David Bryte, Personal Representative of the Estate of Lova E. Bryte, Deceased; David Bryte; Kathy B. Smith; James B. Smith; M.e. Smith, a Minor, by and Through Her Legal Guardian, Kathy B. Smith; Donna J. Miller, Plaintiffs-appellants, v. American Household, Incorporated, Formerly Known As Sunbeam Corporation; Sunbeam Products, Incorporated, Defendants-appellees, Andsears Roebuck and Company, Defendant, 429 F.3d 469 (4th Cir. 2005)

On October 23, 2000, Lova Bryte died in a fire in her Preston County, West Virginia apartment. She was using an electrically heated throw at the time of the fire. Several weeks later, plaintiffs, Mrs. Bryte's personal representative and relatives, brought this action against defendant American Household, Inc. (formerly Sunbeam Corp.), the manufacturer of the throw. Plaintiffs alleged that the electric throw had a defective safety circuit and that this defect caused the deadly fire. At trial, the district court excluded various evidence, including expert testimony, offered by plaintiffs to prove the causation and defect elements of their claims. At the close of plaintiffs' case, the district court granted defendant's motion for judgment as a matter of law on the ground that plaintiffs had not established sufficient evidence of causation. Plaintiffs appeal from the district court's evidentiary rulings and the ensuing judgment for defendant. We affirm.

Though the record on appeal is extensive, the basic facts of this case are not disputed. Mrs. Bryte had begun using the electric throw in question about two years before the fire occurred, after having suffered a stroke. As a result of her stroke, Mrs. Bryte could not pick herself up from her recliner without assistance, and her movement was limited to shifting in the chair. Also as a result of her stroke, Mrs. Bryte used the electric throw throughout the year to keep warm when sitting in her recliner. Typically, she would have had placed over her a blanket or afghan and the throw would be laid on top of the blanket. When in use, the electric throw would be plugged directly in to a wall outlet on Mrs. Bryte's left. When not in use, the throw was draped over the back of the recliner or a love-seat. The throw had not malfunctioned and there was no visible damage to it or its power cord.

Other than the recliner, the living room of Mrs. Bryte's apartment contained a love-seat, a television and VCR, a coffee table, a telephone, and a small table to the right of the recliner, which table had a lamp and a candle on it. Mrs. Bryte did not smoke.

The morning of the fire, Mrs. Bryte and her care-giver, Donna Miller, were in the apartment alone. During the late morning or early afternoon, Mrs. Miller assisted Mrs. Bryte into the recliner and then retrieved the electric throw from the back of the love-seat. Mrs. Miller covered Mrs. Bryte with a blanket, then with the throw, plugged the throw in to a wall outlet to the left of the recliner, and turned it on to the low setting. The throw was not tucked in around Mrs. Bryte's sides. The power cord extended, at Mrs. Bryte's feet, from the throw, and the temperature control rested on the floor to Mrs. Bryte's left below her feet. Mrs. Bryte did not adjust the temperature control herself. At around 11 a.m., Mrs. Miller had lighted the candle that sat on the small table to the right of the recliner. According to Mrs. Miller, at that time the candle was partly burned, and, as far as she saw, the flame had not extended above the rim of the glass candle container. Mrs. Miller also testified that Mrs. Bryte was unable to reach the candle, or the lamp that also was on the table. Mrs. Bryte could, however, reach a part of the table for such things as a denture cup, a Star magazine, a cookbook, a Bible and her glasses.

Mrs. Miller left the apartment at 1:55 p.m. to pick up Mrs. Bryte's granddaughter from school, leaving the candle lighted and the electric throw turned on. Mrs. Miller returned to the apartment at, she estimated, 2:08 p.m., at which time she heard Mrs. Bryte calling for help. Upon entering the apartment, Mrs. Miller saw smoke and fire coming up the left side of Mrs. Bryte's recliner and heard "snapping and cracking." Mrs. Miller did not see any flames related to the power supply cord of the electric throw and she could not identify more specifically the source of the flames, other than the fact that the flame was "down beside [Mrs. Bryte's] chair and up on her leg." Mrs. Miller tried, but was unable, to extinguish the fire. While in the apartment, however, Mrs. Miller was able to see that the candle was still lighted and that neither it nor the lamp had been moved or overturned. The flames and smoke forced Mrs. Miller to flee the apartment to get the children out.

