In this products liability case, the plaintiffs appeal from the district court's action in directing a verdict for the defendant. More specifically, plaintiffs contend the trial judge erred in excluding five proffered exhibits on the grounds of hearsay, irrelevancy, and unfair prejudice. We affirm.

* On July 16, 1982 Linda Emerson, accompanied by Jerome McNair, was driving north along Alabama Avenue, S.E., in the District of Columbia in her 1979 Chrysler LeBaron. Upon approaching the intersection of Alabama Avenue and Irving Place, S.E., the car pulled abruptly to the right and crashed into a telephone pole, injuring both Emerson and McNair. Emerson later died from the injuries she sustained during the crash.

The plaintiffs in this action--Harriet Brooks (the administratrix of Emerson's estate)--and McNair--claim that the car accident was caused by brake piston seizure of the right front wheel. Brake piston seizure occurs when, after the driver has released the brake pedal, the piston remains in contact with the wheel rotor, thereby preventing rotation of the wheel on the axle. The car will then pull in the direction of the non-rotating wheel. The plaintiffs further claim that the alleged brake piston seizure in the 1979 LeBaron resulted from defectively designed brakes. The alleged defect is the "lip-in" dustboot which, according to the plaintiffs, permits corrosive material to enter the caliper bore and retard piston retraction during and after braking.

In support of their theory that the lip-in dust-boot constituted a design defect and was the cause of the accident and that Chrysler knew about the defect but failed to warn customers or to correct the defect, the plaintiffs intended to present to the jury five exhibits (Exhibits 16A, 16B, 17, 26, and 34) relating to a 1978-80 National Highway Traffic Safety Administration, ("HSA"), investigation into brake piston seizure in 1976-80 Chrysler vehicles as evidence of "prior occurrences of piston seizure under substantially similar circumstances." Plaintiffs' Pretrial Brief at 16. Exhibit 17 consisted of 330 consumer complaints. Most of these complaints were in the form of questionnaires that HSA mailed to Chrysler car owners pursuant to its investigation of the brake piston seizure problem.1 In the questionnaires, the owners indentified the year, make, and model of their Chrysler cars, the mileage, and the composition of the steering and braking systems; the owners also detailed their complaints about problems with the vehicle and how those problems interfered with their operation of the car. Exhibit 34 consisted of 98 questionnaires selected from Exhibit 17 by the plaintiffs' expert witness, Dr. James A. Kirk.

Exhibits 16A and 16B were documents obtained from HSA files. Exhibit 16A was comprised of correspondence between HSA and Chrysler regarding the problem of brake piston seizure. Exhibit 16B consisted of two internal memoranda written by HSA's safety defects engineer in which he described the investigation and the consumer complaints.

Exhibit 26 is Chrysler's internal file on the HSA investigation. In addition to the correspondence contained in Exhibit 16A, Exhibit 26 also contained correspondence with the Canadian government's department of transportation, and internal Chrysler memoranda detailing Chrysler's own investigation into the problem and its strategy to deal with it.

Chrysler moved seeking in limine to exclude all five exhibits. It argued that the five exhibits were inadmissible because they were hearsay and because the plaintiffs had not shown that they represented instances of substantially similar prior occurrences; that even if they did represent such instances, they were inadmissible under Fed.R.Evid. 403 because whatever minimal probative value they had was substantially outweighed by the risk of undue prejudice. Chrysler added that Exhibits 16A and 26 were inadmissible because they did not constitute admissions within the meaning of Fed.R.Evid. 801(d)(2) and that Exhibit 16B was inadmissible because it is not a public record within the meaning of Fed.R.Evid. 803(8)(C). Finally Chrysler argued that, even if Exhibit 16B were a public record, it should be excluded because the information forming the basis for the two memoranda was inherently unreliable. The plaintiffs in turn filed their own motion in limine seeking to exclude all references to the fact that Emerson and McNair were not wearing their seatbelts at the time of the accident.

Essentially agreeing with all of Chrysler's arguments, the district judge granted Chrysler's motion in full.2 The district judge, however, offered to reconsider his ruling with respect to several of the questionnaires contained in Exhibit 34 if the plaintiffs would have the owners submitting those questionnaires testify. He further ruled that the plaintiffs' expert, Dr. Kirk, could not rely on any of these exhibits when testifying, but he offered to reconsider this ruling when Dr. Kirk was examined on voir dire. The district judge also granted the plaintiffs' motion to exclude all references to Emerson's and McNair's failure to use seatbelts.

