Science in the Jury Box: Jurors' Views and Understanding of Mitochondrial DNA Evidence

Policy makers, pundits, and scholars have all raised questions about how jurors understand and apply scientific evidence. In the current study, 480 jury pool members observed a mock trial that included expert testimony about mitochondrial DNA (mtDNA) evidence purportedly linking a defendant to a crime. As a group, the jurors showed moderately good command of the biological facts relating to mtDNA evidence, although some jurors made errors in defining mtDNA and in making inferences about its relevance to the trial. Comprehension was higher after jury deliberation and among jurors with more education. A minority of jurors expressed reservations about science, concern about the reliability of the mtDNA evidence, and suspicion that the mtDNA evidence was contaminated. Science in the Jury Box 3 Science in the Jury Box: Jurors’ Views and Understanding of Mitochondrial DNA Evidence Complex scientific evidence has become ubiquitous in both civil and criminal trials (Faigman et al., 2005-2006; Gross, 1991). In the words of one judge: “The demand for expert testimony by litigants has become insatiable” and “an astounding number of ‘expert’ consultants and professional witnesses in virtually every field of human endeavor have arrived on the scene” (Loeffel Steel Products, Inc. v. Delta Brands, Inc., 2005:1106). Yet, lawyers, litigants, and policy makers have voiced concerns about whether juries can comprehend and properly apply complex scientific or technical evidence (for reviews see Cheng, 2005; Diamond & Rose, 2005; Kaye, 2004). These concerns underlie a remarkable trilogy of Supreme Court cases and an amendment to the Federal Rules of Evidence limiting the evidence that juries may hear (Daubert v. Merrell Dow Pharmaceuticals, Inc., 1993; General Electric v. Joiner, 1997; Kumho Tire Co. v. Carmichael, 1999; Federal Rule of Evidence 702). Doubts about the capacities of lay jurors have even prompted the consideration of a “complexity exception” to the right to trial by jury in civil cases (In re Japanese Electronic Products Antitrust Litigation, 1980; Lempert, 1981-1982; Lilly, 2001). Lilly, for example, argues that a complexity exception “seems especially appropriate . . . when a forthcoming trial is likely to be protracted and involve difficult technical or scientific issues” (Lilly, 2001:80). On the criminal justice side, prosecutors and journalists have offered the view that exposure to television shows like CSI have led jurors to become extraordinarily Science in the Jury Box 4 demanding in criminal cases, insisting on nearly infallible scientific evidence linking a defendant to a crime before they will convict (Podlas, 2006; Shelton, Kim & Barak, 2006; Schweitzer & Saks, 2007; Tyler, 2006). Empirical study of the CSI effect is in its infancy and the results are mixed. Schweitzer & Saks (2007), for instance, found that undergraduate students who watched CSI were more critical of forensic evidence than their nonviewer colleagues. In contrast, Podlas (2006) found no significant relationship between viewing CSI and treatment of forensic evidence. Shelton et al. (2006) surveyed jury pool members and found an inconsistent pattern of television viewing and expectations about forensic evidence, but pointed out that close to half of the summoned jurors expected that the prosecutor would offer some form of scientific evidence in every criminal case. The debates over jury competence and the conflicting data over the influence of popular culture indicate that on both practical and theoretical grounds, it is worthwhile to develop a robust understanding of the factors that influence juror expectations, comprehension, and use of scientific evidence. Jury Comprehension of Scientific Evidence Many research studies of actual juries conclude that their fact finding is basically sound, even in cases with complex evidence (Diamond, 2006; Kalven & Zeisel, 1966; Myers, 1979; Hans & Vidmar, 2004; Vidmar, 1998). Research shows, for example, that the strength of the evidence presented at trial is the prime determinant of the jury’s verdict (Hannaford-Agor et al., 2002; Hans et al., 2003; Eisenberg et al., 2005; Myers, 1979; Visher, 1987). A number of studies have documented that trial judges who preside over criminal jury trials agree with the vast majority of jury verdicts (Hannaford et al., 2002; Kalven & Zeisel, 1966). The agreement rate with the jury’s verdict is about the Science in the Jury Box 5 same whether the trial evidence is low or high in complexity (Heuer & Penrod, 1994; Kalven & Zeisel, 1966; Eisenberg et al., 2005). Taken together, this work suggests that whatever problems jurors might have with comprehending trial evidence are not severe enough to produce distinctly different outcomes