The Court orders that a special panel shall be convened pursuant to Administrative Order No. 1996-4 to resolve the conflict between this cause and Sumner v General Motors Corp, 212 Mich App 694 (1995).
The Court further orders that the opinions in this case released September 24, 1996, are hereby vacated.
The appellant shall file a supplemental brief within 28 days of the clerk’s certification of this order. Appellee may file a supplemental brief within 21 days of service of appellant’s brief. Nine copies must be filed with the Clerk of the Court.
LOPEZ v GENERAL MOTORS CORPORATION
Docket No. 164400. Released September 24, 1996, at 9:10 am.; vacated December 30, 1996.
Before: White, P.J., and Bandstra and W. P. Cynar,* JJ.
White, P.J. Plaintiff appeals as of right from a judgment of no cause of action entered on a jury verdict in this products liability case involving the crashworthiness of an automobile and a claim of enhanced injury due to an alleged defect in, and malfunction of, a 1987 Chevrolet Chevette’s three-point fixed-loop shoulder restraint system. I would reverse and remand for a new trial on the basis that the trial court abused its discretion in admitting two videotapes of experimental General Motors (gm) crash tests that were neither performed under conditions substantially similar to plaintiff’s accident nor offered merely to establish general scientific principles.
Plaintiff testified at trial that on the morning of January 10, 1989, she drove the Chevette, purchased by her father, to work, taking her regular route down Sheridan Road, an unpaved, gravel, two-lane road in Durand, Michigan, on which plaintiff and her parents lived. Plaintiff testified that it was pitch black outside and cold and that, at a railroad crossing about one mile from her home, she hit what she later learned was a flatbed railroad car blocking the section of railroad tracks on her side of the traveled portion of Sheridan Road.1 Plaintiff estimated that she was going between [802]*802fifteen and twenty miles an hour2 when the accident occurred. It is undisputed that plaintiff was wearing her lap and shoulder safety belt at the time of the accident. Plaintiff testified that she sat in the driver’s seat with her back all the way up against the seat and that the seat was positioned in the mid-range position. Plaintiff testified that she could not recall the moment of impact and that after the impact she was still sitting in the driver’s seat with the restraint system around her. At the time of the accident, she weighed 125 pounds and was 5’ 3” or 5’ SW tall.
Plaintiff sustained severe injuries, the extent of which was an important issue at trial. The injuries included facial fractures requiring reconstructive surgery, including a crushed nasal septum and column, collapsed orbital floor, and broken cheekbone. Plaintiff also sustained a broken clavicle and bruising to the face, breasts, shoulders, and knees.
Plaintiff’s counsel argued at trial that the shoulder restraint system’s lock bar and ratchet were defective and that the shoulder belt on impact latched but then broke loose, resulting in plaintiff’s head and shoulders being unrestrained, her face traveling forward and hitting the steering wheel, and both shoulders and both sides of her chest suffering serious injuries as well. Plaintiff’s counsel argued that plaintiff was entitled to damages for enhanced injuries, i.e., those resulting from her collision with the steering wheel, which followed her car’s collision with the railroad car, noting that damages for plaintiff’s injuries from the lap belt were not sought, because that belt had functioned properly.
Plaintiff’s father, Mr. Lopez, who had driven the 1987 Chevette several months before plaintiff began using it in December 1986, testified he had no problems with the Chevette’s seat belt, but had never been involved in an accident in that car. Mr. Lopez testified that his wife had complained to him that the passenger seat belt locked up on her when riding on bumpy roads. Plaintiff’s mother, Mrs. Lopez, testified that she went to the scene of the accident and to the hospital and that plaintiff told her at the hospital that she had been going fifteen miles an hour at the time of the accident. Mrs. Lopez testified that she drove the Chevette three or four times and that the seat belt cinched or tightened up on her as she drove down the bumpy road, as well as when she was a passenger in the car. Mrs. Lopez testified that the Chevette could not be driven too fast down Sheridan Road because the road was poorly maintained and the car could not be kept on the road if not driven carefully. She estimated that in the summer, and if it had not rained, one could drive up to thirty miles an hour on that road.
