute, that the scientific principles of shaken-impact syndrome and subdural hema-tomas resulting from extreme accidents are reasonably reliable. Further, we assume that Dr. Rosquist, as an expert was qualified to opine on such matters. Given the above assumptions, we address whether the expert testimony is helpful to the jury in either understanding other evidence in the case or helping the jury resolve a factual issue in this case. Finally, we discuss whether in spite of the helpful nature of the testimony, the testimony should nevertheless be excluded because the evidence's probative value is substantially outweighed by unfair prejudice completing the CRE 702 and CRE 403 analysis.
C. Whether the Evidence Was Useful to the Jury
The testimony at issue here is Dr. Rosquist's description of accident seenar-ios which result in subdural hematomas, an injury that also occurs in shaken-impact syndrome. 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. In re Paoli R.R. Yard PCB Litigation, 35 F.3d 717, 743, 745 & n. 13 (3d. Cir.1994). 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. Id. 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. People v. Saiz, 82 P.3d 441 (Colo.2001) (abuse of the court's discretion in determining admissibility of evidence considered in light of its offered purpose).
The prosecution's proffered reasons during trial for admitting the testimony were (1) to provide a frame of reference for the jury for the impact required to cause a subdural he-matoma; and (2) to show the basis of the expert's opinion that a subdural hematoma is only caused by massive, violent force. We review the testimony with reference to the purposes for which it was offered to the trial court.
1. Accident Testimony as Basis of Dr. Rosquist's Opinion
We address the second proffered reason first because we would find the accident scenarios admissible for that purpose.
At trial, the defense did not contest the testimony that subdural hemato-mas result from massive, violent force. The accident scenario evidence supported Dr. Rosquist's expert opinion that subdural hematomas result from such force. The basis is sound because it is based on empirical data, and under CRE 702,
the ex
pository information assists the jury in better understanding the nature of a subdural hematoma: that it cccurs only in a few known situations. Similarly, the jury understands that a simple fall is unlikely to cause a subdural hematoma.
Generally, the factual basis of an expert opinion is admissible, and cross-examination should reveal the weaknesses in its underpinnings. Here, Dr. Rosquist conceded the limitation of her opinion that she did not know the amount of force required to cause a subdural hematoma in shaken-impact syndrome. Defense counsel, in cross-examining Dr. Rosquist, compelled her to clarify that one could not quantify the amount of shakes required to cause a subdural hematoma and that all sorts of body types and strengths could cause this result. At no time did Dr. Rosquist pronounce that Martinez used force equivalent to that described in the accident scenarios.
Because Heather suffered a subdural he-matoma, and because it is undisputed that massive, violent force causes it, the expert's testimony as to basis of her opinion helped the jury understand the facts of this case. Therefore, we find that Dr. Rosquist's testimony satisfies CRE 702 for the purpose of serving as the basis of her opinion.
2. Accident Scenarios as Frame of Reference
A frame of reference is a portal through which the evidence in a case must be viewed and understood. As such it delimits the boundaries of testimony, and is a much narrower concept than the basis of an expert's opinion.
The validity and fit of a framework must be examined under the province of CRE 702. Generally, frames of reference are used in the context of social frameworks, such as battered women's syndrome, to help a jury resolve other issues in the case such as "due care, intent, or purpose, and who likely did what and why." Christopher B. Mueller and Laird C. Kirkpatrick, Federal Evidence § 350 (1994). Therefore, frameworks require that a jury make an inference from the evidence presented and fit it to the facts at hand. We recently discussed framework evidence in Masters v. People, 58 P.3d 979 (Colo.2002), where the expert, Dr. Meloy, testified about characteristics he regarded as consistent with sexual homicides.
In this case, the prosecution attempted to use a frame of reference for purposes of showing the amount of force that Martinez may have employed. However, unlike Masters, other than the mere similarity in a subdural hematoma, we cannot discern a valid framework in which the characteristics of accident scenarios are consistent with shaken-impact syndrome. Simply put, before we limit a jury's understanding of the force required and biomechanies resulting in subdural hematomas in shaken-impact syndrome to that occurring in the accident scenarios, a jury needs the relationship between shaken-impact syndrome and accident scenarios.
Dr. Rosquist did not offer any links or a fit between the accident seenarios and shaken-impact syndrome. Indeed, she stated that the amount of force used in shaken-impact syndrome to cause a subdural hematoma is unknown; nor could she state how many shakes are necessary as she conceded on cross-examination. As such, she could not show the overlap of the subsets of mechanisms which cause subdural hematomas.
