KENNEDY, Circuit Judge.

Plaintiff David Engebretsen appeals the judgment entered after a jury verdict for defendant Fairchild Aircraft Corporation in this airplane products liability action. On appeal, plaintiff first argues that the District Court erred in denying his motion for judgment as a matter of law. Further, plaintiff argues that the District Court erroneously denied his motion for a new trial because of the admission of two reports by defendant’s expert witnesses that contained inadmissible and prejudicial hearsay; the admission of expert testimony regarding post-incident tests; and the District Court’s refusal to admit certain government reports offered by plaintiff. For the reasons stated below, we affirm.

This case concerns an emergency landing of a Metro III aircraft manufactured by defendant. Plaintiff, a captain for Comair, alleges that, while attempting to land the plane, he sustained a back injury as a result of the plane’s allegedly defective Stall Avoidance System (“SAS”). During the incident in question, plaintiff and First Officer Sean Belcher were flying a Metro III aircraft, a small plane with a capacity for eighteen passengers and two crew members. The Metro III has a flight control known as the elevator, which controls the pitch attitude of the aircraft. The elevator or pitch trim system is used to adjust the elevator and thereby relieve some of the pressure felt by the pilot on the control column or “yoke.” The plane is also equipped with an SAS. The SAS senses an impending stall and serves to warn the pilot of the stall and attempts to avoid the stall by causing a forward or nose-down pressure on the control yoke.

Plaintiff and Belcher testified to the following events. On September 14, 1986, plaintiff and Belcher were flying the Metro III en route from Indianapolis to the Greater Cincinnati International Airport in Covington, Kentucky. While on their final approach for landing, Belcher, who was operating the plane, noticed a forward pressure in the control yoke. Plaintiff assumed control, aborted the landing procedure, and initiated a go- *725 around procedure. Believing the yoke pressure was caused by the SAS, plaintiff turned the SAS clutch switch to “off’ but the forward pressure was not alleviated. At first, Belcher assumed they were still attempting to land and so continued to push forward on the first-officer yoke, while plaintiff was pulling back on his yoke. Plaintiff continued to climb the aircraft, declared an emergency, and requested landing. As plaintiff proceeded to land, he felt an increasing forward pressure on the yoke, requiring both pilots to fly the plane. Plaintiff also attempted to complete the SAS malfunction checklist, which requires pulling the SAS circuit breakers while maintaining back pressure on the control yoke. After pulling two of the four circuit breakers, the pilots felt another increase in the yoke’s nose-down pressure. Plaintiff added power to overcome the yoke pressure and turned for final approách. They felt yet another increase in the nose-down force. Belcher advised the passengers of the emergency and told them to assume crash positions. Unable to fully control the rate of descent, plaintiff had to maintain a high speed in order to reach the runway. He managed to land the plane at 155 knots, approximately fifty-percent faster than normal landing speed. The forward pressure continued on the yoke until all the SAS circuit breakers were pulled. Plaintiff taxied the plane to the gate and the passengers deplaned. Although not then apparent, plaintiff later experienced back pain which he attributes to the incident. At the time of trial, plaintiff was on disability flight status due to his back injury.

Shortly after landing, plaintiff contacted David Wahn, a Comair avionics mechanic, and told him that the SAS pusher had engaged on final approach. Upon inspection of the aircraft, Comair mechanics discovered approximately one inch of water in the aircraft’s belly in the environs of the servo motor and clutch. The water build-up was caused in part by the failure of the fuselage drainage plugs. The electrical connector which joined the servo motor to the aircraft’s wiring had moisture and corrosion in it. By shorting out certain pins on the electrical connector, the mechanics were able to activate the SAS. The incident was reported to the Federal Aviation Administration (FAA). Because no injuries to person or property were reported, the National Transportation Safety Board (“NTSB”) did not fully investigate the incident, leaving it to the FAA. Several other tests were eventually performed by Comair and Fairchild which are discussed at relevant times below.