Plaintiff’s expert, Henry Kowalski, testified that he has a Ph.D. in engineering mechanics and had been a professor, administrator, and senior researcher at GMI Managing Institute (formerly General Motors Institute) for more than twenty-five years, teaching courses in automotive structural analysis and vibration analysis, and heading the experimental stress analysis laboratory. Kowalski testified that the motion of a person restrained in [803]*803an automobile after a frontal barrier impact is “basically a problem in [his area, i.e.,] engineering mechanics, primarily dynamics.”
Kowalski testified that he had examined plaintiffs Chevette and he could not get the spring mechanism that holds the shoulder belt in a certain position to engage. Kowalski testified that the passenger-side belt operated differently. Plaintiff’s counsel sought to qualify Kowalski in three areas: (1) calculating a vehicle’s speed by determining crush or deformation; (2) mechanics of the shoulder harness system; and (3) occupant kinematics. The court ruled Kowalski could not testify regarding accident reconstruction as it relates to speed, but determined he was qualified as an expert in the mechanics of the shoulder harness system and, over defense counsel’s renewed objection, in occupant kinematics.
Kowalski testified at length regarding how the restraint system malfunctioned and that if the shoulder restraint system had functioned properly plaintiff would not have rotated forward and smashed her face into the steering wheel. Kowalski opined that plaintiff would not have sustained such severe facial, clavicular, and other iryuries had the shoulder harness of the restraint system functioned properly, that the injuries were caused by the malfunction of the shoulder harness, and that the 1987 Chevette was not crashworthy as a result. He stated that if the shoulder harness had functioned properly, plaintiff would likely have sustained some bruising to the left shoulder and welts and bruising to the ribs, but not such severe facial injuries. Kowalski testified that the Chevette’s steering wheel had moved forward 1 to Vk inches and that plaintiff’s face hitting it alone could not have accomplished this; it would have required most of the mass of the human body, i.e., the upper torso, rotating forward.
Before the defense called its witnesses, a lengthy discussion was held on the record out of the jury’s presence, during which the court overruled several objections of plaintiff’s counsel to admission of two videotapes of gm tests: an October 1986 sled test simulating a thirty miles an hour collision and a 1977 frontal barrier crash test on a Chevette traveling at 30.7 miles an hour. Plaintiff’s counsel objected to the admission of the sled test on the ground that the sled test involved a test vehicle that was not a standard automobile, and which had been modified, and David Peruski, a defense expert and gm mechanical engineer, had testified at deposition that he did not know what the modifications were. Plaintiff’s counsel objected to the crash test on the bases that it was conducted at 30.7 miles an hour, six miles an hour faster than the speed the defense’s expert estimated plaintiff’s car was going; it was conducted in 1977, more than ten years before plaintiff’s Chevette was manufactured; the test had been conducted to evaluate the difference between a fender metal thickness of 0.033 inches and 0.030 inches, to see if thinner metal could be used later; the thickness of the test vehicle’s fender was unknown and the thickness of the fender of plaintiff’s automobile was unknown; and plaintiff was five feet three inches tall and weighed 125 pounds, while the dummy used in [804]*804the gm tests was five feet seven inches tall and weighed 175 pounds.3 Plaintiffs counsel argued that, according to Kowalski’s computations, the higher speed and heavier dummy in the gm test would have meant that that dummy was subjected to a force more than 2.19 times greater than plaintiff was subjected to in her accident. Regarding both tapes, plaintiff’s counsel argued that there was no proper foundation because the personnel who conducted the tests were not testifying, PerusM’s name appeared nowhere on the test results, and the defense was not calling anyone who could validate the tests or the circumstances under which they were conducted. Plaintiff’s counsel argued that the tests were hearsay and irrelevant and that their prejudicial effect would outweigh their probative value. Finally, plaintiff’s counsel objected that defendant was presenting no one who could testify that the restraint system had operated properly during the videotaped tests and asserted that the seat belt had broken in one of the eight tests defendant ran, which would not be shown at trial. Plaintiffs counsel argued that the instrumentation information defendant provided had none of the necessary information.4