After presenting two witnesses who testified as to the chain of custody of the car after the accident, the plaintiffs' counsel rested the case apparently because he believed that the judge's evidentiary rulings had severely undermined his case and because Dr. Kirk had indicated that he could not testify that there was a design defect if he could not rely on the excluded exhibits.3 Chrysler moved for a directed verdict, which the district judge granted after discussing with the plaintiffs individually the nature of his rulings and the effect of their counsel's action on Chrysler's settlement offer. The plaintiffs' appeal is based on the trial judge's evidentiary rulings; Chrysler cross-appeals from the order excluding evidence of failure to use seatbelts.

At the hearing held on the parties' motion in limine, Chrysler's primary argument in support of exclusion of all five exhibits was that they did not represent examples of substantially similar prior occurrences because the 1978-80 HSA investigation related to a distinct cause of brake piston seizure. According to Chrysler, the investigation revealed only that the lip-in dust-boot was difficult to install properly in its groove or that it was easily pulled out of groove by brake friction. The out-of-groove dust-boot resulting from misinstallation or friction permitted water and other materials from the road (e.g., salt) to seep into the caliper bore causing corrosion on the side of the caliper bore on the land area between the dust-boot and the hydraulic seal. This corrosion could ultimately cause the piston to retard during contraction. Chrysler argued that the exhibits generated by the investigation were therefore irrelevant, see Fed.R.Evid. 401, because the plaintiffs' claim was that the lip-in dust-boot was defective per se in that it failed to prevent corrosive materials from seeping into the bottom of the caliper bore even when it was properly seated in its groove.

Chrysler further argued that the exhibits were properly excludable under Rule 4034 because they were minimally probative on the issue of whether the Emerson vehicle suffered from the particular design defect alleged by the plaintiffs and whether the accident was caused by this defect, and the evidence was likely to be unfairly prejudicial to Chrysler and confusing to the jury.

In support of their claim that the exhibits represented substantially similar prior occurrences and were probative of the central issues in this case, the plaintiffs noted that the subject of the HSA study was 1976-1980 model Chrysler automobiles (including the LeBaron model) fitted with the lip-in style dust-boot mounted on a 2.75-inch-diameter sliding plastic piston and that the Emerson vehicle was a 1979 Chrysler LeBaron fitted with that particular dust-boot. They further argued that the HSA study showed that brake piston seizure was caused by corrosion forming on the caliper bore which caused the vehicle to pull sharply in the direction of the "frozen" wheel. In the instant case, their examination of the Emerson LeBaron completed shortly after the accident revealed that there was corrosion on the bottom of the caliper bore and that the right front wheel was "frozen" on its axle. In addition, the plaintiffs asserted that their witnesses would testify that the Emerson LeBaron pulled sharply to the right--the direction of the "frozen" wheel--shortly before smashing into the telephone pole.

In response to Chrysler's argument that the HSA study was narrowly confined to the problem of out-of-groove dust-boots, the plaintiffs relied on several extracts from the exhibits which the plaintiffs argued demonstrated that the HSA investigation was concerned with piston seizure resulting from a defectively-designed dust-boot that permitted corrosive products to bypass even a properly-seated dust-boot. The first such extraction was a warranty summary drawn from Chrysler's May 8, 1979 letter (contained in Exhibits 16A and 26) to the HSA responding to questions posed by it. The warranty information was set forth in a chart. Column 1 of the chart identified the model year; column 2 identified the model; column 3 identified one warranty code ("Replace caliper 05-81-00") and was itself further broken down into subcolumns labelled "dust-boot misinstalled" and "piston seized"; and column 4 identified another warranty code ("Replace caliper 05-81-05 piston seized"). The chart thus identified the number of warranty claims received by Chrysler from dealers against each of the relevant warranty codes for each model of car. Plaintiffs argued that column 3 demonstrates that only a small percentage of piston seizure was caused by misinstallation of the dust-boot because the subcolumn "dust-boot misinstalled" only accounted for a small percentage of "replace calipers," whereas the "piston-seized" of column 4 and the subcolumn of column 3 accounts for a much larger percentage of "replace calipers."

The plaintiffs also relied on a drawing contained in Exhibit 16A, as well as in 16B, made by HSA's safety defects engineer in which he identified the "area of corrosion" on both sides of the hydraulic seal. The plaintiffs also observed that, at least once, when identifying the source of the problem, Chrysler referred to a "malfunctioning dust-boot" which "allows water to enter the cylinder bore outside of high-pressure seal" (App. at 238, 341) and that Chrysler's engineers referred to the dust-boot problem as a "basic design problem" (App. at 386). Both parties advance these same arguments on appeal.

