McWILLIAMS, Circuit Judge.

This is a diversity action for damages arising out of a two-car automobile collision which occurred on August 5, 1970, on U.S. Highway 89, just north of the town of Sevier in Sevier County, Utah. Mr. and Mrs. Don Julander and their four minor children were in one vehicle which was being driven by Don Julander in a southerly direction on Highway 89. David Surber, age nineteen, was driving the other vehicle involved in the collision, namely, a 1968 Ford Bronco, in a northerly direction on Highway 89. David Surber, age nineteen, was driving the other vehicle involved in the collision, namely, a 1968 Ford Bronco, in a northerly direction on Highway 89. Just north of Sevier, Utah, the Ford Bronco crossed over the center line of the highway into the path of the Julander vehicle, and a head-on collision occurred which had disastrous results for the Julander family. Mrs. Julander, who was approximately seven months pregnant, died as a result of injuries sustained in the collision. Mr. Julander and his four children were all injured in the accident, with several of the children sustaining severe injuries.

Don Julander and his four children brought an action for the wrongful death of Mrs. Julander, as well as for their own injuries, against the driver of the other car, David Surber, and the later's father, who owned the Ford Bronco, and the Ford Motor Company, the manufacturer of the Bronco. Their claim against the Surbers was subsequently settled and the case came on for trial against the Ford Motor Company only.

The Julanders' claim against Ford Motor Company was based on alleged negligence by Ford in its design, manufacture and assembly of the Bronco and, alternatively, their claim for relief was also based on the doctrine of strict products liability. Trial by jury resulted in a verdict awarding the Julanders $200,000 for the wrongful death of Mrs. Julander, and the additional sum of $215,995.26 for their own personal injuries. Ford now appeals the judgments thus entered on the jury's verdicts. We reverse, as in our view the trial court committed error in the admission and exclusion of evidence which under the circumstances simply cannot in good conscience be classified as harmless error not affecting substantial rights of Ford. In view of the fact that our disposition of the matter necessitates a second trial, we shall consider matters other than the grounds upon which we reverse, since certain of them, at least, may well arise upon retrial.

The several grounds urged in this court by Ford as grounds for reversal are summarized as follows: (1) Insufficient evidence of (a) negligence on the part of Ford in its design, manufacture and assembly of the 1968 Ford Bronco or (b) that any negligence in design, if such there be, was a proximate cause of the collision in question; (2) error on the part of the trial court in submitting the case to the jury on an alternative theory of strict products liability; (3) the admission by the trial court over objection of evidence of other accidents involving Broncos, and in particular the admission into evidence of seven civil complaints filed by others against Ford; (4) the exclusion by the trial court of evidence sought to be introduced by Ford relating to its design, manufacture and assembly of the Bronco, and in this connection Ford particularly objects to the exclusion by the trial court of Ford's offer of exhibit B; (5) improper final closing argument by counsel for the Julanders wherein he commented upon the damages issue, when counsel for Ford had made no reference to damages in his closing argument; and (6) excessive damages awarded by the jury for the wrongful death of Mrs. Julander. We shall first make brief comment upon certain of the above matters which we deem to be only peripheral.

Not knowing what the evidence concerning the damages flowing from the death of Mrs. Julander will be on retrial, and not knowing whether the jury on retrial will even find in favor of the Julanders, or, if the jury so finds, not knowing what their assessment of damages in connection with such wrongful death will be, we see no good purpose in here determining whether the jury's award of $200,000 for the wrongful death of Mrs. Julander in the first trial of this matter finds support in the record. Accordingly, we decline to review this particular matter.

As concerns the claim by Ford of improper closing argument on the part of counsel for the Julanders, it is sufficient to note that in our view any error in this regard would under the circumstances be only harmless error not affecting the substantial rights of Ford. At the same time, we recognize the general rule that it is improper for counsel for a plaintiff to inject new matter in his final closing argument which should be limited to a rebuttal to defense counsel's closing argument, and nothing more. For an annotation on this subject, see 93 A.L.R.2d 273 (1964). However, as was said in Thomson v. Boles, 123 F.2d 487 (8th Cir. 1941), cert. denied, 315 U.S. 804, 62 S.Ct. 632, 86 L.Ed. 1204 (1942), appellate courts should exercise great caution in setting aside a judgment because of remarks made by counsel in closing argument during a hotly contested case, and even though the argument be improper, the judgment should not be disturbed unless it clearly appears that the remarks in question unduly aroused the sympathy of the jury and thereby influenced the verdict. In the instant case, the comment of counsel was, under the circumstances, improper, but such comment, in our view, was only harmless error and does not justify reversal.