Defense counsel argued in response that the tests were relevant to plaintiffs negligence claim as evidence of due care in the manufacture of the vehicle based on repeated testing of the seat belt system and that the [805]*805tests “are absolutely critical support for our experts’ opinions about the likely occupant kinematics, the general principles of occupant kinematics in a frontal crash.” Defense counsel argued that both tests showed that “under substantially similar situations, a fully restrained dummy with a restraint system that locks up will in fact have excursion forward, bob their head and hit their head near the steering column.” He further argued that the purpose of both tests was to evaluate the performance of the Chevette belt system,5 that the belts in the tests were the same three-point type of system as plaintiff’s, and that Peruski would testify to that and would also testify that a difference in fender thickness “does not substantially affect the occupant kinematics involved in the frontal barrier crash test.” Defense counsel stated “any nitpicking about a couple of miles an hour difference can be brought as going to weight.” Under questioning by the court, defense counsel conceded it would have been possible to duplicate plaintiff’s accident for purposes of illustration to the jury, but argued it was not necessary to do so. He concluded by arguing that the tapes were “a central part of our defense. These tests absolutely demonstrate to the jury and the Court what the kinematics are in an accident like this.... [Defense experts Peruski and Karen Morley] rely on them for purposes of their opinion. And they should come in as business records subject to whatever cautionary instructions may be appropriate and subject to points that go to the weight of the evidence.”
Plaintiff’s counsel then responded that the videotapes did not qualify under MRE 803(6) as business records and also objected to Peruski testifying about occupant kinematics, arguing that at Peruski’s deposition gm’s counsel (not defendant’s trial counsel) had objected at least twice on the record when plaintiff’s counsel attempted to question Peruski regarding this area, stating that Peruski “has not been offered in the capacity of a biomechanical engineer, nor an expert in occupant kinematics.”
The court admitted the videotapes as business records, indicating that plaintiff’s foundation objection was addressed by defendant’s counterargument that the tapes would illustrate “the principles involved and the principles don’t change even given the difference in speeds, difference in dummy weight, size, et cetera.” With respect to plaintiff’s relevancy objection, the court indicated it would leave it to counsel to point out to the jury the dissimilarities between the test conditions and the actual accident. With respect to qualifying Peruski in the field of occupant kinematics and biomechanical engineering, the court indicated it would “leave it to the qualifications to be established if he is going to be able to give that kind of testimony.”
The defense called Peruski, a mechanical engineer employed by gm since 1963. Peruski testified that around 1986 he spent one year at the Milford proving grounds managing durability and test driving the Fiero, Corvette, Bonneville, Regal, Cutlass Supreme, light trucks, and small vans. [806]*806The Milford proving grounds had facilities for safety research and a development laboratory where sled and crash testing was conducted. PerusM testified that his gm experience included automobile safety testing, automobile safety design, and testing of gm seat belt systems. Around 1989, PerusM went to the safety center as a senior staff analysis engineer and had been principally involved in the evaluation of product performance in the field and the relaying of that experience and knowledge to the designers of future vehicles. PerusM testified that he had evaluated real accidents, had looked at more than one hundred cars involved in crashes, and had testified in approximately five cases previously. Defense counsel asked PerusM to explain to the jury what a sled test is. He explained that in a sled test, unlike in a barrier/actual crash test, the vehicle body and dummy start out stationary, and then the acceleration is in the reverse direction — the vehicle body is accelerated out from under the dummy. PerusM explained that variables are controlled in a sled test, and a reinforced body/sled buck is used.
In contrast, in a barrier/actual crash test an actual vehicle is run into a barrier. PerusM explained that there were limitations to sled tests and that certain results are not simulated, including areas of body yield after a front-end crush. PerusM was involved in sled and barrier tests of the Chevette when it was first manufactured in 1976. He testified that the 1976 model year Chevette was substantially similar to the 1987 Chevette, and all years in between, including the belt restraint system.