Shortly after the fire was extinguished, and before the removal of Mrs. Bryte's remains, Assistant State Fire Marshall Mack Dennis was called to the scene. At that time, Dennis had been employed by the West Virginia State Fire Marshall's Office as a cause and origin investigator for over 20 years. While he was not certified as a fire investigator, he had attended numerous courses in arson and explosive investigation. Dennis' investigation of the fire scene traced the fire's path, starting from Mrs. Bryte's left side and moving around the room. Dennis took photographs of the fire scene and made a not-to-scale diagram of the room where the fire occurred, which located the large artifacts in the room. Dennis took oral, not written, statements from the witnesses to the fire.

Dennis' diagram did not indicate the wall outlets located in the room. He did observe, however, that the wall outlet immediately to Mrs. Bryte's left "had an electrical service cord plugged into it, and the remains of that cord came over to Mrs. Bryte and was [sic] laying across her arm." Dennis' report later characterized the cord as "wrapped around" Mrs. Bryte's left arm. He did not measure the cord or the exact distance between the wall outlet to her left and Mrs. Bryte's body, but approximated the distance to be "two and a half feet." Dennis did not recover or photograph the service cord that he saw draped across Mrs. Bryte's body. Indeed, Dennis did not know at first what this cord was for; only later, after speaking with family members and Mrs. Miller, did he learn that Mrs. Bryte had been using an electric throw at the time of the fire.

Dennis did not know that there was a lamp, or observe any other cord, plugged in to the same wall outlet as the cord found on Mrs. Bryte. But there was another cord, an extension cord running from a wall outlet behind the chair, to the lamp on the table. Moreover, he did not inspect the outlet's wiring. Dennis nonetheless excluded the outlet, and also the table, candle, and lamp as potential origins of the fire based on his discussions with Mrs. Miller and family members. Also, he made this determination without knowing how the candle had been lighted.

Dennis' only report, issued the day of the fire, concluded that the cause of the fire was "improper use of an electric blanket." Specifically, Dennis recorded Mrs. Miller as having told him that the electric throw was "around" Mrs. Bryte. Dennis later may have concluded, for reasons that are unclear due to interruptions in his testimony, that the throw may not have been used improperly. But he nevertheless maintained that the throw was the cause of the fire.

Critically, Dennis did not recover any artifacts or samples from the scene. Dennis testified that he had no reason to do so, nor did he advise the Brytes to preserve the fire scene or any of its artifacts, for the same reason: "I had no reason to believe it [the fire] was other than accidental." Apparently no one, until his attorney did, instructed David, Mrs. Bryte's son, about preserving evidence from the fire scene, and David removed the debris, including the recliner and electrical devices, and disposed of it at a landfill that he owned.¹ 

Plaintiffs filed this suit about six weeks later, on December 1, 2000.²  In addition to Dennis, plaintiffs retained Dr. W.T. Cronenwett as an expert witness to offer opinion testimony regarding the specific cause of the fire. In his first deposition, Dr. Cronenwett opined that the electric throw was the ignition source of the fire. He based that opinion on " [t]he absence of any other ignition source in the area, the snapping and crackling that was heard, the fact that Mrs. Bryte was confined to her chair, did not smoke and had no opportunity or means to set herself on fire, and the fact that there are—have been a number of instances where the Sunbeam electrically heated bedding products have spontaneously burst into flames." More specifically, Dr. Cronenwett testified that the blanket caused the fire due to defectively designed circuitry. He acknowledged, however, that his opinion that the defective circuitry in the throw had caused the fire rested wholly on the conclusion of Dennis.³  Dr. Cronenwett attributed his inability to independently identify the specific source of ignition to the fact that there was "no remaining physical evidence" after the fire.

Dr. Cronenwett was deposed twice, the second time in April 2003. In addition to repeating his earlier questions, counsel for defendant repeatedly asked Cronenwett whether he could identify evidence of a malfunction in the electric throw that could reliably support his conclusion that the electric throw caused the fire. To each such question with respect to various specific defects, Dr. Cronenwett acknowledged that he could not, explaining, for example, that "... I never saw any physical evidence of the blanket, so I don't know what specific—what the specific cause of the ignition was." Moreover, when asked whether there was "any evidence in this case that can reliably support a conclusion that there was any defect in any component part of the electric throw that caused the fire in this case?" Dr. Cronenwett responded "No." (JA 2019.)