Counsel for Ford also contends that the evidence was legally insufficient to carry the case to the jury on either a theory of negligent design or on the basis of strict products liability and that accordingly the judgments should be reversed and the cause remanded to the trial court with directions to dismiss. As a corollary of the argument Ford also contends that there was no evidence that any possible negligent design caused or contributed to the collision. With this argument we are not in accord. We shall first discuss the "negligent design" aspect of the case.

It was the plaintiffs' theory of the case that Ford negligently designed the 1968 Bronco to the end that at certain speeds, depending upon the size and location of its load, the Bronco had a tendency to "oversteer." "Oversteer," according to several of the plaintiffs' witnesses, means that "very slight changes of the steering wheel produce very sharp changes in the direction of the vehicle." In this general connection, the testimony of David Surber, the driver of the Ford Bronco, should first be commented upon.

According to David Surber, just prior to the accident he was following a slow moving camper on Highway 89 just north of Sevier, Utah, and he determined to pass the camper. Pulling towards the center of the road, Surber testified that he saw the Julander vehicle, and another vehicle as well, approaching from the opposite direction, and that he fully realized he could not pass the camper at this point on the highway and accordingly pulled back into his lane, directly behind the camper. Believing that he was too close to the camper, Surber testified that he applied his brakes, "firmly," but not in any sort of an emergency manner, and turned the steering wheel slightly to the right. Then, according to Surber, the Bronco veered sharply to the right, and when he turned the steering wheel back to the left, the rear end of the Bronco "fishtailed" and he veered across the center of the road into the path of the oncoming vehicle driven by Don Julander. Surber testified that just prior to applying his brakes he was going approximately forty-five to fifty miles per hour. At the time David Surber was driving by himself from California to Colorado and was carrying fishing and camping gear for himself and his parents, who were to join him later in Gunnison, Colorado. The fishing and camping gear were in the back of the vehicle and were estimated by several witnesses as weighing two hundred pounds.

Back to plaintiffs' theory of the case. It was their belief that the 1968 Ford Bronco was "oversteered" because of a combination of factors. In this regard it was contended that the Bronco has a relatively high center of gravity, a solid front axle, rather than independent front suspension, and a very short wheel base, all of which resulted in "oversteer" when driven at a speed of more than forty-five miles per hour with a load concentrated in the rear end of the vehicle.

Our study of the record leads us to conclude that there is sufficient evidence to carry the case to the jury on the question as to whether Ford was guilty of a negligent design of its 1968 Bronco. In thus concluding, although there be other evidence on this issue, we would particularly point to the expert testimony of Kenneth Feder and Dr. Michael Kaplan, and plaintiffs' exhibit 31, all of which in our view require submission of the question of negligent design to the jury.

Feder was the general manager and owner of an "auto rescue" service. He examined the Surber Bronco after the accident and otherwise had a rather extensive experience with the Bronco car. Based on his examination of the Surber Bronco, Feder testified that he found no evidence of any mechanical malfunction that contributed to the accident and that the vehicle was "functioning as intended by design." However, this witness went on to testify that in his opinion the Bronco had a tendency to oversteer under certain circumstances and he explained in detail the basis for his opinion.

Similarly, Dr. Kaplan testified at length regarding what he felt was a defective design of the Bronco resulting in a dangerous handling characteristic. Dr. Kaplan was a college professor of Engineering Mechanics and Mechanical Engineering and had done extensive work, both by way of research as well as by way of practical experience, on highway safety in general and on the Bronco in particular.

