The State of Science in the Courtroom: A Scientist’s Reflections From Conversations With Federal Judges

Rebekah Petroff, a scientist who worked at the Federal Judicial Center, examines how judges frequently grapple with making definitive decisions based on uncertain scientific evidence.

In September 2023, I entered the Thurgood Marshall Federal Judicial Building for the first time, excited to start my new role.¹ After a 10-month application process and two-week orientation, I was the lone American Association for the Advancement of Science science and technology policy fellow placed in the federal judicial branch. I had many ideas of what this year might bring, but, as a scientist with absolutely no legal background, first I had to simply learn about the judiciary.

For the rest of 2023, I attended Judicial Committee meetings, became familiar with Federal Rules and judicial policies, and immersed myself in the surprisingly fascinating world of law reviews. As my understanding of the judiciary grew, I realized my “science-centric” beliefs about the judiciary were not always aligned with reality. I had to fill a huge chasm of knowledge to realign my own ideology with the real world. To fill in missing pieces, I was determined to get inside the heads of people who knew much more than I did. With the support of my colleagues at the Federal Judicial Center (FJC), I had the opportunity to meet with seven current and retired federal judges to explore how they perceive and incorporate science in their courtrooms.

Meet the Judges

Collectively, the judges participating in this project had decades of legal expertise.² Prior to sitting on the bench, they had spent 10 to 30 years practicing law, averaging 22 years. In their practice, each confirmed that they had regularly considered scientific evidence in cases, although the amount varied. For example, one judge noted that “no more than five percent” of their caseload as an attorney included scientific evidence, while another reported that scientific evidence was involved in 80 to 90 percent of the cases they worked on. This, however, changed once the judges were appointed to the federal bench.

The judges served on the federal bench for an average of 24 years, ranging from 15 to 38 years. During their tenure as judges, all reported they regularly decided on the admissibility of scientific evidence. Some judges reported lower amounts of scientific evidence in their cases, ranging around 20 to 40 percent. One judge stated that this number was high — “too high to put a number on it.” Each judge had a different experience with science, both as a practicing attorney and while sitting on the bench, but all recognized the inherent presence that science had in the courtroom.

Our conversations were fun and informative and ended up traversing many topics across various types of science in both criminal and civil court cases. While we certainly talked about the day-to-day of their dealings with scientific evidence in criminal and civil cases, Daubert hearings, and related Federal Rules of Evidence, we also chatted about the similarities and differences between science and law, how the judges have shaped their own practices, and tiny treasures of knowledge they want other judges to know.

A Scientist Walks Up to the Bench — And It’s Not the Laboratory Bench

As a scientist, I have an in-depth understanding of how science works; the same could not be said of the law. We scientists are trained to follow the scientific method to gradually reduce uncertainty around facts about our natural world. As a result of this iterative process, scientists are never truly certain. We are quite fine with not absolutely knowing the facts. All we can ever say is that, for these conditions, in this exact experiment, the evidence suggests one version of the truth. We then weigh evidence over time. With enough evidence, we can develop theories, like that of the theory of relativity, which first described gravity. But even theories that are widely accepted and supported by strong evidence are continually reassessed. For example, quantum mechanics has usurped the theory of relativity; the equations we use to describe gravity simply don’t work in the world of quantum mechanics.

Still, I recognize that, as humans, we naturally look for certainty — I know I do in my daily life. Coming into my role at the FJC, I naively thought that one of the goals of the judiciary was to find certainty in matters of the law, as even the most contentious cases are decided one way or the other.

How wrong I was.

In my conversations with the judges, they all pointed out that the law is no different than real life. Even though we may be searching for certainty, there is very little in life (or in a courtroom) of which we can ever be truly certain. As best put by one judge, “There’s uncertainty in almost all circumstances, and one of the things that I’ve learned as a judge is there are virtually always two sides to the story.”

During our conversation, another judge conceptualized the differences between uncertainty in scientific and legal evidence by drilling down on a major difference between science and law. Scientists deal with averages, not individualized data. An epidemiologist, psychologist, or a biomedical scientist will take individuals’ data and examine patterns over the larger group. All conclusions are generalizations that we hope apply to broader populations. But, as this judge pointed out, “Courts individualize.” Judges said that it can be particularly challenging to take generalized scientific data into a court, where the court ultimately must decide facts about an individual. Another judge agreed with this perspective, saying that “science can easily live with generalities, but the legal system can’t.”