PerusM testified that the Chevette shoulder restraint had a vehicle-sensitive retractor, which is intended to make the belt system more user friendly, i.e., allow people to move around in the vehicle absent accident configurations, when the belt locks up. PerusM testified that KowalsM’s observations about the shoulder belt ratcheting teeth being offset was correct, but that it was so by design and not an indication of a defect. PerusM also testified, responding to KowalsM’s testimony, that when he examined the Chevette the belt would not lock up when he pulled on it, but that it was not evidence there was anything wrong with the belt. PerusM testified that the belt was not designed to pin a passenger back against her seat in a frontal collision, that in a thirty miles an hour impact involving a fiftieth percentile occupant, the dummy’s hips would move forward six or eight inches.
PerusM testified that the sled test materials are prepared and maintained in the ordinary course of gm business, prepared by “someone” with knowledge about how the tests were run, and that the setup sheets describe how the test was conducted “to some extent.” PerusM added that only in rare cases are written reports made of the tests. When asked whether the setup sheets tell what kind of restraint system was used in the particular sled test, PerusM answered “to some extent, yes.” When questioned on voir dire by plaintiff’s counsel about the setup sheets, PerusM testified that he did not participate in or witness the specific videotaped sled test and that he recognized the name on the form as that of an engineer he knew, but he had not spoken to him about the videotaped test.
[807]*807Plaintiff’s counsel renewed Ms objection to qualifying Peruski on occupant kinematics, but did not oppose qualifymg Mm with respect to restraint system design, testmg and performance, safety standards, and accident reconstruction. The court qualified PerusM with respect to all areas, stating that it did not find any suiprise to plaintiff or new areas not already discovered or delved into by plaintiff’s counsel at PerusM’s deposition.
Regardmg the sled test conducted in 1986, PerusM testified the vehicle was set up with a tMee-point Chevette belt system identified as “an experimental belt system” m terms of the gmde assembly, i.e., the portion that goes above the occupant’s shoulder, wMch appears to refer to an experimental “D-ring.” PerusM testified that “it’s basically a tMrty mile per hour frontal impact with two fiftieth percentile dummies,” and that “typically” when fiftieth percentile tests were done, the seat was m the mid-position, i.e., m the fourth notch of seven total notches, and the seat belt angle would be adjusted to twenty-six degrees. PerusM later testified that when he exammed plaintiffs veMele M October 1991, the seat was m the third notch and he was confident that that was the position the seat was m at the time of the accident.
When asked about the model year of the sled test Chevette, PerusM answered “certainly would have been m ‘86 or later . . . probably the ‘87.” The basic structure of the Chevette body from 1976 to 1987, accordmg to PerusM, “would have been the same.” When asked if the Mnematics in the sled test were “substantially or materially affected by the experimental D-ring that was used, PerusM answered ‘I would not expect it to ... .’ ”
PerusM testified that the sled test simulated a two-door Chevette strikmg a barrier face at thirty miles an hour usmg a body with no wheels or doors. When asked by plaintiff’s counsel whether the videotape was a duplication of a real situation, PerusM answered, “It simulates it under very controlled, setup circumstances.”
The sled test videotape was then shown, with PerusM explainmg as it played. At various times, PerusM made reference to the differences between the test and the accident, e.g., “[although there are differences between tMs test and the accident configuration, generally speaking, the kinematics will be the same,” and at other times he stated “tMs is not a representation of the accident.” PerusM opined that, with respect to how the occupant moves, plaintiffs movement would have been the same as the dummy’s m the thirty miles an hour sled test videotape, although the amount of forward movement or excursion would “vary slightly.” PerusM was asked on direct examination:
Q. Is that driver dummy there actually contacting the steering wheel m this particular sled test? Can you tell from that shot?
A. Very close to if not in tMs case.
At various times, the videotape was frozen on a particular frame, rewound, and replayed.