As concerns plaintiffs' exhibit 31, Ford produced upon request the written results of a so-called "jury evaluation report" prepared in connection with test rides made by Ford of the Bronco in January 1965, which was shortly before the Bronco was itself placed on the market. Upon trial, the plaintiffs called as an adverse witness for cross-examination one Richard Vogh, who was the supervisor of the evaluation section of the light truck division of Ford, and as a part of their cross-examination of this witness introduced, without objection, the aforesaid exhibit 31.

According to Vogh, the Bronco was tested by various of Ford's employees in January 1965, each of whom would then rate the Bronco's performance on a 10 point scale, a rating of 7 to 10 being "acceptable," a rating of 5 and 6 being "borderline acceptable," and 1 to 4 being "unacceptable." Such tests were not only made on the Bronco, but were also run on similar vehicles sold by Ford's competitors as well as Ford's own F-100 series truck. According to exhibit 31, which was a compilation of the results of this testing, the Bronco rated 7 or above in all categories other than "handling," where it was rated below 7 and therefor was in only the "borderline acceptable" category. And as relates to directional stability and handling, the Bronco tested below all other vehicles thus tested. Needless to say, counsel for the Julanders placed considerable emphasis on exhibit 31. More will be said later about exhibit 31 in connection with the rejection by the trial court of Ford's exhibit B. It is sufficient to say here without further lengthening this opinion that the testimony of the experts Feder and Kaplan, coupled with exhibit 31, was sufficient to warrant, and indeed require, submission to the jury of the issue of negligent design. And on the same basis the issue of causation was also a jury question.

As indicated, the trial court also submitted the case to the jury on the basis of strict products liability and Ford asserts that such was in error. In this regard, it is Ford's position that Utah has not as yet adopted the doctrine of strict products liability and that accordingly it was improper for the trial court to instruct the jury on such. We disagree.

It is quite true that the Utah Supreme Court has not yet decided whether there is in Utah strict liability in tort for defective products. See Perkins v. Fit-Well Artificial Limb Co., 30 Utah 2d 151, 514 P.2d 811 (Oct. 4, 1973). However, in such circumstance a federal trial court must determine for itself what result would probably be reached were the question to be litigated in a state court, and in making its determination the federal court may look to all resources, including decisions of other states, as well as the resident state, federal decisions, and the general weight and trend of authority. Cottonwood Mall Shopping Center, Inc. v. Utah Power Light Co., 440 F.2d 36 (10th Cir. 1971), cert. denied, 404 U.S. 857, 92 S.Ct. 107, 30 L.Ed.2d 99 (1971). And once a federal trial court has made its determination, we, as the reviewing court, are entitled to lean on the judgment of the federal trial judge as being knowledgeable and persuasive in the determination of the law of his resident state and his resolution of the matter should not be disturbed by us unless clearly erroneous. Binkley v. Manufacturers Life Insurance Co., 471 F.2d 889 (10th Cir. 1973), cert. denied, 414 U.S. 877, 94 S.Ct. 130, 38 L.Ed.2d 122 (Oct. 9, 1973), and Vaughn v. Chrysler Corporation, 442 F.2d 619 (10th Cir. 1971), cert. denied, 404 U.S. 857, 92 S.Ct. 106, 30 L.Ed.2d 98 (1971).

Under the authorities above cited, we defer to the trial court's determination that the doctrine of strict products liability applies in Utah. In this same connection, Ford additionally claims that even if the doctrine of strict products liability applies in Utah, such doctrine applies only to the user or consumer, and should not be extended to a bystander, such as the Julanders, and that in any event the doctrine does not apply in a case of alleged negligent design. Again, we defer to the trial court's judgment on the matter. Certainly there is respectable authority that the doctrine does apply to one situated as were the Julanders and that the doctrine is not rendered inoperative in a negligent design case.

In its discussion of strict products liability, the Restatement (Second) of Torts § 402A takes no position on the questions of either bystander recovery or of application of the doctrine to design defects. However, affirmative answers to both questions by the California Supreme Court, which fathered the products liability doctrine, as well as similar recent decisions in other circuits, impress us as being persuasive of that doctrine's post-Restatement evolution.

In connection with the application of the doctrine to a bystander, see such cases as Passwaters v. General Motors Corp., 454 F.2d 1270 (8th Cir. 1972); Wasik v. Borg, 423 F.2d 44 (2d Cir. 1970); and Elmore v. American Motors Corp., 70 Cal.2d 578, 75 Cal.Rptr. 652, 451 P.2d 84 (1969). For an annotation on this particular subject, see 33 A.L.R. 3d 415 (1970).

