The appellant, Dona R. Zimmerman, alleged that she was injured in an automobile collision caused by the negligence of the appellee, Nedra J. Powell. Over Zimmerman’s objection, the court allowed Powell’s accident reconstructionist, Jubal D. Hamemik, Ph.D., to testify that Zimmerman was driving above the speed limit when the vehicles collided. The jury determined that Zimmerman had suffered $17,851.18 in damages, but also determined that she bore 49 percent of the responsibility for the collision and reduced the damages to $9,104.10.
*424
Under Neb. Rev. Stat. § 27-702 (Reissue 1995), the trial court must act as a gatekeeper to ensure the evidentiary relevance and reliability of an expert’s opinion.
Carlson
v.
Okerstrom,
FACTUAL BACKGROUND
The collision occurred at an intersection in Scottsbluff, Nebraska. A yield sign at the intersection required east-west traffic to yield to north-south traffic. As Powell approached the intersection from the west, Zimmerman approached from the south. Although Powell claims that she looked to the south before entering the intersection, she admitted that she did not see Zimmerman. Powell proceeded into the intersection without yielding. Zimmerman slammed on her brakes, but was unable to stop. She stmck Powell’s vehicle in the center of the passenger’s side.
The main focus of this appeal is whether Zimmerman was driving above the 25-m.p.h. speed limit before braking. According to Zimmerman, she approached the intersection at 15 to 20 m.p.h. She testified that before entering the intersection she slowed down and looked both ways. She noticed Powell’s vehicle approaching from the left. Zimmerman claims that she could tell that Powell had not seen her and that Powell was not slowing down. Zimmerman then slammed on her brakes, leaving 20 feet of tread marks.
To contradict Zimmerman’s claim that she was traveling below the speed limit, Powell relied on Hamemik’s testimony. Hamemik has a master’s degree in civil engineering from the University of Connecticut and a doctorate in computational mechanics from the University of Colorado. He is employed by a private engineering firm and has analyzed over 2,000 accidents.
*425 Hamemik testified that before braking, Zimmerman was traveling 30 to 35 m.p.h. He also testified that at impact, Zimmerman was traveling at 15 m.p.h. and Powell was traveling at 20 to 22 m.p.h.
Because of our ruling, we need not recite in detail Hamernik’s testimony on how he arrived at his opinions. Instead, we include only what is necessary to provide context.
Generally, two well-accepted methods for determining the speed of a vehicle involved in a collision are used within the engineering community: the conservation of momentum method and the conservation of energy method. Hamemik claims to have relied on both methodologies. However, to provide reliable results, each method must have reliable underlying data. Despite 160 pages of testimony, it is not clear what data Hamemik needed to make his calculations reliable, nor is it clearly explained where he got the data that he used. Apparently, some of the data were derived from simulations he ran using a “state-of-the-art” computer program called Human Vehicle Environment (HVE).
PROCEDURAL BACKGROUND
Before trial, Zimmerman moved to prevent Hamemik from testifying at trial. In the motion, Zimmerman argued that Hamemik’s opinions were unreliable under
Schafersman
v.
Agland Coop,
Seven days before trial, the court held a hearing under Neb. Rev. Stat. § 27-104 (Reissue 1995), i.e., a Daubert hearing, to determine whether to admit Hamernik’s opinions. After both parties presented evidence, the court took the motion under advisement.
Before opening statements, the court told the parties that it was still unsure whether Hamemik’s testimony concerning the vehicles’ speeds would be admissible. Specifically, the court noted that it was not convinced that either the HVE software or the manner in which Hamemik had used it was reliable. The court, as we understand it, ruled that Hamemik could not give *426 his opinions on the vehicles’ speeds at trial unless Powell first established the methodology’s reliability.
At trial, Hamemik offered further explanation of the methodologies he used to determine how fast the vehicles were going before the collision. This testimony is set out later in our opinion. The court, over Zimmerman’s objection, then allowed Hamemik to testify about Zimmerman’s speed before braking. In overruling the objection, the court did not explain why it had determined that Hamemik’s trial testimony had led it to conclude that his opinions were admissible under Daubert/Schafersman /.
We note that in addition to testifying about the vehicles’ speeds, Hamemik had intended to offer opinions on the magnitude of force Zimmerman would have experienced in the collision and how this force compared to other forces experienced in human volunteer testing. Before trial, the court specifically ruled that Hamemik could not compare the force Zimmerman experienced to forces experienced in human volunteer testing. Furthermore, although the court did not preclude Hamemik from giving his opinion as to the magnitude of force experienced by Zimmerman, Hamemik did not give any such testimony at trial. Thus, we are concerned only with Hamernik’s testimony about the vehicles’ speeds.