from the assessments of ordinary judges across a range of cases. Interviews with jurors and case studies produce more mixed impressions. Jurors themselves say that scientific, statistical, and technical expert evidence is challenging (Cecil, Hans & Wiggins, 1991). After questioning the jurors in a tort case involving toxicological and epidemiological evidence, Sanders concluded that the jury's deliberations did not reflect a full understanding of the case but that the defense lawyers’ presentation and the judge's instructions to the jury may have been contributing factors (Sanders, 1998). From post-trial interviews with jurors who decided an asbestos case, Selvin and Picus (1987) likewise reported that jurors misunderstood some of the scientific evidence about the development of asbestosis. Lempert’s (1993) review of 13 complex jury trials, a number of which included scientific and technical evidence, uncovered some mistakes in jury comprehension but observed that they were often traceable to problems in the attorneys’ presentations of complex evidence or by jury instructions. He concluded that nonetheless juries usually reached defensible verdicts. Experimental mock jury research also identifies areas of vulnerability in lay citizens’ use of scientific evidence. Thompson and his colleagues have undertaken a series of experiments to determine how lay citizens use statistical and scientific evidence (Thompson & Schumann, 1987; Thompson, 1989; Kaasa, Peterson, Morris & Thompson, 2007). Participants in several of these studies have fallen prey to fallacious reasoning Science in the Jury Box 6 about statistical arguments (Thompson & Schumann, 1987; Thompson, 1989). Kaasa et al. (2007) discovered that although mock jurors as a group used statistics appropriately, giving differential weight in line with the diagnosticity of the forensic evidence, a subgroup of jurors who expressed concerns about their ability to handle statistical reasoning did not correctly differentiate between evidence that varied in diagnosticity. The introduction of DNA as forensic evidence has spawned a number of experimental studies designed to examine how lay persons evaluate the presence of a match between DNA samples from a defendant and from a crime scene. In general, the research shows that study participants tend to give statistical information about the DNA match less weight than might be prescribed by probability theory (see, for example, Koehler, 2001; Nance & Morris, 2002, 2005; Schklar & Diamond, 1999). Schklar and Diamond’s (1999) important study found that their participants were concerned about laboratory error and other problems with the forensic DNA samples. The authors pointed out that in evaluating juror competence against a probability model, one needed to take into account jurors’ expectations and presumptions about the quality and integrity of DNA evidence, not simply its statistical power. Thus, taken as a whole, the body of research suggests that juries are likely to be reasonably competent in handling scientific evidence, but that in some circumstances, particularly when statistics are presented, jurors may make systematic errors. Citizens, Science and the Law An inquiry into jurors’ treatment of scientific evidence is important not only because of its practical and policy significance, but also because it has the potential to deepen theoretical understanding of how citizens rely on science within the distinctive Science in the Jury Box 7 context of legal decision making. Scientific evidence is typically presented by experts whose claims rest on broad assumptions about the legitimacy of the underlying science. Jasanoff (1995) writes that while scientific expertise is often seen as straightforward and autonomous, operating outside of the law, this is a simplistic view that ignores the institutional setting of a legal trial: “scientific claims, especially those that are implicated in legal controversies, are highly contested, contingent on particular localized circumstances, and freighted with buried presumptions about the social world in which they are deployed” (Jasanoff, 1995, p. xiv). The selection and preparation of expert witnesses and the adversarial setting within which they testify all shape the content and meaning of expert testimony about scientific evidence (Gross, 1991). Confronted with science in the courtroom, jurors face two often competing sources of authority, science and law. Examining how jurors consider and assess scientific expert testimony can provide a window into citizens’ thinking about these domains of expertise. Systematic study of public attitudes toward science present an intriguing and complex picture. National survey data (National Science Board, 2004; 2006) indicate that most Americans hold very positive general views toward science