As concerns the application of the doctrine of strict liability to design defects, see such cases as Runnings v. Ford Motor Company, 461 F.2d 1145 (9th Cir. 1972); Werth v. Clark Equipment Co., 457 F.2d 1262 (9th Cir. 1966), cert. denied, 409 U.S. 876, 93 S.Ct. 127, 34 L.Ed.2d 129 (1972), and Pike v. Frank G. Hough Co., 2 Cal.3d 465, 85 Cal.Rptr. 629, 467 P.2d 229 (1970).

We now come to those matters which in our view necessitate reversal. As mentioned above, the Julanders as a part of their case called as an adverse witness Richard Vogh, the supervisor of the evaluation section of the light truck division of Ford. During their cross-examination of Vogh, counsel for the Julanders inquired as to whether he (Vogh) was aware of any "legal or written complaints by users of Broncos about steering problems." Vogh replied that he was not. Counsel next inquired as to whether Vogh knew that "numerous legal actions are on the file against Ford." Objection to that question was sustained, whereupon counsel for the Julanders offered copies of seven complaints filed in various courts around the country against Ford, all actions being based on steering problems of one sort or another with the Bronco. These complaints had been produced by Ford in pre-trial discovery proceedings. Over vigorous objection, the trial court admitted into evidence these seven complaints on the grounds that such constituted notice to Ford of steering problems with its Bronco.

In this court the Julanders would sustain the action of the trial court in admitting these seven complaints not only on the ground that such constituted notice to Ford, but also on the ground that such constituted impeachment of the witness Vogh, and was additionally admissible as tending to show negligent design. And in any event, says counsel, any possible error in the admission of these seven complaints was only harmless error. In our view of the matter, the trial court erred in the admission of the seven complaints and this error, particularly when coupled with the trial court's erroneous rejection of Ford's exhibit B, which will be developed shortly, simply cannot be brushed off as harmless error.

Before considering the various grounds upon which it is here argued that the seven complaints were properly received in evidence, we should first examine the complaints in a bit of detail. These seven complaints, which collectively constituted plaintiffs' exhibit 32, were copies of complaints setting forth causes of actions against Ford for injuries sustained as a result of alleged steering problems of one sort or another with a Bronco. The allegations in the seven complaints were understandably couched in general language to the end that it is difficult to tell whether the precise problem encountered by the seven complainants was the same as that claimed to have been encountered by Surber. However, all did pertain to a steering or braking problem.

There is nothing in the record to indicate when Ford was served or otherwise received notice of any of these seven actions. However, the record does indicate that two of the actions were filed before the date of the accident here in question, namely, August 5, 1970, the most recent of the two having been filed on July 20, 1970, only about two weeks before the Julander accident. The record fails to reveal when two of the remaining five actions were filed, and the record affirmatively indicates that the remaining three actions were filed after the Julander accident. As indicated, the actions were brought all over the country, from Tennessee to California. The trial court admitted exhibit 32 on the grounds that such constituted notice to Ford. Generally, where negligence is in issue, prior complaints to a defendant concerning an allegedly hazardous condition are admitted as being probative of the defendant's knowledge. This knowledge, in turn, then operates as a standard against which can be tested the reasonableness of the defendant's conduct with regard to the allegedly hazardous condition. See New York Life Ins. Co. v. Seighman, 140 F.2d 930 (6th Cir. 1944).

Assuming, then, that evidence concerning notice was otherwise germane to the issues raised in this case, it would seem that to be admissible the notice to Ford must have been given prior to the date of the Julander accident (August 5, 1970), and sufficiently so, to the end that Ford could have taken steps to remedy the situation. Here, the record fails to disclose when Ford was served with, or otherwise notified of, any of the seven complaints. Though two of the actions were filed prior to August 5, 1970, three were not filed until well after the date of the Julander accident, and the record does not disclose when the remaining two actions were brought. Without belaboring what we deem to be the obvious, the admission of exhibit 32 simply cannot be upheld on the basis of a "notice" theory, because such did not constitute prior notice. See Ozark Air Lines, Inc. v. Larimer, 352 F.2d 9 (8th Cir. 1965), where in a slip and fall case while boarding an airplane, it was held to be prejudicial error to admit testimony of a fall occurring on a subsequent date to another person while boarding defendant's airplane.