The jury determined that both Powell and Zimmerman had been negligent and that Zimmerman had suffered $17,851.18 in damages from the collision. In determining the negligence of the parties, the jury concluded that Zimmerman bore 49 percent of the responsibility for the collision and reduced the damages that she could recover by that percentage. Zimmerman appealed.
ASSIGNMENTS OF ERROR
Zimmerman assigns, reordered and consolidated, that the court erred in (1) failing to find that Hamemik’s opinions were unreliable under Daubert/Schafersman /, (2) refusing to exclude Hamemik’s testimony as a sanction for Powell’s filing of untimely discovery disclosures, (3) setting the Daubert hearing 7 days before trial, (4) placing the burden on Zimmerman at the Daubert hearing to show that Hamemik’s testimony was unreliable, (5) refusing to allow Zimmerman’s counsel to fully explore the basis for the assumptions made by Hamernik at the *427 Daubert hearing, (6) failing to announce its ruling on the Daubert motion until the trial’s commencement, and (7) allowing the jury to view computer-generated simulations of the collision prepared by Hamernik.
Zimmerman also assigns as error that the jury’s verdict is contrary to the evidence. However, this assignment of error is not argued and therefore we do not consider it on appeal. See
Gilroy
v.
Ryberg,
STANDARD OF REVIEW
The standard of review is discussed in the analysis portion of our opinion.
ANALYSIS
Daubert/Schafersman I Gatekeeping Duty
Section 27-702 governs the admissibility of expert testimony. It 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.”
Until our decision in
Schafersman I,
we had employed the
Frye
test to evaluate the admissibility of an expert’s testimony. Under the
Frye
test, when an expert relied on a scientific principle or discovery, the proponent of the expert’s testimony had to prove that the principle or discovery had “‘gained general acceptance in the particular field in which it belongs.’ ”
Schafersman
I,
In
Schafersman I,
however, we abandoned the
Frye
test and, in its place, adopted the framework set forth in
Daubert
v.
Merrell Dow Pharmaceuticals, Inc.,
Both the
Daubert/Schafersman I
framework and the
Frye
framework seek to exclude unreliable expert testimony, but they go about this in significantly different ways. The
Frye
framework relies on the expert’s field to weed out unreliable testimony. The focus is on whether “ ‘an asserted expertise [i]s believed valid by enough asserted experts. If enough of them [think] so — that is, if the asserted expertise enjoys “general acceptance”— then a court [i]s justified in concluding that the proffered testimony [i]s valid.’ ”
Schafersman I,
By contrast, under the
Daubert/Schafersman I
framework, the burden to weed out unreliable expert testimony is placed directly on the trial court. While general acceptance remains a valid factor to be considered, the trial court cannot “piggyback” its decision onto someone else’s judgment.
Schafersman I,
We recognize that the
Daubert/Schafersman I
framework will frequently require trial judges to grapple with unfamiliar scientific principles. Our recent decisions involving
Daubert/Schafersman I
have shown that this can be a daunting task. Accord
Daubert
v.
Merrell Dow Pharmaceuticals, Inc.,
The trial court’s task is also eased by the
Daubert/Schafersman I
framework, which recognizes a range of reasonable methods exists for distinguishing reliable expert testimony from false expertise. See
Kumho Tire Co.
v.
Carmichael,
The trial court’s discretion, however, is not boundless. As we have discussed, the
Daubert/Schafersman I
framework relies on trial courts to determine whether an expert’s testimony is reliable. Once a party opposing an expert’s testimony has sufficiently called into question “the testimony’s factual basis, data, principles, [or] methods, or their application ... the trial judge
must
determine whether the testimony has a reliable basis in the knowledge and experience of the relevant discipline.” (Emphasis supplied.)
Schafersman 1,
A necessary component of this rule is that a trial court, when faced with a
Schafersman I
objection, “must adequately demonstrate by specific findings on the record that it has performed its duty as gatekeeper.”
