Ask the Expert

Daubert Without Prejudice: Achieving Relevance and Reliability Without Randomness

Harold J. Bursztajn, MD
Linda Stout Saunders, JD
Archie Brodsky, B.A.

Dr. Harold J. Bursztajn, Associate Clinical Professor and Co-Director, Program in Psychiatry and the Law, Harvard Medical School at the Massachusetts Mental Health Center, practices both as a clinician and as a forensic psychiatrist consulting locally and nationally to attorneys and institutions. [1] Linda Stout Saunders, a former Special Judge for Concord District Court (Concord, N.H.), is President-elect of the New Hampshire Trial Lawyers Association and Director of Jurimetrics Associates, which offers consulting services in the area of law and psychiatry. Archie Brodsky is Senior Research Associate, Program in Psychiatry and the Law, Harvard Medical School at the Massachusetts Mental Health Center. The authors are affiliated with Judge and Jury Decision Making, a research and consultation service dedicated to understanding the processes by which judges and juries make decisions when complex scientific evidence is at issue.

The O.J. Simpson verdict raises serious questions about how much impact complex scientific evidence really has on a jury, as against other issues raised, explicitly or implicitly, at trial. What can we learn from this case about how prejudicial factors may affect a jury's consideration of such evidence?

The Simpson case was an object lesson in the unintended consequences of the presentation of complex scientific evidence -- unanticipated ways in which such evidence may be misperceived by the jury. Indeed, the more complex the evidence, the more ways it can be misperceived. When the evidence is as far removed from familiar experience as DNA evidence is, its potential impact must be assessed in the context of the jurors' preexisting biases.

For example, both blood and DNA can evoke implicit associations with race. These associations might not come into play in a trial that did not revolve around racial issues. But in a trial that became as highly charged with racial imagery and symbolism as that of O.J. Simpson, the testimony about blood and DNA testing, as the jurors heard it, might well have amplified the defense focus on an alleged racially motivated police conspiracy. [2] Would you say, then, that evidence that is not in itself unfairly prejudicial may become unfairly prejudicial in light of the jurors' preexisting biases? Exactly. In such cases the prejudicial character of the testimony lies not in the evidence alone, but in the way the evidence may arouse latent biases in the jury when presented in adversarial and anxiety-filled proceedings. And that makes the judge's task of balancing the probative value of proffered testimony against its prejudicial impact [3] all the more challenging. As with any evidence, and perhaps even more than most, the potential prejudicial impact of complex scientific or technical information will vary greatly depending on the form in which it is admitted and presented. Does the U.S. Supreme Court's decision in Daubert vs. Merrell Dow Pharmaceuticals, Inc., [4] which concerned the admissibility of expert testimony about scientific, technical, or other specialized knowledge, bear upon this question?

It is highly relevant. In Daubert the Court ruled that the adoption of the Federal Rules of Evidence in 1975 had superseded the seventy-year-old Frye test, under which the scientific basis of expert testimony must have gained "general acceptance in the particular field to which it belongs" to be admissible. [5] The Court stated, "The Rules--especially Rule 702--place appropriate limits on the admissibility of purportedly scientific evidence by assigning to the trial judge the task of ensuring that an expert's testimony both rests on a reliable foundation and is relevant to the task at hand." [6] The Massachusetts Supreme Judicial Court has followed suit in replacing Frye with a Daubert-like standard. [7] Daubert opened up the question of the admissibility of relatively novel, leading-edge scientific evidence. The Court recognized that, as courtroom testimony draws more and more on advanced science and technology (ranging from DNA testing to computer simulations), trial judges must make increasingly sensitive determinations of the reliability and relevance of such testimony. The Court's decision in Daubert provides guidelines by which judges can make these determinations with the help of the expert witnesses retained by the opposing parties. What we need now is an analogous process for determining the possible prejudicial impact of complex scientific evidence, so that judges will be equally well-informed when they consider the other side of the delicate balance of admissibility.

What would be an example of how complex scientific evidence might have an unfairly prejudicial impact on the jury?