Evidence of prior instances is admissible on the issues of the existence of a design defect and a defendant's knowledge of that defect only if a plaintiff shows that the incidents "occurred under circumstances substantially similar to those at issue in the case at bar." McKinnon v. Skil Corp., 638 F.2d 270, 277 (1st Cir.1981). Accord Hale v. Firestone Tire & Rubber Co., 756 F.2d 1322, 1332 (8th Cir.1985); Edwards v. Consolidated Rail Corp., 567 F. Supp. 1087, 1106 n.29 (D.D.C.1983), aff'd mem., 733 F.2d 966 (D.C.Cir.), cert. denied, --- U.S. ----, 105 S. Ct. 252, 83 L. Ed. 2d 189 (1984). Even when substantial similarity of circumstances is proven, the trial judge has broad discretion to exclude those exhibits under Rule 403. McKinnon, 638 F.2d at 277. See, e.g., Borden, Inc. v. Florida East Coast Ry., 772 F.2d 750, 755 (11th Cir.1985).

We agree with Chrysler that the trial judge did not abuse his discretion in excluding these exhibits under Rule 403.5 The exhibits are minimally probative on the issue of whether the Emerson vehicle suffered from the defect alleged by the plaintiffs and whether that defect caused the accident. We have reviewed these exhibits in considerable detail, and we can find nothing in them that expressly supports the plaintiffs' claims that a properly-seated dust-boot permits the passage of corrosive materials and that it therefore may have caused the accident in this case. In those exhibits the only cause noted for brake piston seizure is the out-of-groove dust-boot. (See e.g., App. at 281, 307, 309, 315, 318, 323, 358, 359, 361, 364, 383, 386, 397, 401.) See also App. at 257 (Design changes6 undertaken to "minimize any potential for misassembly and/or roll out of the boot seal.").7

Nor does the warranty summary support the plaintiffs' claims. At the hearing, Chrysler gave a detailed explanation of the warranty information. Counsel stated that columns 3 and 4 represented warranty codes under which the dealer making repair work would claim reimbursement for certain types of work. Column 3 represented the old warranty code under which the dealer would claim a smaller amount when the caliper assembly needed replacement because the dust-boot was improperly installed (subcolumn 1) or a greater amount when the assembly had to be replaced because the piston had become seized due to an out-of-groove dust-boot resulting from misinstallation or internal brake friction (sub-column 2). Column 4 represented a new warranty code introduced in 1977 which was only applicable to the piston seizure problem. For this reason, no warranty data for 1976 appeared under the new code, while warranty data under the old code dropped dramatically in 1977 and decreased thereafter as dealers adjusted to the new warranty codes. In addition, Chrysler stated that the preparers of the warranty summary would testify that this was the proper interpretation of the data. Chrysler had also identified this distinction in the two warranty codes in its answers to the plaintiffs' interrogatories, noting that each subcolumn in column 3 had its own identifying digits in addition to the code 05-81-00; it had also suggested in an earlier letter to the HSA that the two codes, old and new, existed.

In contrast, the plaintiffs presented no argument for their interpretation of the warranty summary, even though they had the opportunity to depose the preparers of the document. And they present no new arguments on appeal. In addition, their interpretation conflicts with the actual wording of the chart that the warranty code 05-81-05 was simply inapplicable in the year 1976. Moreover, nothing on the face of that summary or in the document containing the summary suggests that the two subcolumns of column 3 are mutually exclusive, or that the "piston seizure" referred to in subcolumn 2 of column 3 could not have been caused by dust-boot misinstallation or an out-of-groove dust-boot from whatever cause. Indeed, at one point, the plaintiffs' counsel appeared to concede that the subcolumns of column 3 were not mutually exclusive. See App. at 778.

The plaintiffs' other arguments in support of their claim that the subject of the HSA study was substantially similar to the defect alleged in this case and hence probative of central issues are also unpersuasive. The diagram identifying corrosion build-up on both sides of the hydraulic seal is not only based on unverifiable sources of information,8 but it is also only relevant insofar as it undermines Chrysler's claim that corrosive materials cannot bypass a properly-seated hydraulic seal. The threshold issue here, however, is whether the HSA investigation suggests that corrosive materials seep past both properly and improperly-seated dust-boots, and nothing in that diagram establishes that corrosive materials enter the caliper bore through a properly-seated dust-boot.9 Moreover, the plaintiffs' claim is that corrosion formed in the bottom of the caliper bore; at most, this diagram shows that corrosion forms on the side. Furthermore, in the August 24, 1979 letter in which the Chrysler engineer refers to a "basic design problem," that same engineer notes that the problem "was not entirely due to improper boot installation ... [but also to] friction between the upper lip and surface of the piston [that] tends to pull the boot out-of-the groove." (emphasis added).

