Science and the challenges of evidence-based forensics

Detectives and scientists have a lot in common. Both collect data, develop hypotheses and test their assumptions. And both pursue evidence in the hope of a definitive finding.

Science was key to the exploits of fictional detective Sherlock Holmes, who developed tests to detect poisons and other clues. Another fictional British detective, Dr. John Evelyn Thorndyke, also applied science, using his training as a pathologist to solve crimes, including using X-rays to reveal a murder victim’s corpse stashed in an Egyptian s­arcophagus. Many more science-minded fictional gumshoes followed, including introverted entomologist Gil Grissom in CSI: Crime Scene Investigation and FBI behavioral analyst Jason Gideon in Criminal Minds.

We continue to be fascinated by the science of solving crimes — true crime is the most popular genre for top-ranked podcasts. But failures of forensic science to accurately ID perpetrators and exonerate the innocent have become as fascinating as the marvels of scientific sleuthing itself. No doubt Sherlock would argue that science is still the solution. Would he be wrong?

In this issue, we examine where forensics falls short and how scientists are working to fix those failures. It’s a challenge, because the work requires both improving the science itself and bettering public understanding of science for a society that includes judges, prosecution and defense lawyers, and jurors.

Despite the shakiness of some of the so-called science, solid, real-deal science can still help make the criminal justice system more accurate and fair. As freelance writer Amber Dance writes, science can’t always provide 100 percent certainty, but it can improve the odds that evidence is tested and evaluated properly (Page 22).

In the courtroom, science has led to huge leaps forward. The introduction of DNA evidence in the 1990s overturned many wrongful convictions. It became the gold standard in suspect identification, identifying notorious criminals like the Golden State Killer (SN: 12/17/18). But it’s not infallible; using very small samples of DNA, a mere few cells, makes it harder to get a positive identification. Efforts are underway to add likelihood ratios to this type of evidence, but the technique isn’t widely used.

Researchers are also testing ways to improve eyewitness identification. Such approaches include being sure that the officer running a photo lineup doesn’t know who the suspect is and inadvertently influence the witness, or ensuring that one person in a lineup doesn’t look dramatically different than the others.

I was fascinated by one study that suggested a way to reduce witness bias when showing lineups. Showing pairs of photos and asking which of each pair looked most like the suspect, rather than showing multiple headshots at once, meant witnesses were less likely to assume that one person in the lineup had to be the perpetrator. But calling a suspect “most likely” may be meaningless in a court where jurors crave certainty, Dance notes. One of the biggest challenges in using science to make the criminal justice system more just is getting the system to accept the caveats and nuance that are inherent to the science itself. Even when it comes to bits of DNA.

Nancy Shute is editor in chief of Science News Media Group. Previously, she was an editor at NPR and US News & World Report, and a contributor to National Geographic and Scientific American. She is a past president of the National Association of Science Writers.