Plaintiff filed suit in the District Court for the Eastern District of Kentucky seeking damages for personal injury. The court bifurcated the issues of liability and damages. A jury trial was held on the issue of liability. The jury returned a verdict in defendant’s favor. The verdict form required the jury to answer three questions:

Question No. 1:

Did the Plaintiff prove by a preponderance of the evidence that the airplane was “unreasonably dangerous” in design and that the unreasonably dangerous condition was a substantial factor in causing the incident which he claims caused the accident referred to in the evidence?

The jury answered this question in the negative.

Question No. 2:

Did Plaintiff, David Engebretsen, prove by a preponderance of the evidence that the Stall Avoidance System (SAS) continued to exert force on the controls after the clutch switch was turned off?

The jury answered this question in the negative.

Question No. 3:

Did' the Defendant, Fairchild Aircraft Corp., prove by a preponderance of the evidence that the Plaintiff, David Enge-bretsen, was guilty of negligence which was a substantial factor in causing the plaintiffs injury referred to in the evidence?

The jury answered this question in the negative. The District Court entered judgment in accordance with the verdict. Plaintiffs post-judgment motions were denied.

II.

Plaintiff first assigns error in the denial of plaintiffs motion for judgment as a matter of *726 law. He argues that the undisputed eyewitness testimony establishes that the aircraft experienced a malfunctioning of the SAS; that the pilots took all reasonable steps to disconnect the SAS; and that the SAS continued to exert a nose-down pressure after plaintiff took all such reasonable steps to disconnect it.

The issue raised by a motion for judgment as a matter of law is whether there is sufficient evidence to raise a question of fact for the jury. Warkentien v. Vondracek, 633 F.2d 1, 6 (6th Cir.1980). This Court must apply the same standard on review. Id. at 6-7. Furthermore, “[i]n reviewing the grant of a judgment [as a matter of law], this circuit has long held that the state law standard controls in a diversity case.” Toth v. Yoder Co., 749 F.2d 1190, 1194 (6th Cir.1984). This case is governed by the laws of Kentucky. Under Kentucky law, in reviewing a trial court’s refusal to enter judgment notwithstanding the verdict, “[a]ll evidence which favors the prevailing party must be taken as true and the reviewing court is not at liberty to determine credibility or the weight which should be given to the evidence, these being functions reserved to the trier of fact.” Lewis v. Bledsoe Surface Min. Co., 798 S.W.2d 469, 461 (Ky.1990). We must give the non-moving party the benefit of all favorable inferences. See Calhoun v. Honda Motor Co., 738 F.2d 126, 133 (6th Cir.1984). In a products liability case, the plaintiff must prove that the product was defective and that “the defect was the ‘probable’ cause of the accident as distinguished from a ‘possible’ cause among other possibilities.” Calhoun, 738 F.2d at 133 (quoting Midwestern V.W. Corp. v. Ringley, 503 S.W.2d 745, 747 (Ky.1973)).

Turning to this case, plaintiff disputes the conclusions reached by the jury on the first two interrogatories. The jury found that the airplane was not unreasonably dangerous and that the SAS did not continue to exert force on the controls after the clutch switch was turned off. Plaintiff argues that the evidence clearly shows that the sole cause of the accident was a malfunctioning of the SAS, which exerted forward pressure on the control yoke; that this malfunction was caused by moisture in the aircraft’s belly; and that the accumulation of such moisture was foreseeable to defendant.

Defendant, conceding that the SAS was unintentionally activated by exposure to moisture, defended on the ground that the amount of water in the belly was unforeseeable and/or detectable by proper maintenance inspection, and that the continued increase in the yoke pressure resulted from water-shifting in the belly and the pilots’ misapplication of the pitch trim. Defendant also argues that there was ample evidence to support the jury’s answer as to the second interrogatory that the SAS was deactivated when switched “off.”