and technology. Yet, they also reveal that a significant proportion of Americans expresses reservations about science and its potential for destructive social change. Furthermore, many Americans have poor understanding of scientific concepts (National Science Board, 2004). In these circumstances, one may well wonder how citizens with different background assumptions about the nature and potential of science evaluate scientific evidence in the contested domain of a legal dispute. Science in the Jury Box 8 The current project explores questions about citizens and science through a mock jury study, employing a criminal trial with expert testimony on mitochondrial DNA (mtDNA) sequencing. The type of DNA testing typically used in crime investigations analyzes nuclear DNA (nDNA) coiled in the nuclei of most types of human cells (Adams, 2005; Kaye & Sensabaugh, 2000). Outside the nucleus is other DNA, contained in organelles known as mitochondria. Although the mitochondrial DNA sequence is about 200,000 times shorter than the nuclear DNA sequence, a human cell contains hundreds or thousands of mitochondria but only one nucleus. Consequently, when the number of cells in biological material recovered from a crime scene is insufficient for nDNA analysis to be performed, mtDNA sequencing is often feasible (U.S. Department of Justice, 2006). MtDNA sequencing can be conducted with DNA extracted from teeth, bones, and even a few strands of human hair. MtDNA is passed through the maternal line, and all individuals in this line of descent have the same mtDNA sequence. Because of its shorter sequence and its maternal lineage, mtDNA is less individualizing than is nDNA. Nonetheless, it has proved useful in forensic identification, and mtDNA evidence has been presented in many American courts (Faigman et al., 2005-2006). The study was designed to ascertain whether members of a jury pool were able to understand this evidence, and how they used it in deliberations and in decision making in a mock jury trial. Findings from the research study are used in the present article to explore the dimensions of juror understanding about biological science in the context of a criminal case and to examine what shapes lay judgments of science in the courtroom. The study also tested the impact of a variety of jury innovations on the decision making of Science in the Jury Box 9 mock jurors. Those results have been presented elsewhere and are not considered in this article. MtDNA evidence is well suited to investigating the theoretical and policy questions raised in this article. One attraction of using mtDNA evidence was that, at the time of the experimental study, it constituted a novel form of scientific evidence that had not been used extensively in the courtroom (Cheng, 2005). In fact, most of our study participants said they had not previously heard about mtDNA. Thus, a case using mtDNA as forensic evidence provided an opportunity to observe how jurors confront a new scientific topic presented in an adversarial context. The biological underpinnings of mtDNA are complex and thus provide a good vehicle for assessing how jurors from a wide range of backgrounds handle the complex expert testimony. Furthermore it offers a concrete setting in which to examine the impact of both positive and negative attitudes toward science. An interesting aspect of mtDNA is that it bears some relationship to the more widely known nuclear DNA. The relationship between mtDNA and nDNA evidence could potentially assist jurors by providing a familiar framework for the understanding of a new scientific concept. However, the relationship to nDNA could also make evidence evaluation more challenging by introducing a source of confusion. The likely preexisting belief of lay jurors is that DNA permits precise identification of individuals. The full sequence of the three billion or so base pairs of DNA encapsulated in the nucleus of a human cell is unique to each individual (with the exception of identical twins). Even though very little of the full genome is analyzed in forensic testing, the number of sites of variation that can be analyzed in an ordinary case are sufficient to distinguish among Science in the Jury Box 10 virtually everyone other than identical twins. However, mtDNA is not unique, and the existing tests of nDNA are more powerful for individualization than is mtDNA sequencing. Do jurors appreciate this difference, or do they think mtDNA evidence is as revealing as nDNA evidence? How is its reliability calibrated? Or do jurors simply dismiss mtDNA evidence as unreliable, since it is not as informative as nDNA? Method Participants A total of 480 jurors, 60 mock juries of eight persons, participated in the study. The participants were drawn from Wilmington, Delaware, residents who appeared at the New Castle