The plaintiffs stated that their expert would testify at trial that the subject of the HSA study was the same defect they alleged existed in the Emerson LaBaron. But the plaintiffs were called upon to establish the probative value of their proffered exhibits before trial and in the ordinary course of events substantial similarity or identity of circumstances must be established before such exhibits will be admitted.10 0] In this regard, we fail to see why the plaintiffs' expert could not have given this testimony as an offer of proof at the proceedings on the motions in limine.11 See Ramos v. Liberty Mutual Ins. Co., 615 F.2d 334, 339 (5th Cir.1980), cert. denied sub nom. Rucker Co. v. Shell Oil Co., 449 U.S. 1112, 101 S. Ct. 921, 66 L. Ed. 2d 840 (1981).

It is also significant that, in arguing for the admissibility of these exhibits, the plaintiffs' counsel appeared to concede that the exhibits identified the out-of-groove dust-boot as the source of the problem. See App. at 738, 743. But see App. at 744. In addition, the design changes were directed toward providing "positive boot seal retention" in the groove, see App. at 280, and there is no evidence that after design changes were implemented the brake piston seizure problem continuel. Finally, these exhibits do not demonstrate that brake piston seizure was a pervasive problem throughout Chrysler's entire fleet of 1976-1980 model cars that caused accidents similar to the one that occurred in this case. These exhibits showed that approximately two percent of Chrysler's cars suffered from the problem and only a few minor accidents had resulted from it.12

Thus, even if the subject HSA had under investigation is conceded to be a "defective design," these exhibits only identify the "design defect" as a dust-boot that, for whatever reason, easily comes out of groove; not a dust-boot that, in every case, permits corrosive matters to enter the caliper bore. As a result, the trial judge was amply justified in concluding that these exhibits were minimally probative on the key issues of plaintiffs' case.

This is particularly true with respect to Exhibits 17 and 34, which contained the questionnaires that formed the basis of the HSA's investigation and the plaintiffs' claim that the lip-in dust-boot caused the Emerson accident. Many of the questionnaires identified brake piston seizure that resulted specifically from an out-of-groove dust-boot or from sources other than the dust-boot (e.g., swollen pistons, defect in the master brake cylinder, warped brake pistons). Others failed to identify the source of the brake piston seizure; still others did not involve brake piston seizure at all. And in none of these complaints did the owner specifically identify corrosion resulting from seepage past a properly seated dust-boot as the cause of the brake piston seizure.

We recognize that these exhibits do not definitively rule out causes other than the out-of-groove dust-boot; see, e.g., App. at 361 (design changes will "reduce" brake piston seizure); App. at 317 (ambiguous test results regarding seepage of corrosive materials by dust-boot); App. at 318 ("We feel that press-on caliper piston dust-boot would alleviate some of the piston sticking problem."); App. at 361 (same); App. at 385 (problem usually associated with out-of-groove dust-boots); and that the defect alleged by the plaintiffs--piston seizure resulting from a poorly-designed dust-boot--is somewhat similar to the "defect" under study by the HSA. We also recognize that evidence of similar incidents may be particularly probative on the issue of whether one particular product contained a specific defect. See Ramos, 615 F.2d at 339; Rhodes v. Michelin Tire Corp., 542 F. Supp. 60, 62 (E.O.Ky.1982). This type of evidence, however, is not particularly probative when, as in this case, it does not even suggest that the defect alleged by the plaintiffs exists or that any of the few reported accidents were caused by that defect, and it only shows the possible existence of another similar, but not identical, defect in approximately two percent of Chrysler's automobiles manufactured during 1976-1980. Cf. Uitts v. General Motors Corp., 411 F. Supp. 1380, 1383 (E.D.Pa.1974), aff'd mem., 513 F.2d 626 (3d Cir.1975) (proffered exhibits minimally probative where they only show accident involving similar, but not necessarily identical, vehicles); Vockie v. General Motors Corp., Chevrolet Div., 66 F.R.D. 57, 61, 63, aff'd mem., 523 F.2d 1052 (3d Cir.1975).