Adding to these complexities is the adversarial nature of legal proceedings. Many judges noted this inherent difference between science and law. As put by one judge, “Uncertainty is a part of what makes a litigation,” as both sides will inherently disagree on the reliability of the evidence or the extent to which something caused something else. Another judge said this difference can make it more difficult to fulfill their responsibilities:

[Attorneys] go out and hire the most glitzy expert that will give their side of the story. So, you’ve got an expert advocate for one side and an expert advocate for the other, but their goal is not presented in a neutral fashion. . . . The entire process is distorted because you don’t have objective people providing the input for the judge to make the decision.

This judge continued, explaining that the added factor of the competing versions of scientific “proof” takes an enormous amount of skill to be able to wade through. But another judge pointed out: “We’re not to determine who’s right or wrong. We’re just to determine whether both sides have valid arguments, and, if they do, then the jury can decide.”

Another complexity involved in evaluating uncertain evidence is time. Time in science and law are simply different. Science has the time to continually reevaluate and reassess previous experiments. In this way, scientists are constantly updating the boundaries of certainty and uncertainty, something that is not feasible in a single case.³ Judges are frequently expected to quickly come up to speed on continually evolving topics. As one of the judges said, “You’ve got to decide [what the jury’s going to hear] in a fairly short time period.” Another judge said that, while looking at one case and hundreds of others, “it takes [an enormous amount] of time to get skilled in testability, error rate, peer review, general acceptance, and standard protocols to look at the competing version[s] brought by the parties.”

At the end of the day, science and law are two different systems, established to fulfill two different societal needs, leading to differences in processes and understanding. In science, as one judge said, “The whole notion is no one ever comes out and says this is the way it is; all inquiry is over.” But in law, inquiry does end; a decision is reached. This gears scientists, advocates, and judges to think and act differently, which can feel like a major obstacle when these fields come together. Put bluntly by one judge, “The way the scientists think about problems and evaluate problems and come to conclusions is very different from the way courts do it.”

Reconciling Uncertain Science in the Deterministic Courtroom

Given these challenges and differences in the systems of science and law, it was unsurprising that each judge has come to think differently about the uncertainty baked into their roles. One judge said that a gut reaction to the uncertainty of challenging scientific evidence is often discomfort, and many judges may try to avoid the crux of the scientific issues because “they won’t get the kind of certainty that they want.” But, this judge continued, judges should “accept the fact that you’re not always going to be feeling sure about it. . . . You still have to make the decision. The system will break down if you don’t make a decision and make it with reasonable speed.”

Other judges had a slightly different approach when dealing with inevitable uncertainty, even in the face of challenging cases. Some rely on the written rulebook. The rules have already defined the degree of certainty required. Criminal charges must be proven beyond a reasonable doubt, and civil claims must be proven to be more likely than not, given a preponderance of the evidence. Similarly, uncertain evidence still has to reach the threshold of admissibility, be reliable, and, as one judge described, be “based on the best we can do.”

Managing Uncertain and Challenging Science in the Real World

How do these judges do the best that they can do? Each judge, of course, had different ways of managing processes, all depending on the particulars of the case at hand and timing of motions and proceedings.

The most common resources that they relied on (mentioned by four of the seven judges) was the Reference Manual on Scientific Evidence.⁴ This behemoth manual is written by independent scientists and legal experts, offering judges and readers in-depth, neutral primers on the science behind topics such as complex statistical models, DNA evidence, and eyewitness testimony. The fourth edition of the manual is in development, adding new chapters on topics including artificial intelligence and computer science. (Full disclosure: Many of the judges interviewed are on the advisory committee for this update.) One of the judges said that the manual offered a “crash course” in science that would allow judges to “be better informed and better able to make a decision about whether an expert really is reliable or not.” The same judge added, “I think Daubert decisions over the next decade are going to be much better if the judges who are making them are turning to the manual, getting educated on the basics of a particular area of science.” Another judge recommended that, even if a judge doesn’t read the manual cover to cover, “they should have a sense of or a good grasp of the first two or three chapters and read the table of contents so that they can manage these very, very important issues that are presented in cases.” But the manual does not come without its own challenges. Two judges suggested that many other judges and most law clerks do not know about the manual, and, even if they do, as two judges said, time spent learning is often limited by other competing factors.

So, with all the tasks judges are juggling, what other tools are available to them?

One judge favored using Daubert hearings to understand the science, while another judge specifically disfavored them. Instead, that judge preferred to learn about applicable scientific material through legal filings and other FJC resources available online.⁵ Another judge mentioned a preference for technical advisors to help them understand complex scientific issues. Three judges talked about holding science tutorial sessions, where they sit down with both sides and spend a half day going through the basic science. In these tutorials, they said, they focused only on understanding the science, leaving the case dispute at the door. One judge described the process:

I brought in the experts for a morning, and I told them or the lead expert for each side, “Spend a morning telling me what you agree on. What is the basis of the science here that everyone can accept and tell me an article or two?” I found that very effective.