Although it is unreasonable to conclude that the link between the force in shaken baby cases and accident seenarios must be known to a certainty, Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 590, 113 S.Ct. 2786, 125 LEd.2d 469 (1993), we require some link in addition to the same injury before we will allow evidence to be presented to the jury. Id. at 591, 113 S.Ct. 2786 ("(absent credible evidence [of fered to the jury] supporting such a link) evidence that the moon was full on a certain night will not assist the trier of fact in determining whether an individual was unusually likely to have behaved irrationally on that night"). Finally, the prosecution sought to introduce accident scenario evidence as a
frame of reference because babies cannot be shaken in a test scenario; but the paucity of scientific studies, empirical or otherwise, is not a sufficient basis to admit evidence that has no relevance for the purpose it is offered.
Because Rule 702's helpfulness standard requires a valid scientific connection, annun-ciated to the jury, without more foundation, we cannot sanction such a frame of reference. Although this purpose is not a valid one for which the evidence may be admitted, the evidence is admissible nonetheless to provide a basis of Dr. Rosquist's opinion as explained above. We perform a CRE 4083 analysis only for the admissibility of accident scenarios as the basis of the doctor's testimony.
D. CRE 403
CRE 403 favors the admissibility of evidence, but evidence should be exeluded when it is unfairly prejudicial, when it has an undue tendency to suggest a decision on an improper basis. People v. District Court, 869 P.2d 1281 (Colo.1994). For the purpose of serving as a basis of the doctor's testimony, the images of babies falling from several story buildings or of unrestrained children in automobile accidents are shocking and could leave an indelible image in the minds of the jury. Although we find the testimony is helpful to the jury, its prejudicial nature must be weighed against its probative force. Other courts have admitted this evidence in passing and without comment for various purposes,
and we cannot find that the trial court abused its discretion in admitting Dr. Rosquist's testimony.
III. EVIDENCE OF ACCIDENT SCENARIOS TO ASSIST IN A MENS REA DETERMINATION IS NOT DIS-POSITIVE OF MENS REA
The prosecution and amici have expounded on the necessity of accident scenarios to establish a defendant's mens rea: a perpetrator alone with a baby when it is shaken and slammed will likely not admit to the amount of force used. However, necessity is not the proper basis for assessing the admissibility of evidence; rather the evidence must satisfy CRE 702 and CRE 408.
Additionally, we are skeptical that the introduction of accident scenarios, alone, can be used to establish the mens rea of knowing conduct. - Indeed, such evidence alone is neither direct nor cireumstantial evidence of the amount of force that a particular defendant may have used. Evidence suggests that many parents and care-givers are unaware of the dangers of shaking a baby.
However, intent may be shown by cireum-stantial evidence such as the serious nature of the injuries; admissions by the perpetrator; a failure to seek medical care; indicators of prior injuries suggesting repeated abuse; the brutality of the attack; the disparity of the size and strength between the victim and the perpetrator; the defendant's lack of remorse and efforts to avoid detection; the concealment of the victim's body; and finally motive. Charles A. Phipps, Responding to Child Homicide: A Statutory Proposal, 89 J.Crim. L. & Criminology 535, 575 (1999). Thus, we caution that to rely solely on accident scenario evidence to prove intent may in some cases fail to meet the constitutional burden of proof beyond a reasonable doubt.
IV. REBUTTAL CLOSING ARGUMENT
We briefly address the effect of the prosecution's rebuttal closing argument on the admissibility of Dr. Rosquist's evidence in the first instance. The prosecutor urged the jury to use Dr. Rosquist's testimony as a
frame of reference, from which it could be reasonably inferred that Martinez used force equivalent to that described in the accident scenarios.
Although we would find that the prosecutor's inference may not necessarily follow from the expert's testimony-we do not agree with the prosecution's determination of quantum of force necessary to cause a sub-dural hematoma when the expert stated that the same is unknown-it does not control our decision that the trial court did not abuse its discretion in admitting Dr. Rosquist's testimony. The testimony was admissible for the purposes of showing the basis of the doctor's opinion and, therefore, an improper inference in closing argument does not make that evi-denee inadmissible after the fact.
v. CONCLUSION
A trial court admitted the testimony of an expert, Dr. Rosquist, describing horrific accident scenarios, to explain to a jury that subdural hematomas result from massive, violent force. The court of appeals reversed the conviction, reasoning that inadmissible expert testimony was used to establish the central question in this case of whether Martinez knowingly caused the death of a child by a person in a position of trust. Although we agree that the use of accident seenarios in this case to prove mens rea was improper, we find that the trial court did not abuse its discretion in admitting this testimony because it was helpful to the jury as the basis of the expert's opinion. Accordingly we reverse the judgment of the court of appeals.
Justice COATS does not participate.