County Courthouse in response to summonses for jury duty from October 14 through December 16, 2003. Potential jurors volunteered to participate in the study if they were not needed to serve on a regular jury. After jury pool members received their usual orientation from the court, one of the researchers addressed the jury pool, describing the research project and the chance for individuals from the jury venire to participate in the study if they were not needed for jury duty. The purpose of the study, the approximate time commitment, the fact that the jurors would be asked to complete questionnaires and deliberate with other mock jurors, the fact that the group discussions would be videotaped, and the fact that they would receive $50 as remuneration were all described. During the subsequent roll call, a juryoffice staff member typically asked jury-pool members to indicate whether they were interested in participating in the study should they not be needed for jury service that day. On four separate days, the volunteer rate, calculated as the proportion of volunteers to the total number of jurors present, was an average of 74%, ranging from a low of 64% to a Science in the Jury Box 11 high of 97%.When it was clear that no more jurors or only a small number would be needed for actual trials, court staff randomly selected a set of jurors from their master list of the remaining jurors who had previously volunteered for the mock jury study. Typically, there were at least sixteen volunteers, enough for two eight-person mock juries. Demographic Characteristics. The demographic characteristics of the mock jurors and the pool of jurors who reported for jury duty during the period of our study were comparable. A total of 3,381 jurors reported for jury duty during that time period. We compared the gender, race, and age proportions in the mock jury sample and in the full pool of people reporting for jury duty. Women comprised 53% of the pool and 52% of the sample; whites were 77% of the pool and 79% of the sample. Likewise, the representation of specific age ranges for the two samples were all within one to two percentage points of each other. Reported educational attainment was somewhat different in the two groups. However, the differences at the lowest and highest education levels were modest, and the other reported education differences were most likely due to the fact that the court’s jury pool questionnaire and our study questionnaire differed in the options provided. (The research project questionnaire included the “some college” option but the jury pool questionnaire did not.) Five percent of the jury pool and 2% of the mock jury sample said they had less than a high school degree; 49% of the jury pool reported a high school degree whereas 55% of the mock jurors said they were either a high school graduate or had some college courses; 33% of the jury pool and 29% of the mock jurors had college Science in the Jury Box 12 degrees; and 12% of the jury pool and 14% of the mock jurors had post-graduate education. In any event, a good range of educational backgrounds was represented. In sum, the mock jury sample constituted a reasonably close reflection of the jury pool in New Castle County, Delaware. Like the jury pool, it was predominantly white, about half female, and included a variety of educational backgrounds and ages. Jurors’ Background in Science. Most mock jurors had at least some high school courses in science and mathematics. They reported an average of 10 mathematics and science courses in high school and college, with a mode of 4 courses. The range was relatively wide, stretching from zero to 48 courses. A substantial proportion (196, or 43%) of the mock jurors reported some job experience related to mathematics or science. Of these, 77 said that the experience was moderate or substantial. The relevant job experience included the following: insurancerisk management work, chemistry, biotechnology, electrical engineering, science and mathematics teaching, dirt-grade calculations and ground-water contamination studies, medical technology, new drug testing, cardiac surgery, research science at a large research organization, computer programming, nursing, and laboratory technology for research and development of monoclonal antibodies. In sum, most jurors had taken at least some mathematics and science courses in high school or college and about a fifth of the sample had substantial mathematics or science experience on the job. Procedure After a group of jurors had been assembled, the participants were taken to conference rooms in the courthouse equipped for displaying a videotaped mock trial and for recording their deliberations. Here, the jurors completed an initial questionnaire Science in the Jury Box 13 asking for individual views about the reliability of