Regarding the second videotape, PerusM explained that the frontal-barrier test mvolved a modified, two-door 1977 Chevette going 30.7 miles [808]*808an hour. He testified that he believed the objective of this test was to analyze the crash performance of the vehicle with a reduced front fender of 0.030 inches, rather than 0.033 inches. Peruski did not know if after this test gm placed the thinner fender on its Chevettes, and he did not know if a 1987 Chevette would have had the thinner fender or not. When asked if he contended that the frontal-barrier test is a recreation of the Lopez accident, Peruski answered “[cjertainly not, no,” then stated:
I think the video does a fine job of demonstrating how the vehicle .. . crushes overall, what the occupant kinematics are and equally important, what the relationship is between the vehicle and the occupant or in other words, as the vehicle is crushing, what the occupant is doing.
As the videotape played, Peruski indicated that the dummy in the test moved forward and made contact with the steering wheel area. Peruski testified that some degree of involvement of the steering wheel is “typical in this vehicle, yes.”
Peruski testified that he inspected plaintiffs vehicle in October of 1991 and inspected the restraint system. Peruski calculated that plaintiff’s vehicle was going “roughly 25 miles per hour. I can’t say that that’s exact. . . there is some error in that.” Peruski testified that there was evidence the shoulder belt had functioned properly, including significant loading and deformation to the pillar guide assembly. Peruski opined that the restraint system in plaintiffs car performed as it was supposed to and that given plaintiff’s height and weight and the other kinematics in the case, he would expect plaintiff to have had less involvement with the steering wheel than in the crash test, but her chance of having some contact with the steering assembly was great. Peruski further testified that he did not believe that, as plaintiff’s expert testified, the belt may have engaged and then broken loose.
On cross-examination, Peruski was questioned about the exact weight of the dummy used in the test. Plaintiff’s counsel pointed out that Peruski had testified that the dummy weighed 167 or 168 pounds, while defense expert Morley testified in her deposition that the dummy weighed 175 pounds. Peruski stated he was just speaking from his recollection and his recollection of the weight differed from Morley’s. Peruski also conceded that the dummy would have been subjected to forces approximately double the forces plaintiff was subjected to in the accident. On cross-examination, Peruski testified that he did not know the difference between the D-ring used in the restraint system in the 1986 barrier impact test and the D-ring in plaintiff’s restraint system. Peruski also testified that the 1986 test gave no information regarding the level of force with which the dummy made contact with the steering wheel.
[809]*809Defense expert Morley testified by videotaped deposition regarding occupant kinematics and biomechanics (how bodies react to forces).6 Morley testified that she examined plaintiff’s vehicle in October 1991, the same time Peruski did. She opined that plaintiff had been wearing her seat belt and that the collision had not been fully frontal, but a bit skewed to the driver’s left. Morley testified that she had reviewed six papers, some presented at conferences of the Society of Automotive Engineers, dealing with injuries belted drivers suffer in collisions. Morley testified that the literature and examination of the restraint system in plaintiff’s Chevette showed that in accidents such as the one plaintiff experienced, a fully restrained driver is expected to strike the head and face area into the steering wheel when the restraint system functions properly. Morley noted that plaintiff’s car seat was in the third notch position, as had Peruski, and adopted Perusld’s calculation of plaintiff’s accident speed as twenty-five miles an hour, testifying she had not done an independent accident reconstruction. Regarding defendant’s videotapes,7 Morley testified that they demonstrated her description of occupant kinematics in a frontal-type collision, and that although the speeds differed from the speeds involved in plaintiff’s accident,
they are comparable in terms of showing the kinematics of the dummy going into the restraint system, fully loading the restraint system, and that phenomena of the head bobbing into the steering wheel rim and hub area. And the significance is, is that in each of these eight fully frontal tests, there is contact made by that dummy with that steering wheel assembly. It’s an expected type of contact.