Goebel,
This does not mean, however, that trial courts must “ ‘recite the
Daubert
standard as though it were some magical incantation.’ ”
Goebel,
Thus, the record must include more than a recitation of the Daubert/Schafersman I boilerplate and a conclusory statement that the challenged evidence is or is not admissible. A trial court adequately demonstrates that it has performed its *431 gatekeeping duty when the record shows (1) the court’s conclusion whether the expert’s opinion is admissible and (2) the reasoning the court used to reach that conclusion, specifically noting the factors bearing on reliability that the court relied on in reaching its determination. See 29 Wright & Gold, supra. When the court fails to make these findings, it abdicates its gate-keeping function. Goebel, supra.
Standard of Review
We digress momentarily to discuss our standard of review. Generally, a trial court’s ruling in receiving or excluding an expert’s testimony which is otherwise relevant will be reversed only when there has been an abuse of discretion.
Carlson
v.
Okerstrom,
Trial Court’s Inadequate Daubert/Schafersman I Ruling
Here, the trial court held a Daubert hearing to determine whether to admit Hamemik’s opinions. At the hearing, both parties offered evidence on the admissibility of Hamemik’s testimony, and the trial court took under advisement Zimmerman’s motion to exclude the testimony. Before opening statements, the court told the parties:
I am concerned because of the way Dr. Hamemik did [his analysis] was that he used simulation software to try to replicate the actual action in the accident, and the Court is told virtually nothing about the simulation software except that its name is HVE and it’s software manufactured by General Dynamics, but its reliability I know nothing about.
*432 ... I am not satisfied that Dr. Hamernik is absolutely precluded from having sufficient basis to do this scientifically validly [sic] just because I don’t know yet about this simulation software, whether it was reliable and whether it will accomplish some deficiencies that Dr. Hamernik had on inputs .... Although it’s my suspicion that this simulation software was used and that is how he came up with the numbers that he did for things like the angle of the cars after the impact, the distance traveled after impact and so forth.
So what I’m going to do is I’m going to exercise the discretion in [Neb. Evid. R.] 705, which requires that the expert disclose the underlying facts or data before I will allow his opinion. And some things that I am deficient in the underlying facts or data to make it scientific is any explanation about this software, how the simulations work, whether it’s reliable and whether it’s reliable enough information for Dr. Hamernik to fill in the variables that he is missing when he then seeks to reconstruct the accident.
Some of those variables that he is missing is [sic], as a I said, what the angle the car was after impact, he’s coming up with that from simulation software, in my opinion. Whether those will fill in the holes or not will just depend on his testimony at trial whether he’ll be able to establish it. I am requiring that he give that underlying facts and data before I will allow his opinion.
Later, the following exchange occurred between the parties’ counsel and the court:
[Counsel for Powell]: Okay. Just so I’m understanding the Court’s ruling, [Hamernik] can come in and testify about the speeds, it’s the forces that you are needing more information about? In other words, the fact that he has calculated . . . Zimmerman going 30 miles an hour before she applied her brakes, the fact that he has calculated she was going 15 miles an hour at impact, those are fine opinions?
THE COURT: I’m sure if you check, the 15-mile-an-hour impact comes from a myriad of assumptions through simulations that he ran, and he needs to establish that those *433 simulations were both reliably done and can furnish sufficient reliable information for him to conclude what that speed was at impact, because my hunch is that he is simply taking some photographs and estimating crush from photographs to get the speed at impact, and that may not be sufficiently reliable. That certainly is not sufficiently reliable according to some of these reconstructionists, you would need more information than a photograph of one car to come up with all of that.
THE COURT: I don’t know how he’s doing it. I know he did not explain it at the hearing.
[Counsel for Powell]: You need for him to explain to you in more detail how he had done his calculations?
THE COURT: No, explain to the jury. I’m going under [rule] 705, so I am requiring that he furnish to the jury, quote, “the underlying facts and data,” before I will let him give any opinions at all. I’m not precluding him from giving any opinions, I’m requiring that he’ll have to establish the underlying facts and data to show that his opinion is scientifically valid and it’s not just a product of a whole bunch of unreliable simulations.
[Counsel for Zimmerman]: I just have a little question. You don’t want the information outside the presence of the jury before?
THE COURT: No, no, no. We’ve had plenty here. We spent almost a day and I still have no idea about this simulation software, not one word except the name.
THE COURT: I think he’s qualified as a witness, I’m not concerned about that, I’m concerned whether he had enough data to put into the formulas to get the conclusions he did.
THE COURT: I’m not limiting his testimony, I’m just saying it’s going to be done in a certain way. He’ll just give his opinions and then go on, and in cross-examination to reveal its basis he’ll have to establish the basis first, which is somewhat opposite from the way experts are allowed to testify many times.