Prior to trial, defendant moved for summary judgment based on failure of causation and plaintiffs' spoliation of the physical evidence of the fire. Defendant also moved to exclude the testimony of Dennis and Dr. Cronenwett (and Messrs. Dallas and Hull) under Fed.R.Evid. 702 and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469 (1993), on the related ground that neither had a reliable basis for his opinion that the electric throw or a defect therein had caused the fire. In his Report and Recommendation, the magistrate judge recommended denying defendant's motions, except as to Dallas, whose opinion he deemed unreliable. Defendant objected to the Report and Recommendation; the district court overruled those objections but treated the motions as pending, took them under advisement, and instructed the parties that it would rule on them at trial.

The case was tried on December 8, 2003. At trial, plaintiffs attempted to offer the expert testimony of Dennis and Dr. Cronenwett in support of the claims that (1) the electric throw was the cause of the fire, and (2) a design defect in the blanket was specifically responsible for ignition. Contrary to the Report and Recommendation, the district court considered plaintiffs' proffers and determined that, under Daubert, neither expert had a sufficiently reliable basis for his opinion, and it thus excluded Dennis' and Dr. Cronenwett's testimony as to causation. The district court also excluded plaintiffs' evidence about other, similar accidents and reports of defects. The district court then concluded that plaintiffs had not introduced evidence sufficient to permit a jury finding that the electric throw had caused the fire. Therefore, at the conclusion of plaintiffs' case the district court granted defendant's Rule 50(a) (1) motion for judgment as a matter of law. This appeal followed.

Plaintiffs assert five assignments of error:

I. THE TRIAL COURT FAILED TO PROPERLY APPLY THE "MALFUNCTION THEORY" OF CIRCUMSTANTIAL EVIDENCE IN PRODUCTS LIABILITY CASES WHEN THE PRODUCT IS DESTROYED.

II. THE TRIAL COURT MISAPPLIED THE FLEXIBLE DAUBERT ANALYSIS OF RULE 702 TO EXCLUDE EXPERT TESTIMONY FROM MACK DENNIS AND DR. CRONENWETT BECAUSE THERE WAS NO TESTING.

III. UNDER THE "MALFUNCTION THEORY" OF CIRCUMSTANTIAL EVIDENCE, OTHER RELEVANT EVIDENCE OF PRODUCT DEFECT AND SAFETY SHOULD HAVE BEEN ADMITTED.

IV. EITHER WITH OR WITHOUT THE IMPROPERLY EXCLUDED EVIDENCE, THERE WAS SUFFICIENT CIRCUMSTANTIAL EVIDENCE TO SUBMIT THE CASE TO THE JURY.

V. THE TRIAL COURT'S CONDUCT DURING TRIAL DEMONSTRATED BIAS, PREJUDGING OF THE CASE, AND OVERALL MEDDLING IN THE PRESENTATION OF EVIDENCE SO AS TO DEPRIVE PLAINTIFFS OF DUE PROCESS AND A FAIR TRIAL BEFORE AN IMPARTIAL TRIBUNAL. (Appellants' Br. i, ii.)

Appellate review of a district court's interpretation or application of state law is de novo. See James v. Circuit City Stores, Inc., 370 F.3d 417, 421-22 (4th Cir. 2004). Conversely, the district court has broad latitude in ruling on the admissibility of evidence, including expert opinion, and we will not overturn Daubert evidentiary rulings with respect to relevance and reliability absent an abuse of discretion. See Gen. Elec. Co. v. Joiner, 522 U.S. 136, 141-42, 118 S. Ct. 512, 139 L. Ed. 2d 508 (1997). A district court abuses its discretion if its conclusion is guided by erroneous legal principles, or if it rests upon a clearly erroneous factual finding. See Westberry v. Gislaved Gummi, A.B., 178 F.3d 257, 261 (4th Cir. 1999). Finally, we review de novo the district court's grant of judgment as a matter of law, see Freeman v. Case Corp., 118 F.3d 1011, 1014 (4th Cir. 1997), cert. denied, 522 U.S. 1069, 118 S. Ct. 739, 139 L. Ed. 2d 676 (1998), considering the evidence in the light most favorable to the non-moving party, see Brown v. CSX Transp. Inc., 18 F.3d 245, 248 (4th Cir. 1994).

Taken together, we must determine whether the admissible evidence, viewed most favorably to plaintiffs, would have allowed the jury to render a verdict in their favor under West Virginia law.