Passing on the Baton of Knowledge and Experience

At the end of our conversation, I asked the judges what they would want a new judge to know about evaluating scientific evidence in the courtroom. They thoughtfully offered their advice based on their decades of experience. One judge mentioned that they recommend spending time struggling with the science, even when it felt challenging: “I worked really hard to try to understand, even though I realized I was the dumbest person in the room.” While I sincerely doubt that contention, I agree that learning the basics of complex topics can be overwhelming for a newcomer. I know that I have felt like the dumbest person in the room over my year at the FJC on more than one occasion. As echoed by another judge, “This is not stuff you can do on the fly, and, in the proper case, there are these technical advisors that you may want to consult.” During these sessions, one judge advised, “Make sure that the experts and the lawyers talk to you in English, not in scientese.” Another judge recommended that law clerks can be helpful in the learning process as well.

Nearly every judge discussed the utmost importance of a judge’s responsibility in understanding both the basic scientific principles of scientific evidence and how courtroom rules may apply, noting, “There’s a fair number of court decisions that say, well, if you got a close question and there’s plausible arguments, it’s a jury issue, and that in my view gives up the judge’s gatekeeping role. You just don’t throw close questions to the jury.” One judge simply stated, “Understand the rule of admissibility — 702.” They continued to emphasize it is a judge’s duty to screen the science in the courtroom and avoid junk science. To do so, judges need to “understand the role and understand our job in terms of making that initial determination by a preponderance that the factors within the rule are met.” Above all, judges recommended that others remember their role. As one judge said, “When you have to decide whether a scientist is [and] whether the opinions of a scientist are reliable enough to be presented to the jury, you’re the gatekeeper.”

Science and Law — One and the Same?

In the 30 years since the role of “gatekeeper” was officially given to federal judges, science and technology have only become more prevalent in the courts. Although judges deal with complex scientific evidence on a regular basis, I noticed that most judges still had some reservations about dealing with it. Many judges said they were far from scientific experts and that this field was far from their own legal home. Yet I see something different. Most of the scientists, lawyers, and judges I know share many characteristics that drove them to pursue these equally daunting professions. Like my scientist friends and colleagues, every judge I met this year is a highly intelligent, naturally curious, and thoughtful person committed to their role.

As the ever-changing scientific landscape has continued to evolve, the judiciary’s approach to scientific evidence will likely follow. Some may be disheartened to know it probably won’t get any easier as judges will continue to face new, more complex scientific evidence. After my year at the FJC peeking into the law, though, I firmly believe that the federal judges peeking back into my world of science are well-equipped to tackle this great task, now and well into the future.

---

REBEKAH L. PETROFF is an award-winning PhD scientist with over a dozen peer-reviewed publications. She was the 2023–24 American Association for the Advancement of Science science and technology policy fellow placed at the Federal Judicial Center.

CHELSEA S. QUEEN is a recent PhD graduate in legal psychology from The University of Texas at El Paso. She was a student research assistant at the Federal Judicial Center while working on this project.

---

 

1. Views expressed here are those of the authors or judges solely. Views do not necessarily reflect the views of the Federal Judicial Center, American Association for the Advancement of Science, or any other associated institution.
2. All judges who were interviewed for this story read and reviewed it before publication. With permission from the judges, the participants were: David G. Campbell, senior judge, U.S. District Court for the District of Arizona; Nancy D. Freudenthal, senior judge, U.S. District Court for the District of Wyoming; Nancy Gertner, retired judge, U.S. District Court for the District of Massachusetts; Paul W. Grimm, retired judge, U.S. District Court for the District of Maryland; Jed Rakoff, senior judge, U.S. District Court for the Southern District of New York; Patti B. Saris, judge, U.S. District Court for the District of Massachusetts; and Thomas D. Schroeder, judge, U.S District Court for the Middle District of North Carolina.
3. As famously put by Judge Richard Posner, “[T]he courtroom is not the place for scientific guesswork, even of the inspired sort. Law lags science; it does not lead it.” Rosen v. Ciba-Geigy Corp., 78 F.3d 316, 319 (7th Cir. 1996).
4. Federal Judicial Center, Reference Manual on Scientific Evidence 129 (3d ed. 2011).
5. See, e.g., Federal Judicial Center, Manuals, Monographs & Guides, https://www.fjc.gov/research/manuals-monographs-guides (last visited Apr. 27, 2025).