In the trial court, counsel for the Julanders never got the opportunity to state the grounds upon which he offered exhibit 32. In this court, counsel, in addition to adopting the grounds upon which the trial court admitted the exhibit, urges two alternative grounds. In this regard, counsel suggests that the exhibit was admissible as tending to impeach the testimony of Vogh. The record does not support such contention. Vogh, both before and after the admission of exhibit 32, testified that he, through personal knowledge or by hearsay, knew nothing at all about the various complaints contained in exhibit 32, and there is absolutely no evidence which in anywise disputes his testimony in this regard. Accordingly, exhibit 32 in itself nowise contradicts or impeaches Vogh's testimony.

Counsel also suggests that exhibit 32 is itself probative evidence of negligent design on the part of Ford in its design of the 1968 Bronco. Evidence of "other accidents" is sometimes admissible to prove primary negligence, but such evidence should be carefully examined before being received to the end that the circumstances of the "other accident" bear similarity to the circumstances surrounding the accident which is the subject matter then on trial. Prashker v. Beech Aircraft Corporation, 258 F.2d 602 (3d Cir. 1958). Such evidence in the instant case is singularly lacking. In sum, the admission of exhibit 32 was erroneous and we shall discuss later the further contention by counsel that any error in this regard was harmless. We shall next discuss the exclusion by the trial court of Ford's exhibit B.

As a part of pre-trial discovery, the Julanders took the deposition of Vogh in Detroit, Michigan, some eleven months prior to the trial of the case. At that time, counsel asked Vogh to produce certain documents of Ford relating to the Bronco and it was in this setting that Vogh produced the report which upon trial was introduced as plaintiffs' exhibit 31. As referred to above, exhibit 31 indicated that when tested in January 1965 the Bronco was given by Ford itself a "borderline acceptable" rating as to handling and directional stability.

At the time of the taking of this deposition, counsel inquired as to whether Vogh knew of any other reports concerning the testing of the Bronco before it was placed on the market. Vogh said he knew of no such further testing or a report thereof, whereupon counsel asked that Vogh make a further search to see if there were any subsequent reports of a similar nature. Thereafter, Vogh discovered a further report of testing conducted in June 1965 in a file cabinet of one of his subordinates. Vogh did testify that subsequent to the January testing certain modifications had been made and in the June testing, according to the report, the Bronco received a much more favorable rating when tested as to handling and directional stability. Counsel for Ford immediately advised counsel for the Julanders of the existence of this subsequent report and furnished counsel with a copy of the same, some ten months prior to trial. Upon trial, counsel for the Julanders indicated that he had in fact not received this second report. However, this breakdown in communication between counsel, if such it be, was not the basis for the trial court's refusal of the offer of the report concerning the tests conducted in June 1965, which document upon trial was, as indicated, marked as Ford's tendered exhibit B. Accordingly, any refusal to admit Ford's exhibit B cannot be upheld upon the basis of a misunderstanding between counsel.

As concerns exhibit B, Vogh testified that he did not personally conduct any of the testing the results of which were set forth in the exhibit. However, according to Vogh, exhibit B came from the files of one of his subordinates, a project engineer, who was under his (Vogh's) supervision and control, and Vogh further testified that the subordinate's duties included testing of the type recorded on the report. Vogh explained that Ford had no central filing system, as such, which accounted for the project engineer maintaining his own files. Exhibit B was on a company form which, again according to Vogh, was customarily used to record the results of evaluation testing, and the form itself was one in use in June 1965. Vogh went on to testify that Ford had a policy of destroying such reports three years after their date of compilation, and that it was only by accident that exhibit B was still in existence, it being more than three years old. All of which, we note, was equally true with regard to plaintiffs' exhibit 31.

