Complex Litigation

Admissibility of Test Evidence

By Michael Hoenig - New York Law Journal - July 12, 2010

Our trigger for discussing admissibility of tests and demonstrations is a decision on June 16 by an intermediate appellate court in Florida, Mitsubishi Motors Corp. v. LaLiberte,1 reversing a multimillion dollar judgment in a products liability case. The front seat passenger of an automobile that rolled over multiple times at high speed, though seat-belted, was partially ejected through the rear passenger window and subsequently died. The significance of the decision is how the court distinguished between two evidentiary scenarios lawyers and even some judges all too frequently conflate or confuse.

Application of the venerable evidentiary doctrine of "substantial similarity," i.e., that demonstrations or tests proffered into evidence should be substantially similar to the facts of the accident, hangs in the balance. There is a marked distinction between a litigant's offer of a test, experiment or demonstration aiming to recreate, reproduce or approximate the actual accident circumstances as opposed to a proffer of demonstrative test evidence to illustrate, demonstrate or prove visually a fact that is relevant and probative on the controversy. The latter scenario, usually accompanying, illustrating or confirming some aspect of expert testimony, is not governed by the "substantial similarity" rule. Yet, many lawyers and some judges blur the distinctions. The new decision is informative on the point. More about the case later. First, we elaborate a few cardinal points about test and demonstrative evidence generally.

Cases involving complex technical issues can be won or lost on the battle waged over the admissibility of a particular test or experiment. This is particularly true in products cases where the performance of a particular design or part in the accident is the subject of vigorous dispute by adverse experts. Sometimes the key evidentiary battle may not be admissibility of a particular test and its results but over how that information is presented to the jury, for example, via a dramatic film, videotape or computer simulation.

Demonstrative evidence generally is the expert's "support" medium by which the jury can see and "feel" the involved structures instead of only picturing them via word images or crude illustrations on a sketch pad. If the time, labor and expense in performing and recording tests or experiments by qualified experts is worth the gambit, it makes little sense for proponents (or opponents) of such evidence not to prepare properly for the battle over admissibility.

One must note the powerful role of trial court discretion. Trial judges have a first-hand view of the proceedings, the witnesses, the test videotapes themselves and their quality and informational value, as well as a front-row seat as to possible prejudicial impact. Appellate courts tend to defer greatly to such opportunities of observation.

The proponent or opponent of demonstrative test evidence has to ask a number of basic questions in checklisting the potential arguments pro and con. Is the test evidence "relevant"? Remember, relevant evidence is admissible unless there is some valid reason or rule that should keep it out. Is the objective of the offered experiment to show results under conditions that duplicate or substantially approximate the actual conditions in issue at trial? Or is the test evidence merely intended to illustrate a scientific principle or some technical aspect of the expert's testimony that might assist the jury?

If the demonstrative materials would help the jury, do their prejudicial aspects outweigh the advantages of letting them in? Can the reasons for admissibility be explained and potential prejudice minimized by a so-called "limiting" or "curative" instruction by the judge to the jury defining the limited purposes of such evidence? Other pertinent questions will flow from the foregoing.

'Substantial Similarity'

The objective of the "substantial similarity" requirement is to prevent misleading the jury, which may attach exaggerated significance to a "scientific" test or demonstration. The legal standard is not a test's duplication of actual conditions. Instead, the relevance of experimental evidence depends on whether or not the experiment was performed under conditions substantially similar to those of the actual occurrence sought to be proved.

The "substantial similarity" criterion is fact-dependent. Obviously, a scintilla of "similarity" is not enough. Neither is exact replication of conditions required. Between these extremes lies a gray area ripe for advocacy and trial court discretion. "Fairness" concerns and "relevance" criteria play a significant role. The burden of demonstrating substantial similarity of test conditions to accident conditions is on the proponent of the demonstrative evidence.

Sometimes, however, the test or experiment is not intended to duplicate, reenact or even approximate "substantially similar" conditions governing an occurrence but is offered only to illustrate or demonstrate a scientific principle. Alternatively, it may illustrate only one aspect of the expert's reasoning process leading to an opinion. In such cases imposing the "substantial similarity" requirement would wreak havoc with the objective of assisting the jury to understand technical matters. Thus, courts have fashioned a different test of admissibility as well as a curative element that sometimes must accompany the evidence—a special instruction to the jury that puts the test evidence in appropriate context and explains the limited purpose for receiving it.

In considering test evidence, courts also analyze the probative value-versus-prejudice calculus reflected in Federal Evidence Rule 403 and equivalent state rules of evidence. Thus, even if the proponent of test evidence establishes substantial similarity, for example, the test will not be admitted if "its probative value is substantially outweighed by the danger of unfair prejudice" or the danger of "misleading the jury." Decisions declaring and illustrating the foregoing principles are cited in the corresponding endnote.2

Despite the evolution of these historical evidentiary approaches to admissibility of test evidence, there is a "new kid on the block," so to speak, that materially affects the admissibility of test evidence. This is the judicial "gatekeeping" task thrust upon courts in the Daubert/Kumho Tire era in which courts are charged with the obligation to police expert testimony for reliability as well as for relevance.3 Daubert and Kumho posited that scientific, technical and specialized evidence must be reliable and elaborated a number of nonexclusive reliability criteria helpful to gatekeeping judges. The Daubert reliability standard and admissibility factors have been incorporated into Federal Rule of Evidence 702 and the Notes of the Advisory Committee.