Sometimes technical evidence finds its way into the courtroom even when the scientist responsible for developing the technique finds the use made of it objectionable. For example, a psychological test was introduced in court to show that a defendant was malingering a psychiatric illness. However, by the interpretive protocols established by its developers, the test in this case did not produce a valid profile of the defendant for legal purposes (or, worse yet, even for clinical purposes). Thus, the test probably should have been ruled inadmissible because it was neither relevant nor reliable when used improperly. Yet when the profile it did produce was presented to the jury in a graphic display, especially without any notation that it was invalid, even the most vigorous testimony to that effect by the expert witnesses retained by the opposing side could not overcome the impact of the initial visual display, with its trappings of scientific precision.

Sometimes the selective admission of psychological evidence can be prejudicial in that it gives the jury an unbalanced view of the psychological dynamics of the case. For example, in a recent New Hampshire case in which an attorney was accused of sexually abusing a client, [8] the prosecution was permitted to introduce a treating clinician's testimony based on the use of questionable techniques such as regression under hypnosis and anatomical dolls. The expert retained by the prosecution was allowed to testify that the alleged victim's delayed "recall" was a classic example of dissociation. The expert retained by the defense, on the other hand, was not permitted to testify about the psychological process of projection, which might have led the alleged victim to "recall" an event that did not actually occur. This, in spite of the fact that both dissociation and projection are recognized as processes in the American Psychiatric Association's Diagnostic and Statistical Manual (Fourth Edition). In this instance, evidence that in itself need not have been unfairly prejudicial may have had an unfairly prejudicial impact because it was not balanced by equally relevant and reliable opposing evidence.

If evidence is introduced that may be unfairly prejudicial, won't the judge's instructions to the jury to disregard it, or to consider it only in some limited context, cure the problem?

Empirical research on how juries reach decisions confirms what attorneys and judges know from practical experience -- that juries, notwithstanding the judge's instructions, cannot easily dismiss what they have seen and heard. [9] Therefore, the potentially prejudicial character of any testimony needs to be considered thoroughly before the evidence is introduced. How can complex scientific or technical evidence be evaluated for prejudicial impact? The courts have not addressed the determination of prejudicial impact of complex scientific evidence in the systematic way in which the Supreme Court in Daubert outlined a process for determining the reliability and relevance of that evidence. The Daubert decision gave trial judges some guidelines for exercising their responsibility to rule on the admissibility of scientific evidence. In order for trial judges to strike a proper balance between the probative and the prejudicial, it would be useful to have similar guidelines on the other side of the scale. A judge may need to consider the possible prejudicial impact of highly complex evidence while taking into account preexisting juror biases that may be subtle and not at all evident in voir dire, but which may emerge in the course of the supercharged emotional atmosphere which one side or the other is invested in creating. Judges who evaluate the prejudicial impact of complex scientific evidence might find it helpful to have flexible guidelines for judicial determinations similar to those articulated for reliability by Daubert. They also might wish to solicit "second opinions" at the trial level, from experts in the area of individual and group decision making under conditions of uncertainty, rather than simply await the de facto judicial second opinions delivered from hindsight at the end of a long appeals process.

Why can't such determinations be left to the cultivated intuition of judges?

Intuition may work well when it comes to common, familiar types of evidence. Indeed, wise, experienced judges can often make very sound decisions even when faced with a kind of evidence with which they have no prior experience. But even the most experienced judge, when presented with unfamiliar forms of evidence or presentations of evidence (such as computer simulations), may want to take time out for some second opinions about its possible prejudicial impact.

How can the judge obtain these second opinions?

By conducting hearings outside the presence of the jury, in which experts in the psychology of information processing and individual and group decision making could address the question of the unfairly prejudicial impact of the proffered evidence. Such experts, while increasing the cost to the opposing parties at the trial level, could decrease the enormous financial and emotional costs of a hindsight-based appeals process. Also, the trial judge would make the final determination of the validity and relevance of the experts' opinions in a manner less subject to the possibility of a reversible error.

Would the experts who render opinions on the potential prejudicial impact of the proffered evidence be the same experts who render opinions on its probative value -- that is, experts on the scientific or technical field in question?