We first address assignments of error I and II which state plaintiffs' principal claim on appeal, which is that the district court erred in applying Daubert rather than substantive West Virginia law as to causation. In essence, plaintiffs rely on several West Virginia cases and argue that, when a product allegedly causing a fire has been destroyed by that fire, plaintiffs are allowed to rely on the "Malfunction Theory" and circumstantial evidence to prove that a defect in that product caused the fire. The argument goes that evidence as to causation is not controlled by Daubert and the Federal Rules of Evidence, rather by the substantive law of West Virginia which is claimed to embrace the "Malfunction Theory" of proof of causation.

Plaintiffs characterize the district court's decisions on evidence as reversible disregard for West Virginia law. The following colloquy at trial succinctly illustrates plaintiffs' argument and the district court's explanation for rejecting it:

Plaintiffs' attorney: " [T]he law in the Fourth Circuit and in West Virginia is that an expert need not review product remains to arrive at an opinion as to the probable defect in it."

Court: "The evidence [law] is the Federal Rules of Evidence."

As the transcript shows, plaintiffs essentially advocate applying West Virginia law rather than the Federal Rules of Evidence where, as here, products-liability claims, destruction of physical evidence, causation, and expert testimony converge.

The position of the district court was correct and has been decided in Cavallo v. Star Enterprise, 100 F.3d 1150 (4th Cir. 1996). In that case the district court, as here, entered a judgment for the defendant, having excluded expert testimony as to the cause of the various chronic illnesses of the plaintiff. The expert witnesses had offered testimony that the plaintiff's exposure to petroleum vapors from jet fuel spilled or leaked by the defendant had caused the plaintiff's illnesses, upon which she sued. The court excluded the evidence because it was not supported by appropriate validation, as required by Daubert v. Merrell Dow, 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469 (1993). The position of the Cavallo plaintiffs in Virginia, similar to that taken here by the Bryte plaintiffs in West Virginia, was that Virginia law should apply as to the admissibility of the testimony:

The Cavallos assert that, because their claims are grounded in Virginia state law and were originally filed in Virginia state court, the admissibility of expert testimony should be governed by Virginia law.

However, we followed our decision in Scott v. Sears, Roebuck & Co., 789 F.2d 1052 (4th Cir. 1986) and affirmed the district court, quoting from Scott, 789 F.2d at 1054:

Unlike evidentiary rules concerning burdens of proof or presumptions, the admissibility of expert testimony in federal court sitting in the diversity jurisdiction is controlled by federal law. State law, whatever it may be, is irrelevant.

100 F.3d 1150, 1157.

This argument has been rejected by other Courts of Appeals in similar cases dependent on expert testimony. For example, in Ealy v. Richardson-Merrell, Inc., 897 F.2d 1159 (D.C. Cir. 1990), the defendants appealed the district court's denial of their motion for judgment N.O.V. in a Bendectin birth-defects case. Plaintiffs defended the verdict in their favor by arguing that Erie Railway Co. v. Tompkins, 304 U.S. 64, 78, 58 S. Ct. 817, 82 L. Ed. 1188 (1938), required the district court to follow an earlier District of Columbia Court of Appeals decision that had affirmed a jury verdict for different plaintiffs based on the "same four types of expert data" relied upon by the federal plaintiffs. Ealy, 897 F.2d at 1163. The D.C. Circuit agreed with defendants that reliance on state (decisional) law was not required. Ealy, 897 F.2d at 1160 (" [t]he admissibility of expert testimony [is] a procedural matter which does not implicate the principles underlying Erie") (citation omitted). The court explained its holding, that the district court abused its discretion in admitting unreliable expert testimony accepted by the state court, in terms applicable to this case:

Unfortunately for [plaintiffs], the only evidence they have on the issue of scientific causation is the testimony of their experts. Once that testimony is deemed inadmissible, it is self-evident that the evidence is insufficient to support a verdict for the plaintiffs. Contrary to the [plaintiffs'] assertions, however, this court is enforcing a procedural rule governing admissibility; it is not supplanting substantive state rules on the sufficiency of evidence.

Ealy, 897 F.2d at 1163; cf. Daubert v. Merrell Dow Pharms., Inc., 43 F.3d 1311, 1322 (9th Cir. 1995) (affirming summary judgment for defendants: " [p]laintiffs do not quantify this possibility [of causation], or otherwise indicate how their conclusions should be weighted, even though the substantive [state-law] legal standard has always required proof of causation by a preponderance of the evidence"), on remand from 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469 (1993). These decisions highlight the application in the federal courts of the process that plaintiffs argue is a "misapplication" of West Virginia law, rather we are of opinion that it is an application of the rules of evidence under Daubert.