Ford sought to introduce exhibit B on the ground that it was a record made in the ordinary course of business and therefore admissible under 28 U.S.C. § 1732(a) as an exception to the hearsay rule. In ruling on the admissibility of exhibit B, the trial court candidly conceded that it was not entirely "clear about the matter," but nonetheless denied the offer on the ground that the report couldn't have been kept in the usual course of business since, under company policy, it should have been destroyed because it was more than three years old and had been preserved only by accident. In other words, the basis for the trial court's refusal to permit the introduction into evidence of Ford's exhibit B was that by company rule it should have been destroyed, and by accident it was not; ergo, the record could not have been kept in the ordinary course of business, as in the ordinary course of business it would have been destroyed. This reasoning does not appeal to us and no case has been drawn to our attention which supports it.

In Palmer v. Hoffman, 318 U.S. 109, 63 S.Ct. 477, 87 L.Ed. 645 (1943), the Supreme Court declared that the provisions of 28 U.S.C. § 1732(a) should be given a liberal interpretation to the end that the "anachronistic rules" which gave rise to its need should be done away with. In United States v. Re, 336 F.2d 306 (2d Cir. 1964), cert. denied, 379 U.S. 904, 85 S.Ct. 188, 189, 13 L.Ed.2d 177 (1964), it was observed that the theory behind 28 U.S.C. § 1732(a) is to avoid the need of producing the person who actually participated in the transaction. Hence, the fact that Vogh himself did not participate in the testing conducted in June 1965, and the recording of the results of such testing, is not in itself reason to reject the offer of exhibit B.

To the same effect as Re, see Sabatino v. Curtiss National Bank of Miami Springs, 415 F.2d 632 (5th Cir. 1969), cert. denied, 396 U.S. 1057, 90 S.Ct. 750, 759, 24 L.Ed.2d 752 (1970), where it was held that 28 U.S.C. § 1732(a) does not require that the foundation testimony come from one who himself kept the records or had supervision over them. In Sabatino, it was also observed that federal rules and practice favor the admission of evidence, rather than its exclusion, if the proffered evidence has any probative value at all. The probative value of exhibit B is at once apparent to all, not only for its own intrinsic probative value but also because the exhibit tends to soften and ameliorate the devastating impact of plaintiffs' exhibit 31, which for obvious reasons was relied on most heavily by the Julanders in their effort to convince the jury that Ford had foisted on the general public a vehicle whose handling and directional stability qualities were borderline, at best. As noted above, plaintiffs' exhibit 31 also came from the files of Ford and it too was more than three years old. Suffice it to say, the trial court erred in refusing to admit into evidence exhibit B on the grounds that under company policy it should have been destroyed, and only by accident had survived the shredder.

Counsel for the Julanders alternatively argues that even if the trial court erred in admitting exhibit 32 and in rejecting exhibit B, the error was only harmless error not affecting any substantial right of Ford. We agree that if this error be deemed harmless error, then under 28 U.S.C. § 2111 and Fed.R.Civ.P. 61 we should not disturb the judgment. And we are further aware that appellate courts have been more quick to find harmless error in the case of the erroneous admission or exclusion of evidence than in other areas. Our search reveals that the test for determining the existence of harmless error, or no, is neither easy to state nor easy to apply, and in the last analysis such resolution generally depends upon the facts and circumstances of the case at hand. For background material on this matter, see Wright and Miller, Federal Practice and Procedure § 2881.

In our considered view, the error in the instant case cannot be glossed over by labeling it as only harmless error. The fact that seven others had brought suit against Ford over the Bronco was of course damaging to Ford. And in this regard counsel in closing argument advised the jury that they had only seen the "tip of the iceberg." Similarly, exhibit B was an important piece of evidence from the standpoint of Ford. It shows that when tested in June 1965 the Bronco as concerns handling and steering tested 8 on the scale of 10, well within the "acceptable" category, and higher incidentally than three of its competitors. Additionally, exhibit B tended to neutralize the impact of exhibit 31, which the Julanders relied on so much. This error was not harmless, and therefore we have no choice but to reverse.

We of course know not what the evidence will be on retrial. So, recognizing that the evidence on retrial may not be the same as adduced upon the first trial of this matter, we reverse, and remand with directions that the retrial of this case be consonant with the views herein expressed.

Judgment reversed.