State courts adhering to the Frye evidentiary standard applicable to novel scientific evidence perform the same or similar screening function. An example is Nassau County Supreme Court Justice Thomas P. Phelan's decision in Santos v. State Farm Fire & Casualty Co. rejecting defendant's proffer of expert opinions based on a computer fire dynamics simulation model used to plot how a residential fire would spread and to thus conclude that the fire had been deliberately set. The opinion and an article about the decision is found in the Law Journal of July 6.4

'Reliability' Factor

One of Daubert's principal reliability factors is "testability" which, in the U.S. Supreme Court's words, means: is the expert's theory or technique one that can be and has been tested? The Advisory Committee's notes to Rule 702 further describe the "testability" factor in these words: "whether the expert's theory can be challenged in some objective sense, or whether it is instead a subjective conclusory approach that cannot be assessed for reliability." Courts have held unreliable, for example, experts' untested opinions.5

Gatekeeping for reliability also affects admissibility of computer simulations and reconstructions.6 Reliability and admissibility have become somewhat meshed issues. A test or demonstration may be "relevant" or look "substantially similar," yet flunk the "reliability" test. Since "testability" is a factor in assessing reliability, both plaintiffs and defendants may want to conduct testing to substantiate their own experts' methodologies or opinions and, in turn, to shoot down those of the opposing experts.

The importance of this "new reliability kid on the block" is reflected in Samuel v. Ford Motor Co.,7 a remarkably detailed opinion excluding certain driving maneuver tests performed under the supervision of plaintiff's expert and intended to show that the 1993 Ford Aerostar van was defective because it had a propensity to roll over. One problem with the tests was that the professional driver exceeded the "normal range" of maximum steering inputs a nonprofessional driver is willing to make in an unexpected event. A second problem was that the steering rate, i.e., how fast the driver turned the wheel, was well above the average maximum steering rate for all successful drivers. Thus, the steering maneuvers were relatively extreme and violent affecting the reliability of the results and the expert opinion based on them. We discussed the Samuel case in a column which also quoted the court's "Daubert/Kumho Worksheet," an information outline the parties had to complete.8 The "worksheet" is still a useful outline to help measure the indicia of reliability that test evidence may (or may not) possess.

The influence of the Daubert/Kumho reliability era and a reinvigorated Federal Evidence Rule 702 upon admissibility of test evidence was discussed lucidly by trial specialist Jonathan Hoffman in his 2008 law review article, "If the Glove Don't Fit, Update the Glove: The Unplanned Obsolescence of the Substantial Similarity Standard for Experimental Evidence."9 We discussed some of the issues and highlights of Mr. Hoffman's article in our May 2008 column, "Admissibility of Test Evidence: A Fresh Approach."10 Clearly, out-of-court test evidence, live and animated videos, computer simulations, and other illustrative demonstrations now take on a layer of additional complexity.

In the case, plaintiff contended that the subject seat allowed too much slack during the accident while other designs would not. Plaintiff further asserted that the seatback was defective because it yielded rearward while other designs would not. To defend, conducted "spit tests" in which a vehicle with a seated surrogate of decedent's size is essentially turned on a spit to show how far the body can reach even when seatbelted as preferred by plaintiff's expert. The trial court excluded these defense tests because they could mislead the jury.

To counter the opinion of plaintiff's expert regarding the seat design, also offered photographs and the results of "pull tests" which showed how much force was needed to deform seats found in other cars which plaintiff's expert said were superior. Then there were "sled tests" simulating rear collisions at 32 mph to show the dangers of building stiffer seatbacks. Defendant also sought to show National Highway Traffic Safety Administration (NHTSA) rear collision testing performed to evaluate fuel systems but which also showed the movement of test dummies within the vehicles. The trial court excluded the defense tests, even those conducted by NHTSA.

The appellate court reversed for an abuse of discretion. The purpose of the defense tests was not to replicate the accident to prove that the seatback operated in a manner different from that proposed by plaintiff's expert. Rather, the tests supported otherwise valid defenses to the defective product claims. The trial court should not have applied the "substantial similarity" doctrine to evidence that was not intended to recreate the accident. Further, the demonstrative evidence depicted situations "sufficiently dissimilar" to the actual accident so that "there was little chance that the jury might be misled or confused." The trial court's probative-versus-prejudicial analysis also was "fatally flawed."11


Pitched battles over test evidence can be expected in products liability cases. It is necessary to discern clearly what the purpose of the testing is and then to apply the foregoing principles to advocate for or against the demonstrations. To the historical principles governing admissibility of tests one must now add the threshold "reliability" issue which can affect the very viability of expert opinions.

Michael Hoenig is a member of Herzfeld & Rubin.

1. 2010 Fla. App. LEXIS 8651 (Fla. Ct. App., 4th Dist., June 16, 2010), motion for rehearing en banc filed on July 1, 2010.

2. See e.g., Crossley v. General Motors Corp., 33 F.3d 818 (7th Cir. 1994); Muth v. Ford Motor Co., 461 F.3d 557 (5th Cir. 2006); Harvey v. General Motors Corp., 873 F.2d 1343 (10th Cir. 1989); Jackson v. Fletcher, 647 F.2d 1020 (10th Cir. 1981); Swajian v. General Motors Corp., 916 F.2d 31 (1st Cir. 1990); Whitten v. Michelin Americas Research & Development Corp., 2008 U.S.Dist. LEXIS 60415 (W.D. Tenn. July 25, 2008); Bado-Santana v. Ford Motor Co., 482 F.Supp. 2d 192 (D. P.R. 2007).

3. Daubert v. Merrell Dow Pharmaceuticals Inc., 509 U.S. 579 (1993); Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137 (1999).

125 Broad Street, New York, New York, 10004 - Phone: 212-471-8500 - Fax: 212-344-3333
©2004 -2014 Herzfeld & Rubin, P.C.