different types of evidence, including eyewitness evidence, evidence provided by crime victims, police evidence, expert evidence, and DNA evidence. They rated the reliability of evidence on a five-point scale, where 1 corresponded to “not at all reliable” and 5 indicated “extremely reliable.” Jurors also responded to 7 items taken from the National Science Board (2004; 2006) to measure attitudes toward science. Four items aimed to assess the promise of science, while three items measured reservations about science. (Table 1 provides the items.) These items used a 4-point scale ranging from strongly agree to strongly disagree. The videotaped trial was then played for the jury, and jurors completed a second questionnaire that asked for initial reactions to the trial and mtDNA evidence. The researcher then provided each jury with a Jury Verdict form, instructed the jury to select a foreperson or presiding juror, turned on the video camera to record the group discussion, and left the room to await the completion of the jury’s deliberation. Once the mock jury had reached a unanimous verdict or declared itself hung, mock jurors completed a final questionnaire, asking for reactions to the jury’s verdict, their own individual views, mtDNA questions, and support for different jury reforms. Following completion of the final questionnaire, the mock jurors were debriefed and received payment. The videotapes of 57 of the 60 group deliberations were transcribed, and the software program Atlas.ti was used to examine mock juror deliberations qualitatively. Videotapes of the remaining three deliberations were defective and could not be transcribed. Materials: The State v. Jones Mock Trial Science in the Jury Box 14 The mock trial was based on State v. Pappas (2001), the first appeal in Connecticut from a successful prosecution relying on mtDNA evidence. We adapted material from the trial transcript and the reported decision of the Connecticut Supreme Court upholding both the admission of the mtDNA evidence and the defendant’s conviction. However, as described below, some evidentiary details were modified for the purposes of our research. The mock trial, filmed in a courtroom, included introductory instructions by a judge, opening statements by a prosecutor and a defense attorney, witness testimony, including competing experts who discussed mtDNA evidence, closing arguments, and legal instructions by the judge. An actual judge and practicing attorneys played those roles; the experts were a law professor specializing in forensic use of DNA and a biology professor. Other roles were played by actors. The mock trial pertained to an armed robbery of a bank by a lone, masked gunman who fled from the bank, pursued by local police. The police found the robber’s discarded blue sweatshirt and stolen currency. Laboratory examination of the sweatshirt revealed two human head hairs in the hood. An anonymous call directed the police to the defendant, Kevin Jones. The police collected a sample of his head hair. The two samples of hair were sent to the FBI crime laboratory for DNA analysis. The defendant was arrested and charged upon learning that his mtDNA matched that found in the sweatshirt hairs. In the actual trial, the jury heard only from the prosecution’s expert witness regarding the mtDNA evidence; the defendant did not present an expert. We modified the nonscientific factual evidence so that it was more ambiguous, making the mtDNA evidence more crucial to the jury’s decision. We also modified the prosecution’s expert witness testimony and added testimony from a defense expert who Science in the Jury Box 15 disputed the prosecution expert on several points. PowerPoint slides developed by the FBI were adapted for use by the prosecution expert witness; and we produced additional slides for the defense expert witness. The prosecution and defense expert slides were shown in the videotape during the experts’ testimony. Results Jurors’ Attitudes toward Science The initial questionnaire administered to jurors in the study included the seven National Science Board (2004; 2006) items tapping attitudes toward science (four on the promise of science, and three on reservations about science). As described earlier, the national survey data indicated that most Americans hold very positive general views toward science and technology. The same held true for our mock jury sample, as shown in Table 1. For instance, in 2004, 91% of a national sample agreed that “[s]cience and technology are making our lives healthier, easier, and more comfortable.” A comparable 95% of our mock jurors agreed with that statement. Similarly, 86% of a 2004 national sample and 82% of the mock jurors agreed with the statement that “[b]ecause of science and technology, there will be more opportunities for the next generation.”