We find evidence in the record to support defendant’s position. First, evidence supported a finding that the amount of water in plaintiffs airplane was unforeseeable. Co-mair mechanic David Wahn testified on cross-examination that he had never seen that much water in the belly of one of these aircraft before. David Spare, another Co-mair avionics mechanic who worked on the aircraft with Wahn, stated in deposition that he likewise had never seen that much water in one of these planes. Finally, James Martin, a Comair maintenance mechanic who was sent to drain the water from the aircraft, testified on cross-examination that he had never seen or heard of that much water accumulation in a Metro III before or since the incident. He stated further that “[w]e considered this to be a very unusual event, yes. It was not common.” Plaintiff points out that Fred Reina, a former field representative for Fairchild, said in deposition that it was not uncommon for water to accumulate in a Fairchild aircraft. The resolution of such conflicts in testimony, however, is a question for the jury. Furthermore, with respect to the cause of the water build-up, Heaslip testified for the defense that if Co-mair had properly complied with its own inspection procedure, one inch of water would never have built up. Taking the evidence as true, and drawing inferences in defendant’s favor, the jury was entitled to find that an unusually large amount of water existed in the belly of plaintiff’s plane, and that defendant should not be held responsible *727 for not anticipating such an accumulation when designing the SAS.

The record also contains evidence that the SAS ceased to exert force after the switch was turned off and two circuit breakers had been removed. Defense experts testified that switching the clutch switch to off must, under all circumstances, deactivate the stick pusher. Also, Wahn testified that, during his inspection of the plane following the incident, although he experienced a stick push when he turned the SAS switch on, the SAS deactivated when he turned the switch off. Comair Avionics Manager Jim Dachille, who performed tests on the plane’s clutch switch both at Comair and later at defendant’s facility, testified on cross-examination that whenever the switch was turned to the off position, the SAS pusher was deactivated. In fact, every time Comair mechanics turned the clutch switch off or pulled the circuit breakers, the SAS deactivated, even when the servo motor was completely submerged in water.

It is undisputed that First Officer Belcher continued to push on his yoke for a short period after plaintiff aborted the landing procedure. As discussed above, evidence indicates that there was excess water in the belly and that the SAS deactivated when turned off. In addition, Jack Morgan, one of defendant’s experts who performed flight tests on the plane, testified that only the misapplication of the pitch trim system by the pilots could have caused the nose-down force experienced by the pilots during the incident. This evidence supports a finding that, once the SAS was switched off, the continued force on the yoke was caused by a forward-shifting of the water in the belly and/or pilot error. We conclude that evidence in the record entitled the jury to find for defendant on the first two interrogatories.

III.

Plaintiff next contends that the District Court erred by denying his motion for a new trial. Plaintiff presents three grounds for a new trial, the first of which is that he was prejudiced by the District Court’s belated admission of two reports (Exhibits 48 & 49) by defendant’s experts after closing argument and jury instructions had been given.

Exhibit 48 is a report prepared by Mr. Terry Heaslip. Heaslip is a professional aviation accident investigator hired by defendant to determine the cause of the incident. Heaslip was qualified at trial as an expert in aviation accident reconstruction and investigation. The report contains various “Findings and Opinions.” These are assertedly based on Heaslip’s evaluation of the incident’s circumstances, his knowledge of the aircraft and its history, and his analysis of flight test results. He testified that he reviewed, among other things, the pilots’ statements, various manuals, tapes, wiring diagrams, reports, and the production line of the aircraft.

Exhibit 49 is a “Preliminary Report” prepared by Jack Morgan and R.H. Blackwell, who are employees of defendant. Morgan is defendant’s chief accident investigator. The report summarizes various tests done on the aircraft and its components by Comair and defendant’s employees. It also describes certain flight tests done by Morgan designed to determine the cause of the incident. The report contains a “Findings and Conclusions” section.