The fact that these exhibits were minimally probative in this case does not automatically justify their exclusion. The Federal Rules of Evidence generally favor the admissibility of relevant evidence, leaving it to the parties to argue the weight that should be given that evidence, see S. Saltzburg & K. Redden, Federal Rules of Evidence Manual at 85 (3d ed. 1982); 1 D. Louisell & C. Mueller, Federal Evidence Sec. 91, at 638 (1977). As a result, Rule 403 requires that the probative value be substantially outweighed by the risk of unfair prejudice, trial delay, and jury confusion before relevant evidence will be excluded.

At least one court has observed that "[p]roof of prior accidents or occurrences are not easily admitted into evidence because they are often the result of unfair prejudice, consumption of time and distraction of the jury to collateral matters." Uitts, 411 F. Supp. at 1383, citing McCormick, McCormick on Evidence Sec. 200 (2d ed. 1972). Although we do not necessarily agree with this observation, it is clear that in this case admission of the exhibits in this case would lead to this very result. With respect to Exhibits 17 and 24, the questionnaires are riddled with complaints about problems with the other automobiles, unrelated to brake piston seizure, by customers dissatisfied with Chrysler's refusal to extend the warranty to cover the repairs of the dust-boot; highly inflammatory remarks about Chrysler, including comments about poor workmanship in general and about Chrysler's failure to cover the problem under the warranty; and third-party hearsay comments by the owners' mechanics and friends regarding the prevalence of brake piston seizure in Chrysler cars. In addition, the other exhibits suggest, as the plaintiffs note, Chrysler was somewhat cavalier in its dealing with the agency in admitting the scope of the problem and the possibility of a safety hazard, and "hardnosed" in its approach to warranty repairs. As such, these exhibits were "unfairly prejudicial" within the meaning of Rule 403. Advisory Committee Note to Fed.R.Evid. 403 (" 'Unfair prejudice' within its context means an undue tendency to suggest decision on an improper basis, commonly, though not necessarily, an emotional.").

Furthermore, it cannot be doubted that the trial inevitably would have been delayed as Chrysler would have attempted to rebut the substance of each of the 330 complaints or to distinguish the nature of the complaints contained therein from the alleged defect in this case. See Ellis v. International Playtex, Inc., 745 F.2d 292, 303, 305 (4th Cir.1984). The trial would have been further delayed by an extensive foray into the nature and extent of the HSA investigation, and the jurors' attention would have been diverted from the specific claim of the plaintiffs to the subject under study by the HSA. See also supra note 10.

We therefore conclude that the district judge properly concluded that the minimum probative value of these exhibits was substantially outweighed by the risk of unfair prejudice, trial delay, and jury confusion and that he therefore did not abuse his discretion in excluding these exhibits under Rule 403. In reaching this conclusion, we also note that the plaintiffs made no effort to delete from these exhibits obviously prejudicial, irrelevant hearsay material. See McCormick, Handbook of the Law of Evidence, at 112 (2d ed. 1972) (trial court can properly exclude an entire exhibit if parts of that exhibit are inadmissible and the proponent of the evidence has made no attempt to differentiate between the admissible and the inadmissible). We also note that the offer of proof made by the plaintiffs at the hearing on the motions in limine as to the probative value of these exhibits was simply inadequate to rebut Chrysler's convincing arguments of low probative value, prejudice, confusion and trial delay, and that the plaintiffs were given another opportunity to make a better offer of proof--an offer which they declined. Under these circumstances, we believe the trial judge was amply justified in excluding these exhibits. A judge understandably might be more reluctant to exclude proffered exhibits under Rule 403 where, as here, those exhibits evidently form the core of the plaintiffs' case. But the plaintiffs simply cannot demand that the judge admit wholesale, unedited prejudicial exhibits that will cause substantial delay in the trial particularly when they make little effort to first convince the court of the probative value of these exhibits.

The plaintiffs also complain that the district judge improperly prohibited their expert from relying on these exhibits in formulating his opinion about the existence of a design defect in the Emerson LeBaron and the cause of the accident. See Fed.R.Evid. 702. This argument is without merit. Shortly before permitting the plaintiffs' counsel to rest his case, the district judge made clear that the scope of Dr. Kirk's testimony was still open to reconsideration when Dr. Kirk was examined on voir dire. See supra at 1193 and note 11.

We rule that the trial judge did not abuse his discretion in excluding the exhibits in question under Rule 403 or improperly preclude the plaintiffs' expert from relying on these exhibits. We therefore do not reach the other issues raised by this appeal and by Chrysler's cross-appeal. The judgment of the district court is affirmed.