*434 [Counsel for Powell]: Okay.
THE COURT: If it doesn’t work out to make it scientifically reliable . . . you will be in the spot of having to tell the jury in opening you think you’ll get those opinions in and discover you were not able to, I can’t prejudge that.
These pretrial discussions show that the court believed — and we think correctly — that it did not yet have enough information to determine whether Hamemik’s opinions on the vehicles’ speeds were admissible under Daubert/Schafersman I. Specifically, the court expressed concern whether the HVE software was reliable and whether the manner in which Hamemik used it was reliable. The court, as we understand it, ruled that before Hamemik could offer his opinions at trial, Powell would have to establish the reliability of Hamemik’s methodology.
At trial, Hamemik gave the following explanation of the HVE program:
There’s software out there . .. produced by ... Engineering Dynamics Corporation .... They’ve developed programs over the years to look at both energy and momentum for car crashes. This program was initially developed by . . . the National Highway Transportation Safety Administration, but now has evolved over the years into a program that they call HVE, which stands for Human Vehicle Environment.
This program matches the physics associated with a car crash and it models both energy and momentum simultaneously at the same time. When you use this program, you produce something called a simulation. It looks like an animation, but it’s a simulation, actually how the two cars interact and move and the damage that will result.
So I’ve checked my hand calculations with HVE. In fact, I ran HVE to get a ball park idea of how the cars would likely move, conducted hand calculations, then subsequent to that I’ve done complete simulations of this accident.
[HVE] is state of the art. It was first released in 1992. Engineering Dynamics has been producing software for, I think, 25 years now. HVE is the latest and greatest of their software, it’s well-validated. They’ve ran hundreds of tests that actually take cars and crash them at known *435 speeds, measure the change in speed, the post-impact speeds and where the cars end up at rest. Their software matches all these values within 1 percent. It’s a very scientific, very accurate program.
You can recreate accidents with this program, but you cannot make it do whatever you want, you have to obey the law of physics. And the input will — in this accident, if you know how the cars roughly approach, you can calculate where the cars should end up, it will actually show you that. It will predict the damage, so you can modify the speeds until the damage matches and also that the cars end up where they’re supposed to end up, roughly.
. . . You can’t violate the laws of physics. This program takes in account the entire forces of the roadway, the locations of what we call the center of gravity. . . .
HVE takes into account the location of the [center of gravity], the weights of the vehicle and the tire forces with the roadway and the stiffnesses or the strength of the vehicles so you can account for the deformation and the post-impact movement, and they all have to match in order to get the right solution. I cannot simply say that I want the car to end up here and this car ends up here, it won’t do that unless I obey the laws of physics.
Shortly after this testimony, Powell’s counsel asked Hamemik what opinions he had reached as a result of his accident reconstruction. Zimmerman’s counsel objected, and the court stated, “Overruled. You may give your opinions.” The transcript also contains a written order, noting that Zimmerman’s “Daubert Motion was . . . overruled in part.”
The record, however, contains only the court’s conclusion; there is no analysis. The court should have explained why Hamemik’s trial testimony was sufficient to show that HVE and the manner in which he used it were reliable. For example, if the court believed that HVE was reliable because Hamemik suggested that it was widely accepted as quality software within the engineering community, it should have said so. Because the court failed to explain its reasoning, we conclude that it abdicated its gatekeeping duty.
*436 Prejudice to Zimmerman
Our conclusion that the trial court erred by failing to perform its gatekeeping duty does not, however, end our inquiry. Not every error justifies a new trial; only an error which is prejudicial to the rights of the unsuccessful party does so.
Gourley
v.
Nebraska Methodist Health Sys.,
Some courts have held that when the trial court fails to make the required findings, the appellate court should conduct the
Daubert/Schafersman I
analysis on the appellate record. See,
Kinser
v.
Gehl Co.,
But in our view, this improperly shifts the gatekeeping duty from trial courts to appellate courts. There is a reason that trial courts, rather than appellate courts, bear the gatekeeping duty. “Trial courts . .. have a much broader ‘array of tools which can be brought to bear on the evaluation of expert testimony’ . . . .” 4 Jack B. Weinstein & Margaret A. Berger, Weinstein’s Federal Evidence § 702.02[6][c] at 702-29 (2d ed. 2004). An appellate court is limited to a cold record and thus is not in a position to perform the gatekeeping role in a manner that is fair to the parties. Cf.,
Goebel
v.