During trial, defendant called Morgan and Heaslip as expert witnesses. Both experts testified to their post-incident investigations. 1 Although neither expert referred to his report on direct or redirect examination, plaintiff referred to portions of the exhibits during cross-examination in an attempt to impeach the witnesses with alleged errors in the reports and inconsistencies between the reports and the witnesses’ present knowledge and testimony.

At the close of defendant’s case, defendant moved for admission of, among others, Exhibits 48 and 49. Plaintiff objected to the admission of the reports on the ground that they contained prejudicial hearsay not within any exception. He also argued that although he referred to portions of the reports on *728 cross-examination, that does not entitle defendant to enter the entire documents. The court reserved its ruling. After the parties completed closing arguments and the court gave jury instructions, the court decided to admit Exhibits 48 and 49. At no time prior to that had plaintiff expressed a need for an early decision on the exhibits’ admissibility, nor did plaintiff, once the court decided to admit the exhibits, ask that any part of them be redacted.

In his motion for a new trial, plaintiff reiterated his objection to the exhibits on hearsay grounds. He also argued, as he does now, that the late time at which the court admitted the documents prejudiced his case because he was prevented from asking the court for a limiting jury instruction regarding the reports and from arguing the significance of the reports to the jury in closing. In denying plaintiffs motion for a new trial, the District Court held that the documents were admissible under Fed. R.Evid. 702 & 703 because the allegedly objectionable statements were conclusions by qualified expert witnesses, not factual hearsay statements. Because the witnesses could have referred to their reports on direct examination, the court reasoned, their admission was proper. The court also ruled that the reports were properly admitted under the “completeness doctrine” of Fed.R.Evid. 106 because plaintiff had referred to them on cross-examination. The court further held that any statements in the reports that were objectionable were harmless because nearly every contested finding in the reports was explored extensively on direct and cross-examination of the experts. Finally, the court noted plaintiffs failure to request redaction of any of the statements.

This Court reviews a denial of a motion for a new trial under an abuse of discretion standard. Tobin v. Astra Pharmaceutical Products, Inc., 993 F.2d 528, 542 (6th Cir.), cert. denied, — U.S. -, 114 S.Ct. 304, 126 L.Ed.2d 252 (1993). We wifi find an abuse of discretion only when the Court has “a definite and firm conviction that the trial court committed a clear error of judgment.” Logan v. Dayton Hudson Corp., 865 F.2d 789, 790 (6th Cir.1989). However,

while it is not the province of appellate courts to review the discretion of trial courts if exercised within legal bounds, it is the duty and right of appellate courts to determine whether, in the exercise of the discretion committed to it, the trial judge applied correct legal standards.

Mannino v. International Mfg. Co., 650 F.2d 846, 849 (6th Cir.1981).

We conclude that the District Court applied an incorrect legal standard on one of the grounds on which the court denied plaintiffs motion for a new trial. In particular, the District Court erroneously concluded that Federal Rules of Evidence 702 and 703 permit the admission, on direct examination, of testifying experts’ opinions contained in written documents. Rules 702 & 703 provide:

Rule 702. Testimony by Experts

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. Rule 703. Bases of Opinion Testimony by Experts

The facts or data in the particular case upon which an expert bases an opinion or inference may be those perceived by or made known to the expert at or before the hearing. If of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject, the facts or data need, not be admissible in evidence.

Fed.R.Evid. 702 & 703. Rule 702 permits the admission of expert opinion testimony not opinions contained in documents prepared out of court. See Fed.R.Evid. 702. Rule 703 allows a testifying expert to rely on materials, including inadmissible hearsay, in forming the basis of his opinion. Rules 702 and 703 do not, however, permit the admission of materials, relied on by an expert witness, for the truth of the matters they contain if the materials are otherwise inadmissible. See Paddack v. Dave Christensen, Inc., 745 F.2d 1254, 1261-62 (9th Cir.1984). *729 Rather, “Rule 703 merely permits such hearsay, or other inadmissible evidence, upon which an expert properly relies, to be admitted to explain the basis of the expert’s opinion.” Id. Moreover, such materials should not be admitted if the risk of prejudice substantially outweighs their probative value. See Fed.R.Evid. 403; Nachtsheim v. Beech Aircraft Corp., 847 F.2d 1261, 1270-71 (7th Cir.1988). When inadmissible materials are admitted for explanatory purposes, the opposing party is entitled to a limiting instruction to the jury that the evidence may be considered “solely as a basis for the expert opinion and not as substantive evidence.” Paddack, 745 F.2d at 1262.