No. The judge is in a position to decide on the probative value of evidence by listening to experts in the field of knowledge on which the evidence bears. However, to render an opinion on the subtle, multifaceted interaction between a particular type of evidence (or presentation of evidence) and jurors' preconceptions, taking into account community attitudes, requires expertise in the very process of decision making. Both the judge determining the admissibility of evidence and the jury weighing the evidence are making decisions under conditions of uncertainty. How people process information to make judgments under conditions of uncertainty, whether in the medical clinic or in the courtroom, is itself the subject of a growing body of scientific knowledge. [10] This science of decision making is a necessary foundation for understanding the psychological impact of various kinds of evidence in different social and legal contexts. It is the primary area of expertise on which judges need to draw in assessing potential prejudicial impact.

What are some of the questions on which this expertise in decision making can shed light? To begin with, there is the psychology of prejudice -- the historical and cultural roots of racial, ethnic, and gender stereotyping and the psychological functions that stereotypes serve for the individual or the community. In cases where a person's mental state is at issue, stereotypes about madness may enter into the jury's thinking, sometimes in conjunction with racial, ethnic, and gender stereotyping. [11] (It is essential, therefore, to address the possibility of unfair prejudice in insanity-defense cases, where the evidence is by its very nature unfamiliar, testimony is proffered about states of mind alien to common sense and experience, and it is heard through a filter tinged with preconceptions about mental illness.) The judge's task of bringing these influences to light is complicated by the fact that prospective jurors' biases may not be apparent in voir dire, either to the attorneys or to those being questioned. People sometimes unintentionally, albeit skillfully, deny and conceal their biases, both to others and to themselves.

For this reason, in highly sensitive cases, a single examination in voir dire may not be sufficient to discover the latent preconceptions that may bias a juror's consideration of the evidence. In a series of psychological experiments, the subjects (some of whom were judges) assessed a hypothetical situation differently when one word in the vignette was replaced with a seemingly equivalent word. [12] The fact that a small variation in language can lead people to make different choices indicates that prejudicial trial testimony may appear in quite subtle forms. Experts in the psychology of decision making can use this experimental model to explore underlying biases that might be triggered by the testimony being considered for admissibility.

Are jurors' reactions more uniform and reliable when the evidence is statistical -- that is, when it is expressed in numbers rather than words?

It is a mistake to assume that juries can reliably separate the wheat from the chaff when it comes to expert testimony. With statistical evidence, the selection and arrangement of the numbers and the way they are framed with words give attorneys and experts ample opportunity to engage the preconceptions of jurors. In weighing the testimony of opposing experts, statistical or otherwise, jurors fall back on shorthand decision strategies, called heuristics, that we all use to process the deluge of information that hits us daily. [13] For example, a few of these unconscious or half-conscious rules of thumb have to do with the sequence in which the jury hears the two experts:

Anchoring and adjustment -- The juror makes a provisional judgment on the basis of the testimony heard first and then adjusts his or her opinion on the basis of later testimony. With this heuristic the earlier testimony remains dominant.

Availability -- The juror decides on the basis of the most readily available information. This may be the more recent testimony or that which was presented in the most comprehensible form.

Vividness -- The juror decides on the basis of the information he or she recalls most vividly. This may be the more recent testimony or that which made the stronger impression when it was presented.

Keep in mind that these are only a few of the largely automatic thought processes that may come into play, especially in the service of preexisting beliefs.

Can verbal testimony or instructions be overwhelmed by the direct impact of visual evidence?

Yes. For example, when John Gotti was acquitted of Federal racketeering and conspiracy charges in New York in 1987, the jury seems to have been strongly influenced by a chart prepared by defense counsel detailing the lengthy criminal records of seven prosecution witnesses against Gotti. [14] In a case of alleged sexual abuse of children (such as the Fells Acre case in Middlesex County), not allowing the defendants to face their accusers gives the jury a powerful visual image of the defendants as people from whom children must be protected. However firmly the judge may instruct the jury to disregard the physical placement of the defendants and the children, jurors may have difficulty dismissing the natural inference that "where there's smoke, there's fire." These are the kinds of dynamics about which a judge may seek advice of an expert in decision making.