Without significant exception, plaintiffs relied on Dennis' testimony to establish that the throw caused the fire.⁴  But Dennis did not exclude all or even most of the other possible sources of the fire. As set forth above, it is undisputed that Dennis did not physically examine the lamp, the candle, the cord that remained which he found on Mrs. Bryte's arm, or the wall outlet, or its wiring, which supplied electricity to the throw. Moreover, he was unaware that an extension cord was plugged into the same outlet as was the blanket, and therefore he obviously did not inspect it. These omissions are particularly revealing in light of Mrs. Miller's testimony that she could not unplug the throw because she "would have gotten burned if [she] tried to pull [the cord] out," which suggests fire on the left side of the chair near the outlet, even if not in or on the outlet itself. It is clear that such possibilities have not been excluded in any methodical or reliable fashion. And we are not obliged to credit Dennis' say-so supporting his own reliability by way of excluding other causes. See Kumho Tire Co. v. Carmichael, 526 U.S. 137 at 157, 119 S. Ct. 1167, 143 L. Ed. 2d 238 (1999).

Plaintiffs also assert that Dennis adequately excluded the most likely alternative source of the fire, the candle, based on Mrs. Miller's observation that the candle was still lighted when she arrived at the scene, as well as on the evidence that Mrs. Bryte could not reach the candle. We do not think that Dennis' explanation is sufficient to explain away the candle. Our doubt arises not because of Dennis' reliance on Mrs. Miller's observations, but because the fact that she saw the lighted candle on the table in itself cannot exclude it as a cause of the fire, rather the contrary inference is the more logical. This is a matter of common sense: for example, the candle's flame might have burned the lamp shade above, as defendant suggested when cross-examining Mrs. Miller, or either the blanket or the throw may have been ignited upon contact with the candle which may have been within reach of the decedent. Notably, though Mrs. Miller and Mrs. Bryte's daughter testified that the candle was out of Mrs. Bryte's reach, Dennis did not independently investigate why what appeared to be the table's remains were found close to the recliner, and his only explanation for this was the implausible one that a fireman had kicked the debris. The table with the candle on it was described by Dennis as "immediately to your left of" the recliner, and a photo in evidence shows the table as adjacent to the recliner.

Daubert aims to prevent expert speculation, and our review of the record convinces us that Dennis' failure to independently evaluate the open flame in the room cannot be reconciled with the reliability mandate. Dennis was permitted to rely on what Mrs. Miller saw, but not on her conclusions about the cause of the fire. As to the candle he essentially did the latter.

We also note that, at the beginning of the trial, the justifiable impression of the district court, the depositions of most of the witnesses having been taken, when Mrs. Miller left to get the child, "there was a candle lighted on a table within reach of the decedent."

Plaintiffs also argue that Dennis performed a methodical walk-around inspection, which led him to the origin of the fire. Dennis' investigation did indeed determine that the left side of Mrs. Bryte was the "area of origin" for the fire—but he could not identify the "point of origin." And, as seen above, Mrs. Miller testified as well that the fire was on the outside of the chair, and on Mrs. Bryte's left side as well. Thus, though both Dennis' "area of origin" finding and Mrs. Miller's testimony implicated the area where the outlet and the two power cords were located, Dennis did not closely examine this area and was unaware of a cord lying inches from the recliner, plugged in to the same outlet as the throw. This is inconsistent with the NFPA standards, which require investigators to exclude " [a]ll other reasonable origins and causes." NFPA 921, § 2-3.6 (1999 ed.).

Furthermore, Dennis' causation testimony is called into question by the fact that Dennis may have sought to explain his conclusion about "improper use" of the throw, based upon the fact he "had information that ... [the throw] had been tucked down in beside the victim in the chair and which I later learned the old style electric blankets ..."