In concluding that the expert reports were properly admitted, the trial court reasoned that the experts could have referred to the exhibits during direct examination and therefore defendant could properly move for their admission at a later time. As discussed above, however, Rules 702 and 703 carve out a narrow exception to the rule against the admission of hearsay. Heaslip and Morgan were entitled to testify as to their opinion, and rely on inadmissible evidence. Neither their written opinions nor the materials on which they relied were admissible under Rules 702 and 703. Thus, the District Court’s conclusion that Rules 702 and 703 permit the admission of the reports was erroneous.

The second ground on which the District Court upheld the admission of Exhibits 48 and 49 was the “rule of completeness” embodied in Fed.R.Evid. 106. Rule 106 provides:

Rule 106. Remainder of or Related Writings or Recorded Statements

When a writing or recorded statement or part thereof is introduced by a party, an adverse party may require the introduction at that time of any other part or any other writing or recorded statement which ought in fairness to be considered contemporaneously with it.

Fed.R.Evid. 106. The District Court held that “plaintiff opened the door to admission of these documents by questioning the defendant’s experts as to their content, thus waiving the right to later object.” The court concluded that Rule 106 entitled defendant to have the entire reports admitted.

We conclude that plaintiffs cross-examination of the experts did justify admission of one report, and at least certain portions of the second. With respect to Heaslip’s report, Exhibit 48, admission of the entire report was within the trial court’s discretion. Throughout cross-examination of Heaslip, plaintiff questioned Heaslip about his investigation of the incident. He challenged many of Heaslip’s conclusions and various findings and assumptions on which Heaslip’s conclusions were based. Much of plaintiffs inquiry was directed at statements contained in Heaslip’s report. Plaintiff attempted to show that many statements contained in the report were either inconsistent with Heaslip’s direct testimony or with Heaslip’s present knowledge of the incident’s circumstances. Plaintiff continually - asked Heaslip whether defense-witness Morgan was the source of his information and, at one point, plaintiff asked whether Heaslip had failed to record certain test results because they did not favor defendant’s ease. The tenor of plaintiffs cross-examination was to discredit Heaslip’s direct testimony by demonstrating or implying that Heaslip was biased in favor of defendant; had based his conclusions on facts later learned to be false; had relied extensively on Morgan; and that Heaslip’s conclusions generally varied at different times.

Plaintiffs lengthy impeachment of Heaslip by use of his report entitled defendant to introduce other statements in the report to rebut the charge of inconsistency and bias. Plaintiff argues on appeal that defendant could not introduce prior consistent statements contained in Heaslip’s report because the statements do not satisfy the requirements of Fed.R.Evid. 801(d)(1)(B), which exempts from the definition of hearsay any prior consistent statement which “is offered to rebut an express or implied charge against the declarant of recent fabrication or improper influence or motive.” Fed.R.Evid. 801(d)(1)(B). Plaintiff may be correct that Rule 801(d)(1)(B) is not satisfied here. Although plaintiff did' imply an improper influence, the alleged influence was present at the *730 time of Heaslip’s prior statements and thus the report would not rebut a charge of recent influence. Rule 801(d), however, applies to statements offered for their truth. Unless the rule’s requirements are met, the statements are hearsay if offered as substantive evidence. Here, however, the statements were offered to rehabilitate Heaslip’s credibility. The trial court has greater discretion to admit prior consistent statements to rehabilitate an impeached witness, by clarifying or explaining his prior statements alleged to be unreliable, than if the statements are offered for their truth under Rule 801(d)(1)(B). As the Court of Appeals for the Seventh Circuit stated:

prior consistent statements offered for the limited purpose of rehabilitating a witness’s credibility [are] not subject to the strictures of Rule 801(d)(1)(B).... This use of prior consistent statements for rehabilitation is particularly appropriate where, as here, those statements are part of a report or interview containing inconsistent statements which have been used to impeach the credibility of the witness. Prior consistent statements which are used in this matter [sic] are relevant to whether the impeaching statements really were inconsistent within the context of the interview, and if so, to what extent. This rehabilitative use of prior consistent statements is also in accord with the principle of completeness promoted by Rule 106.

United States v. Harris, 761 F.2d 394, 399-400 (7th Cir.1985) (citations omitted). Thus, the court’s discretion to admit prior out-of-court statements for rehabilitative purposes is not limited by Rule 801 and, therefore, provided they are relevant (Fed.R.Evid. 402), the court may admit the statements for that purpose. In this case, we conclude, the trial court did not abuse its discretion in admitting Heaslip’s report in light of plaintiffs cross-examination. 2

The Morgan report, Exhibit 49, was also admissible for similar reasons although not in its entirety. Plaintiff cross-examined Morgan with reference to his report. Plaintiffs questions related to various post-incident flight tests performed by Morgan. The part of the cross-examination that referred to Morgan’s report focussed on the purpose of the flight tests and the relevance of those tests given that the test conditions differed in some respects from the incident on plaintiff’s plane. Plaintiff suggested that Morgan designed the tests to achieve a result that placed blame for the accident on the plaintiff. Of the six statements in Morgan’s report that plaintiff objects to, two of them were inquired into on cross-examination. These statements were conclusions about the cause of the extreme forces on the control yoke allegedly experienced by plaintiff during the incident. Because plaintiff attacked the bases of these conclusions and Morgan’s motive for the tests, the District Court could admit, for rehabilitative purposes, the portions of the report that pertained to these issues. The remaining four statements in the report to which plaintiff objects were not within the scope of plaintiff’s cross-examination and do not appear to be rehabilitative although we do not perhaps have as full an appreciation of all the evidence as did the trial judge. These statements were conclusions that water in the plane’s belly was caused by improper maintenance; that there was great confusion in the cockpit during the incident; and that the plane’s SAS computer and squat switch were fully operational. We conclude that these statements, which favor defendant’s case, do not appear to have been admissible even for a limited purpose.

Two further questions then arise as to whether plaintiff forfeited his right to exclude these statements by failing to ask for their redaction and, if not, whether the error warrants a new trial. The party opposing *731 the admission of statements within a document has the burden of requesting redaction and forfeits the right to exclude the statements if he fails to make the specific request. See Camps v. New York Transit Authority, 261 F.2d 320, 322 (2d Cir.1958). In this case, plaintiff objected to introduction of the entire reports on the ground of hearsay. When the court denied this motion and ruled the reports admissible, plaintiff had the responsibility of requesting redaction of specific statements. Plaintiff never requested redaction, nor pointed out which statements were objectionable. Thus, we conclude, plaintiff forfeited his right to object to admission of the statements within the reports. 3

Further, we agree with the District Court that any error in the reports’ admission was harmless. Although the documents contained prejudicial statements, those statements were testified to directly by the witnesses and, with the exception of a few statements in the Morgan report, explored on cross-examination. Thus, the risk of prejudice was only the cumulative effect of having the jury see the statements in writing after already hearing them in court, not that the jury was exposed to statements of which they otherwise would not have been aware. Also, although plaintiff could not rebut the reports specifically in his closing argument, he was able to rebut the content of the reports through his arguments by rebutting the testimony to the same tenor. There is nothing in the reports that the jury had not already heard or which the plaintiff had not had the opportunity, on cross-examination and closing argument, to discredit. We find no reversible error in the admission of Exhibits 48 and 49.