At Morley’s deposition, plaintiff’s counsel objected, on hearsay grounds, to the two videotapes being marked as exhibits. The videotaped tests were not shown during Morley’s videotaped deposition. Morley opined that the restraint system in plaintiff’s car had been activated, functioned properly, and that the injuries plaintiff suffered, including the broken left clavicle, were consistent with a properly functioning restraint system. On cross-examination Morley testified that with respect to the abrasions to plaintiff’s right shoulder and breast, she could not explain why plaintiff suffered these injuries. Over defense counsel’s objection, plaintiff’s counsel also presented Morley with an article entitled “Diagnosis of Seat Belt Usage in Accidents,” which indicated clavicle fracture was not a prevalent injury among those who use three-point shoulder belts. Morley testified that the dummies used in the videotaped tests were fiftieth percentile [810]*810male dummies weighing about 175 pounds and being about five feet seven inches tall. When Morley was asked on cross-examination if it was true that the seat belt had broken in one of the eight videotaped tests, she responded “the driver lap belt webbing separated during the test.”
On rebuttal, plaintiffs expert Kowalski testified that Peruski’s calculation of plaintiff’s speed at the time of the accident was flawed and that, assuming Peruski’s crush calculations were correct, the absolute maximum speed was 24.67 miles an hour and not twenty-five miles an hour give or take a few miles an hour, as Peruski testified. Kowalski then testified that Peruski’s crush calculation was flawed and that his figure of 16.85 inches was high. He further testified that the force involved in the thirty miles an hour frontal test crash was twice that involved in plaintiff’s crash, assuming defendant’s twenty-five miles an hour figure, that a restraint system has to do less work when there is less force on it, and that the difference in weight between plaintiff and the dummy was a significant factor in calculating the force. Kowalski further testified that, if plaintiff was traveling at 20-21 miles an hour, the difference in force between her accident and the videotaped tests would be more than three times.
Kowalski testified that several of the teeth in plaintiff’s shoulder belt retractor were chipped, and not as a result of the accident. He testified that those chipped teeth did not match the passenger-side belt, Peruski’s exemplar belt, or a belt Kowalski got from a junk yard and could have permitted plaintiff’s belt to engage and then slip in the accident. On redirect examination Kowalski reiterated that, had plaintiffs shoulder restraint system functioned properly, she would not have hit the steering wheel.
The jury by special verdict form found defendant had not been negligent and had not breached an implied warranty, never reaching the issue whether plaintiff suffered any enhanced injuries.
I
We first address plaintiff’s argument that the trial court abused its discretion in admitting the videotaped sled and crash tests over plaintiff’s objection that the tests were conducted under conditions dissimilar to plaintiff’s accident and that the tapes lacked foundation, were hearsay, and were more prejudicial than probative. This Court’s recent decision in Sumner v General Motors Corp, 212 Mich App 694; 538 NW2d 112 (1995), published after the trial court’s decision and the filing of the parties’ briefs in this matter, is controlling.
This Court held in Sumner that results obtained in out-of-court experiments are not admissible unless the conditions of the experiment are sufficiently similar to those involved in the particular case. Id. at 696, citing Przeradski v Rexnord, Inc, 119 Mich App 500; 326 NW2d 541 (1982), remanded on other grounds 417 Mich 1100.19 (1983).8 In Kirk v Ford [811]*811Motor Co, 147 Mich App 337, 343-344; 383 NW2d 193 (1985), this Court found that the trial court did not abuse its discretion in admitting films, photographs, and reports of crash tests where the plaintiff had not established that they were performed under conditions substantially similar to the accident in question, because they were intended to illustrate general principles relating to the vulnerability of under-the-floor fuel tanks in rear-end collisions and the evidence did not purport to have been generated under similar conditions. Thus, the Court reasoned they were admissible because they were not offered for the purpose of duplicating or recreating the accident, but merely to illustrate certain general principles. See also Gorelick v Dep’t of State Hwys, 127 Mich App 324, 336; 339 NW2d 635 (1983)(holding that where motion pictures are offered to recreate the scene of the accident, they are not admissible unless they portray conditions almost identical to those prevailing at the time of the accident itself; conversely, where offered merely to illustrate certain general principles, differences in surrounding conditions do not require the film’s exclusion).9
As was done in Sumner, defendant here purported to offer the videotaped tests to illustrate general scientific principles, and not to recreate plaintiff’s accident, but the record contradicts this assertion. Defendant’s trial brief stated the videotaped tests were conducted under “substantially similar conditions,” implying it was recreating plaintiff’s accident. At trial, the defense repeatedly discounted the differences between the conditions under which the videotaped tests were conducted and plaintiff’s accident. In its appellate brief, defendant argues simultaneously that it did not offer the videotapes to recreate the accident, that the conditions were substantially similar, and that “slight” differences in circumstances do not make a test inadmissible. In addition, defense witnesses repeatedly testified to the effect that the test results could be generalized to the facts of the accident.