Denver and Rio Grande Western R.R. Co.,
Thus, when a trial court fails to make the requisite findings, the losing party will usually be prejudiced. Only if the admission or exclusion of the expert’s testimony did not affect the result of the trial unfavorably for the party against whom the ruling was made will a court’s abdication of its gatekeeping
*437
duty be deemed nonprejudicial. See,
Dodge
v.
Cotter Corp.,
Here, Powell conceded that she failed to yield at the intersection, but argued that Zimmerman was contributorily negligent because she was driving above the speed limit as she approached the intersection. The jury agreed and concluded that Zimmerman bore 49 percent of the responsibility for the collision. Because Hamemik’s testimony was the centerpiece of Powell’s comparative negligence defense, we cannot say that the admission of his testimony did not affect the result of the trial.
Remedy
Zimmerman contends that we should remand for a new trial on all issues. Powell argues that Hamemik’s testimony was relevant only to the issues whether Zimmerman was contributorily negligent and the apportionment of damages and that therefore Zimmerman cannot relitigate her damages. We agree with Powell.
Partial retrials are permissible if it clearly appears from the record that the issue to be retried is so distinct and separable from the others that a trial of it alone may be had without injustice.
Blue Valley Co-op
v.
National Farmers Org.,
The jury verdict form specifically shows that the jury determined (1) both Powell and Zimmerman had been negligent, (2) Zimmerman incurred $17,851.18 in damages, and (3) Zimmerman was 49 percent responsible for the collision and thus could recover only 51 percent of her damages, or $9,104.10. Hamemik’s testimony focused on whether Zimmerman was driving above the speed limit before she began braking. Thus, the court’s failure to conduct its gatekeeping function tainted the jury’s determination of Zimmerman’s contributory negligence and the apportionment of damages. If we were to remand, these issues would have to be retried. Moreover, because a jury would be unable to apportion damages without an understanding of *438 Powell’s negligence, the issue of Powell’s negligence would also have to be retried.
Hamemik’s testimony, however, did not taint the issue of Zimmerman’s damages. Zimmerman and her family and friends testified about her injuries and how those injuries had affected her life. Medical personnel who had treated her also testified, and she introduced a summary of her medical bills into evidence. To counter Zimmerman’s evidence on damages, Powell relied on cross-examination and a private investigator to rebut Zimmerman’s evidence of damages. Hamemik’s testimony did not touch on Zimmerman’s damages; he focused only on how fast the vehicles were going before the collision.
Nor can we say that if we were to grant a retrial, the issue of damages would be so interwoven with the negligence and apportionment damages issues that a partial retrial would be unfair to either party. We fail to see how knowledge of the nature and extent of Zimmerman’s injuries would be necessary to resolve those issues.
Thus, were we to remand, we would limit the issues to be tried to the parties’ negligence and the apportionment of damages. Partial remand, however, is not required. Because of our conclusion that Zimmerman cannot relitigate damages, the most she could recover at a new trial would be $17,851.18. In her brief, Powell conceded that if we were to decide that the court erred in allowing Hamemik to testify, “there is no reason to retry the case and this Court can modify the Judgment... so as to allow [Zimmerman] the sum of $17,851.18, the sum awarded by the jury before any deduction for [Zimmerman’s] contributory negligence.” Brief for appellee at 45. In other words, instead of remanding for a new trial, Powell has asked us to modify the judgment so as to give Zimmerman the best possible result she could receive if the case were retried. Because of this concession, we choose not to remand for a new trial. Instead, we modify the judgment so that Zimmerman receives $17,851.18.
Zimmerman’s Other Assignments of Error
Finally, it is necessary for us to comment on Zimmerman’s other assignments of error. In each of these, she argued that the *439 court erred in allowing Hamemik to testify. As a result, if we were to conclude that any of these assignments of error had merit, the remedy would be the same as what we have already granted to Zimmerman. As a result, we need not address Zimmerman’s remaining assignments of error.
CONCLUSION
We conclude that the trial court failed to perform its Daubert/Schafersman I gatekeeping function and that this error tainted the issues of contributory negligence and apportionment of damages, but not the amount of Zimmerman’s damages. Because of the concessions made by Powell, however, we do not remand for a partial new trial. Instead, we modify the court’s judgment so that Zimmerman receives $17,851.18.
Affirmed as modified.