IV.

Plaintiffs third argument is that the District Court’s admission of defense expert testimony regarding post-incident tests was erroneous because the circumstances of the tests were not substantially similar to those of the incident. The first test which plaintiff challenges was done in September, 1991, by Morgan, defendant’s chief accident investigator and expert witness. Morgan initially sought to determine the amount of water that had been in the aircraft’s belly during the incident. As discussed above, when Co-mair mechanics inspected the aircraft on the ground following the incident, one inch of water was observed in the fuselage. Morgan determined an average deek angle for the aircraft by measuring the deck angles of nine aircraft and averaging them reaching a figure of one degree nose-up. He then took an old Metro II' fuselage, assertedly with the same dimensions as the Metro III, hitched it to one degree nose-up, and filled it until one inch of water was in the fuselage at the place observed on" plaintiffs airplane.

Morgan then performed flight tests on the aircraft. In the test flights, the plane was loaded such that the center of gravity was at the forward most limit for normal flight. This was done by placing 400 pounds of luggage in the forward baggage compartment and placing three passengers in the first three rows of the cabin. They conducts ed various test flights to determine a possible cause of the yoke pressure. Morgan testified that only the misapplication of the pitch trim system by the pilots, coupled with water-shifting in the belly, could have caused the nose-down force experienced by the pilots during the incident.

Plaintiff objects to the testimony concerning these tests because, he argues, the test conditions differed in significant respects from the conditions of the incident. First, plaintiff challenges Morgan’s estimate of the amount of water that existed in the plane during the incident. He argues that Morgan’s conclusion was unfounded because the deck angle of the plane during the incident was never determined and because Morgan failed to consider that the debris which clogged the fuselage drains during the inci *732 dent might have prevented the water from shifting forward. Plaintiff challenges the flight tests on the ground that the test plane did not have water in its belly; Morgan did not induce an SAS malfunction during the tests; and because Morgan loaded the plane such that the center of gravity was more toward the front of the plane than was the case in the incident. He argues that the tests were irrelevant and prejudicial, and therefore the District Court should have excluded the testimony about them.

The question is whether the post-incident tests performed by Morgan and other defense experts were probative of what occurred during the incident and, if so, whether the risk of unfair prejudice substantially outweighed their probative value. Defendant concedes that the tests did not duplicate the incident. They did not pretend to. The asserted purpose of the tests was to determine what factors could have caused the extreme forces in the yoke allegedly experienced by plaintiff during the incident. Defendant’s theory, supported by credible evidence, was that the SAS could not exert the force experienced, especially after it was switched off. The flight tests were designed to create conditions that might result in such force. Establishing an alternative explanation is certainly probative of what happened. The tests were initially designed to maximize front-load pressure to determine whether, under the most likely circumstances, extreme yoke force could be induced. The fact is, however, that even under these conditions, the test plane did not experience the forces reported by plaintiff. With a negative result, the test was probative because it established that the plane’s center of gravity, i.e. the forward shift of water in the belly, could not have been the sole cause of the yoke force. Besides the effect of weight load on the yoke pressure, the flight tests also included experiments with the trim application. This was to determine whether the water load in combination with trim application could have caused the incident. Although the test plane did experience increased yoke pressure, it did not approach the force experienced by plaintiff. This supports at least two conclusions: that the SAS was also a factor, as plaintiff alleged, or that the pilots did not experience the force complained of. In either event, the test results were probative. Furthermore, although some tests were done with the passenger weight further forward than in plaintiff’s aircraft, some tests were done with the passenger weight toward the rear of the plane. These tests thus lacked the excess front loading complained of by plaintiff. In short, while not duplicating the incident, Morgan’s tests were probative of the incident. Moreover, the risk of any unfair prejudice appears low considering plaintiff’s extensive cross-examination into the shortcomings of the tests. We conclude, therefore, that the trial court’s admission of Morgan’s testimony was not an abuse of discretion.