I conclude, first, that defendant did not establish the requisite similarity. I disagree with defendant’s assertion that the differences in circumstances [812]*812between the videotaped tests and plaintiff’s accident were “slight.” The record establishes that the conditions of plaintiff’s accident were dissimilar in numerous ways to the conditions of the videotaped tests. In addition to significant disparities in the speeds of the vehicles, height and weight of plaintiff versus the anthropomorphic dummies, seat positioning, and resultant discrepancies in force exerted in a frontal collision, the defense could not establish with certainty that the restraint system operated properly in the tests, because the records of the tests did not so indicate. Further, although the sled test was conducted to test the restraint system and was conducted on a 1987 Chevette, PerusM testified that an “experimental” belt system was used in the test and that it had a D-ring. The second videotape, the actual crash test, was not conducted to test the restraint system, but to test fenders of varying metal thiclmesses. That test was conducted ten years before plaintiff’s automobile was manufactured, and PerusM testified he did not isnow if plaintiff’s car had the tested thickness of fender. Moreover, Morley testified plaintiff’s collision was not fully frontal, but rather skewed to the left slightly. While PerusM testified that some of the dissimilarities would not be expected to alter the results, significant dissimilarities in, and unanswered questions about, the crash test conditions remained.
Regarding the height, weight, and speed dissimilarities, defendant took the position that the general principles were the same and that the tapes illustrated the general principles. However, the general principles of occupant kinematics were not in dispute. The question was whether the severity and extent of plaintiff’s injuries, given her weight, height, and speed of travel, indicated that the shoulder harness had malfimctioned after impact. The jury was shown tapes of dummies experiencing frontal collisions wherein the dummies’ faces appeared to hit the steering wheel. Testimony focused not only on the general movement of the dummies in relation to the vehicle, but also on the extent of the movement and whether the dummies hit the steering wheel. Thus, while defendant paid lip service to the principle that the circumstances of the depiction need not be substantially similar if the depiction is offered merely to illustrate general principles, the tapes, in effect, became a focal point and were used as evidence to prove defendant’s position on a specific factual question, i.e., how the seat belt was supposed to perform in this accident. See Sumner, supra at 696-697.
The defense was in essence using the tests to prove that the severity and extent of plaintiff’s injuries were the normal results of a frontal collision. The record establishes that the videotapes were used well beyond the illustration of general scientific principles and their admission was error. Under these circumstances, I am unable to conclude that the videotapes of the tests were properly admitted as limited to establishing general principles of occupant Mnematics or as having been conducted under conditions substantially similar to plaintiff’s accident.
Crucial to the liability issue was the question whether under these accident conditions, a properly functioning seat belt would be expected to restrain this plaintiff to the extent that she would not have suffered the injuries she did. The experts disagreed with regard to this issue. However, [813]*813defendant’s experts’ testimony was buttressed by the videotapes depicting the dummies of a dissimilar size, traveling at a dissimilar speed, hitting the steering wheel. Like the Court did in Sumner, supra, I conclude that the evidence addressed a major issue in the case and was visually compelling.10 Therefore, the admission of the tapes was not harmless.
H
In view of our disposition of the videotape issue, we need not address plaintiff’s claim that the trial court erred in permitting Peruski to testify [814]*814regarding occupant kinematics after defense counsel stated during discovery that he would not testify in this area. Finally, we do not address plaintiff’s argument that the trial court’s enhanced injury instructions were erroneous. This issue was addressed in Sumner, and Sumner should provide adequate guidance on remand.
We reverse and remand for new trial. We do not retain jurisdiction.