The more we explore this area, the more complex it becomes.

Exactly. To make that complexity more manageable, judges in some cases may wish to request expert reports from specialists in the psychology of decision making before deciding on the admissibility of potentially prejudicial evidence. Although this terrain is admittedly treacherous, informed evidentiary determinations by trial judges may build up a useful body of case law on other kinds of prejudicial evidence besides prior-act testimony, about which a considerable body of case law already exists. [15] If, in spite of (or in the absence of) all precautions, prejudicial evidence is introduced, what remedies are available? Attorneys who believe that testimony unfairly prejudicial to their client has been introduced can try to neutralize this testimony in questioning their own experts and in closing argument. They can also ask the judge to hold a hearing to weigh the prejudicial character of the testimony already given. If the judge (with or without such a hearing) believes that the testimony in question really is unfairly prejudicial, the judge can admit the error and instruct the jury to disregard this testimony precisely for that reason. The judge can do likewise when unfairly prejudicial statements are made in closing argument. However, if the judge's instruction is felt by an attorney to be ineffective in curing the harm done by the introduction of the unfairly prejudicial evidence, the attorney can move for a mistrial. Finally, the attorney has the option of appealing the verdict. Still, since these after-the-fact remedies tend to be relatively ineffective, they are better avoided with some thorough preparation.

To improve that preparation, are there any additional remedies that are not now being fully utilized?

There is tremendous untapped potential in the use of direct questioning of expert witnesses by jurors and judges. Typically, jurors submit written questions to the judge, who, after screening them with the respective attorneys, poses them to the witness. The Federal Rules of Evidence establish the right of a federal trial judge to question witnesses, [16] and various federal and state courts have held that it is within a trial judge's discretion to permit questions from jurors. [17] Experiments in a number of jurisdictions have shown that juror questioning of witnesses can enhance juror participation and responsibility, especially in complex cases. [18] If this approach were more widely practiced, juries in cases involving complex or otherwise potentially prejudicial evidence would, in that sense, be acting more like grand juries, which are permitted to ask questions of the attorneys or witnesses before them. Judges could then use the methods of experienced group therapists [19] to engage jurors in a collaborative dialogue with witnesses that would change rather than reinforce any preexisting juror biases. For their part, experts, once alerted to possible prejudicial reactions, can address these by the way they present their own evidence. For example, an expert introducing DNA evidence to a tentative, suspicious jury could begin by talking about the many ways in which DNA testing is used, including nonlegal contexts (such as diagnosing prenatal conditions) and noncriminal legal contexts (such as helping adopted children find their biological parents). [20] As background, the expert might relate the story of the woman who claimed for decades to be the Russian Czar's lost daughter Anastasia. Her DNA was tested posthumously against that of Britain's Prince Philip (a descendant of the Russian royal family) and that of a man related to the Polish factory worker who skeptics believed -- and the DNA ultimately showed -- had transformed herself into "Anastasia." Familiarizing jurors with the wider range of uses of DNA evidence can help create a platform for open-minded consideration and thoughtful deliberation by jurors otherwise caught up in the flood of emotion which a contentious criminal trial can unleash.