While the testimony of Dennis was cut short before he had explained his use of the phrase "improper use" with respect to the throw, the proffered testimony of Dr. Cronenwett with reference to causation is instructive. The "snapping and crackling noises" referred to throughout and referred to by Mrs. Miller are "characteristic of burning ... energized PTC cable" not the burning of an energized power or service cord, for, "generally, when a service cord short-circuits it explodes once and opens a circuit breaker. It would not continue to snap and crackle." Thus, "the burning of an energized service cord for an appliance would be inconsistent with a snapping and crackling sound described by Ms. Miller." Dr. Cronenwett had no information that the circuit breaker was tripped in the line into which the throw had been plugged. But because snapping and crackling came from the burning PTC cable⁵ , he logically excluded a burning of the power cord to the throw as the cause of the fire. Apparently unknown to Dr. Cronenwett, however, the circuit breaker, in the line into which the throw was plugged, was tripped. When it was tripped is not shown in the record. That it was tripped is shown by the testimony of David Bryte, the son of the decedent: " [T]he breaker on that particular circuit was tripped." With the power cord possibly removed as the origin of the fire, and the burning of the PTC cable as the origin of the snapping and cracking, it then is relevant to consider the testimony of Dr. Cronenwett that "the PTC wire could also burn from being attacked by an outside fire source as well." But does, or doesn't, the tripped circuit breaker make the various cords (or their appliances) plugged into the circuit, candidates for the origin of the fire.

All of these reasons serve to support the decision of the district court in excluding the causation testimony of Dennis, and we are of opinion the district court did not abuse its discretion in excluding that testimony. It described the record at this point as "so speculative that a jury has to guess to determine whether there is liability of either the warranty or negligence for the jury to speculate on."

In the third assignment of error, plaintiffs challenge the district court's exclusion of other evidence such as "other accidents including the same product." Here, too, we find no abuse of discretion. See Westberry, 178 F.3d at 261.

Plaintiffs rely largely on this court's decision in Riley v. De'Longhi Corp., 238 F.3d 414 (4th Cir. Oct.30, 2000), in support of their argument regarding other relevant evidence.⁶  In Riley, we determined that four of the five factors applicable to analyzing the sufficiency of evidence of a product defect under Maryland law favored reversing summary judgment for the defendants.

Under Riley, the rule in Maryland in deciding whether there is sufficient circumstantial evidence of a product defect to proceed to trial, is that the court considers five factors: (1) expert testimony as to possible causes; (2) the occurrence of the accident a short time after sale; (3) same accidents in similar products; (4) the elimination of other causes of the accident; and (5) the type of accident which does not happen without a defect.

As to each factor:

(1) The testimony of Dennis, that the fire originated in the throw, was inadmissible, so there was no expert testimony as to the possible cause.

(2) The fire in this case occurred some two years after the sale, so there might be a question of fact as to whether the accident occurred a short time after sale.

(3) As the district court correctly held, there was no proof that the facts in this case were the same or similar to what transpired in the other cases, so there was no showing of the same accident in similar products.

(4) In this case the elimination of other causes of the accident has not been shown, as we have related above, so there has been no showing of the elimination of other causes of the accident.

(5) Since there is no showing of the cause of the accident, there has been no showing of a type of accident which does not happen without a defect. Thus, even under Riley, the evidence of "other accidents involving the same product" is not admissible, only one factor possibly being favorable.

A general logical fallacy similar to that mentioned just above undermines plaintiffs' theory that the district court was required to admit evidence of substantially similar incidents and reports implicating a Sunbeam throw. That is, plaintiffs sought to introduce such evidence without establishing their premise: the salient characteristics of this incident, as well as of the incident to which comparison is sought. The district court commented:

And I agree with [plaintiffs' counsel] one hundred percent [that it is proper to bring evidence of other similar incidents]; and once you get a witness who testifies that this is similar to what went on in these other cases, I might let it come in....

We are of opinion that the district court did not abuse its discretion in denial of evidence of "other accidents involving the same product."

The fourth issue is plaintiffs' argument that "with or without the improperly excluded evidence, there was sufficient circumstantial evidence to submit the case to the jury." We review de novo the district court's grant of judgment as a matter of law. See Freeman v. Case Corp., 118 F.3d at 1011, 1014 (4th Cir. 1997). Though we consider the evidence in the light most favorable to the non-moving party, Freeman, 118 F.3d at 1014, given our affirmance of the district court's decisions as to the admissibility of evidence, the required result is judgment for the defendant. With no witness as to ignition, there simply was no evidence of causation other than Dennis,' and via that, Dr. Cronenwett's, testimony. As in Ealy, "the only evidence [plaintiffs] have on the issue of scientific causation is the testimony of their experts. Once that testimony is deemed inadmissible, it is self-evident that the evidence is insufficient to support a verdict for the plaintiffs." 897 F.2d at 1163 (emphasis added). We are of that same opinion here and so hold.

The judgment of the district court is accordingly

AFFIRMED. ⁷