Finally, plaintiff argues that the District Court erred in refusing to admit three exhibits (501, 509, 510) prepared by governmental agencies concerning the Fairchild Stall Avoidance System. The court excluded the reports on the ground that the incidents reported were not substantially similar to plaintiffs incident.

We find no abuse of discretion in refusing to admit Exhibit 501. As the trial court noted, none of the incidents referred to in the document involved an SAS exposure to water. The trial court also reasoned that, because the document is authored by a governmental agency, the jury would be prejudiced by its appearance of authority notwithstanding its lack of probative value. Finally, we note that the accidents referred to in the report not only involved different causes of the SAS activation, but the SAS was always deactivated by the pilot. In this case, there is no dispute that the SAS unintentionally activated, the dispute concerns the cause of the continued force after the SAS switch was turned off. Exhibit 501 does not bear on this issue.

Exhibit 509 is an NTSB accident report concerning a fatal crash of a Fairchild Metro III aircraft in North Carolina in 1988. Although the court excluded this document, it did allow plaintiff to use the document to cross-examine Heaslip about the Metro Ill’s accident history. Plaintiff wanted to introduce this exhibit primarily because it con *733 tains a reference, apparently to the incident in this action, of an unwarranted SAS malfunction in a Metro III aircraft during an approach to the Greater Cincinnati International Airport. In addition, the report referred to a perceived malfunction of the SAS in the North Carolina accident. Plaintiff argued for admission of the document under the public record exception to the hearsay rule as an investigation conducted pursuant to law. See Fed.R.Evid. 803(8)(C).

We conclude that the court did not abuse its discretion in excluding Exhibit 509. The North Carolina accident, which was the focus of the report, was determined to have a different cause than that alleged here. Although there was a perceived malfunction of the SAS, the probable cause was determined to be the fault of the flight crew, captain, and the airline. There was also no indication of water exposure. Furthermore, assuming the investigation of the North Carolina incident was admissible under Rule 803, the reference to the instant incident was not. That reference was not a finding of the report but rather a reference to a finding in another report. Thus, the statement was hearsay within the document. Iii addition, the NTSB did not investigate this incident and therefore the origin of the referenced report is unclear.

Exhibit 510 is an accident report issued by the French equivalent of the NTSB. The report concerned a fatal air crash in November of 1988. The report does mention a malfunction of the SAS. However, the report goes on to state that any proof of what happened was destroyed in the accident. In addition, the accident occurred on takeoff not landing, and there was no mention of water accumulation or a failure of the SAS to deactivate when turned off. These differences, we conclude, leave the trial court’s exclusion of Exhibit 510 within its sound discretion.

YI.

For the foregoing reasons, the judgment in favor of defendant is AFFIRMED.

. Besides objecting to the admission of the experts' reports, plaintiff also challenges Morgan's testimony relating to his investigation. The latter issue is addressed in part IV, infra.

. We do not intend to suggest that a party's right to admit prior statements in a document for rehabilitative purposes necessarily entitles that party to submit the document as an exhibit to the jury. Permitting the document to go to the jury room could unduly prejudice the cross-examining party. Instead, the trial court may consider admitting the statements in the document through testimony or by having the document read to the jury. If the court determines that admitting the document as an exhibit is necessary, the court should instruct the jury as to the limited purpose of the exhibit. We leave that determination to the sound discretion of the trial court.

. We note, however, our disagreement with the trial court’s procedure in admitting the documents. The court delayed its ruling until after completion of closing arguments and jury instruction's. A ruling at the time objection is made, or at least before the close of evidence, better enables the parties to proceed with their case accordingly. We conclude, however, that plaintiff still had the opportunity to request redaction or a limiting jury instruction or even to reopen arguments if they were necessary.