  1. Questions regarding this or future columns may be addressed to Harold J. Bursztajn, M.D., 96 Larchwood Drive, Cambridge, MA 02138. Tel. (617) 492-8366, Fax (617) 441-3195 (e-mail:
  2. "'Unfair prejudice' within its context means an undue tendency to suggest decision on an improper basis, commonly, though not necessarily, an emotional one." Fed. R. Evid. 403 (Advisory Committee's Note). The Supreme Court of Vermont has elaborated as follows: "Evidence is unfairly prejudicial if its primary purpose or effect is to appeal to a jury's sympathies, arouse its sense of horror, provoke its instinct to punish, or trigger other mainsprings of human action that may cause a jury to base its decision on something other than the established propositions in the case." State v. Bruyette, 158 Vt. 21, 604 A.2d 1270, 1274 (Vt. 1992) (brackets and quotations omitted).
  3. Fed. R. Evid. 403.
  4. Daubert v. Merrell Dow Pharmaceuticals, Inc., 113 S. Ct. 2786 (1993).
  5. Frye v. United States, 293 F. 1013, 1014 (D.C. Cir. 1923).
  6. Daubert, 113 S. Ct. at 2790.
  7. Commonwealth v. Lanigan, 419 Mass. 15 (1994).
  8. New Hampshire v. Ross, Hillsborough County, Northern District, Superior Court S-94-441, 94-1628 (1994).
  9. "There are many procedures at trial whereby jurors are instructed to ignore information but in which they apparently do not. These include admonitions to disregard trial testimony, [and] instructions to use evidence for one limited purpose...." Jonathan D. Casper and Kennette M. Benedict, The Influence of Outcome Information and Attitudes on Juror Decision Making in Search and Seizure Cases, in Inside the Juror: The Psychology of Juror Decision Making 65, 82 (Reid Hastie ed., 1993) (citations omitted). "When seemingly probative information is ruled inadmissible because of due process violations, jurors may nevertheless incorporate it in their probability-of-guilt assessment." Robert J. MacCoun, Experimental Research on Jury Decision-Making, 244 Science 1046, 1047 (1989) (citations omitted). "The ways in which inadmissible or extralegal evidence may influence a juror can range from blatant intentional disregard of the judges' instructions to subtle alterations in the construal of the rest of the evidence in jurors who make a sincere effort to ignore the inadmissible evidence and believe that they have succeeded." Phoebe C. Ellsworth, Some Steps Between Attitudes and Verdicts, in Inside the Juror: The Psychology of Juror Decision Making 42, 47 (Reid Hastie ed., 1993) (citations omitted).
  10. See Harold J. Bursztajn et al., Medical Choices, Medical Chances: How Patients, Families, and Physicians Can Cope With Uncertainty (Routledge 1990) (1981).
  11. See Sander L. Gilman, Difference and Pathology: Stereotypes of Sexuality, Race, and Madness (1985).
  12. Harold J. Bursztajn, Benzion Chanowitz, Thomas G. Gutheil, and Robert M. Hamm, Micro-effects of Language on Risk Perception in Drug Prescribing Behavior, 20 Bull. Am. Acad. Psychiatry & Law 59 (1992). See also Harold J. Bursztajn et al., Medical and Judicial Perceptions of the Risks Associated With Use of Antipsychotic Medication, 19 Bull. Am. Acad. Psychiatry & Law 271 (1991).
  13. See R. M. Hogarth, Judgment and Choice: The Psychology of Decision (1980); Judgment Under Uncertainty: Heuristics and Biases (Daniel Kahneman et al. eds., 1982).
  14. Edward R. Tufte, Envisioning Information 30-31 (Graphics Press, 1990). See also Photographs and Maps Go to Court (Larry Gillen ed., American Society for Photography and Remote Sensing, 1986); Gregory P. Joseph, Modern Visual Evidence (1989).
  15. See Edward J. Imwinkelreid, Uncharged Misconduct Evidence (1984 & Supp. 1994); Linda Stout Saunders, Annual Survey of New Hampshire Evidence Law - Part I: Focus on Rule 404(b), 17 Trial Bar News (New Hampshire Trial Lawyers Ass'n, forthcoming December 1995).
  16. Fed. R. Evid. 614(b).
  17. Carrie Shrallow, Note, Expanding Jury Participation: Is It a Good Idea? 12 U. Bridgeport L. Rev. 209 (1991).
  18. Id. at 221-226. See also Sand and Reiss, A Report on Seven Experiments Conducted by District Court Judges in the Second Circuit, 60 N.Y.U. L. Rev. 423 (1985); Heuer and Penrod, Increasing Jurors' Participation in Trials: A Field Experiment with Jury Note-taking and Question Asking, 12 J. L. & Human Behav. 231 (1988).
  19. See Max Day and Elvin V. Semrad, Psychoanalytically Oriented Group Psychotherapy, in Psychoanalytic Techniques: A Handbook for the Practicing Psychoanalyst 511 (Benjamin B. Wolman ed., 1967).
  20. Patricia M.L. Illingworth, personal communication.