Most of us won’t be on the jury of a forensics-based trial or be accused of a crime, but for those that are, and for those who regularly watch true crime documentaries, it should be known that many of the forensic techniques we grew up hearing about are not actually accurate.
Earlier this year, ProPublica published an article that included a list of things to look out for when determining whether forensic analysis is sound science or junk science. The list includes:
- It has limited or no scientific evidence or research supporting it.
- It is presented as absolutely certain or conclusive, with no mention of error rates.
- It relies on subjective criteria or interpretation.
- It oversimplifies a complex science.
- It takes just a few days to become an “expert.”
With this information in mind, let’s take a look at some forensic techniques that have been featured in movies, television shows, and true crime documentaries but that are actually not sound science at all.
Bite Mark Analysis
Serial killer Ted Bundy was convicted in part on bite mark analysis, and the technique was used in the following decades. In recent years, however, the technique has been widely criticized as unreliable, though it is still used in courtrooms.
Bite mark analysis falls under the broad terms of forensics known as pattern matching, where an “expert” looks at evidence collected at a crime scene and evidence taken from a suspect, compares the two, and makes a subjective determination as to whether there’s a match.
At least 35 people have been wrongfully convicted based on bite mark analysis, The Intercept reported last year, including Steven Mark Chaney, who was convicted in 1987 for a murder he didn’t commit. Chaney had nine alibi witnesses and there was no actual evidence linking him to the murders of John and Sally Sweek, but two forensic dentists testified that an alleged bite mark found on John’s arm matched Chaney’s teeth.
Journalist Radley Balko has extensively documented the problems with bite mark analysis, questioning how accurate it could possibly be.
“For example, we don’t know what percentage of the human population’s teeth are capable of leaving a certain pattern of marks on human skin,” Balko wrote last month. “We don’t even know if such a calculation is even possible. And if it were possible, we don’t know that human skin is capable of recording and preserving bite marks in a way that’s useful (the studies that have examined this suggest it isn’t).”
“The characteristics of a bite mark could be affected by the angle of the biter’s teeth, whether the victim was pulling away at the time of the bite, the elasticity of the victim’s skin, the rate at which the victim’s body heals, and countless other factors,” he continued.
Many cases use dogs who are trained to detect the scent of dead bodies to determine whether a crime has occurred and to connect suspects to that crime. But their accuracy has been called into question in recent years. In some cases, dogs have found scents that couldn’t possibly have existed.
Take the missing girl Madeleine McCann. Her parents were labeled suspects in her disappearance four months after their daughter vanished, based largely on a cadaver dog indicating the scent of death on one of McCann’s baby toys and on her mother’s clothes.
The problem, however, is that it is unclear whether McCann was even killed, but speculating that she was killed the night she died, the cadaver dogs would have had to pick up minute traces of a scent four months after the murder. As Slate reported in 2007, it is unclear how long that scent could linger, with a former Scotland Yard dog handler suggesting the scent wouldn’t last longer than a month.
And to be clear, the dogs didn’t actually find any human remains.
Some attorneys have convinced a Wisconsin judge that some of these death-sniffing dogs were accurate just 22-38% of the time. Even prosecutors who denied those attorneys’ claims said the dogs had a success rate of just 60-69%, which is hardly accurate enough to convict someone.
A cadaver dog was also used as evidence against Scott Peterson, who was convicted in 2004 of killing his wife, though many question whether he could have actually been the murderer.
In the Peterson case, a trailing dog (similar to a cadaver dog but trained to follow scents in the air) allegedly followed his wife’s scent from a parking lot at the Berkeley Marina, where Peterson kept a small fishing boat, to a nearby pier. The dog allegedly followed this scent four days after Peterson’s wife, Laci, was reported missing. The dog was given Laci’s sunglasses and hairbrush, both of which could have had cross-contamination from Scott Peterson, who had maintained his innocence and told police he went fishing on the day Laci went missing.
Contact or Trace DNA
When we hear the letters DNA, we immediately think science and indisputable fact, but there is a type of DNA evidence that is not reliable and has been used to wrongly imprison many – contact, or trace DNA.
We leave behind trace amounts of DNA on everything we touch. This is called contact DNA. It may only be a few skin cells, but it is there. That contact DNA can then be transferred to another surface without us touching it. For example, if you shake someone’s hand, they’ll have traces of your DNA left behind, and then if they touch something else, they could leave those traces of your DNA as well as their own DNA on that object.
There isn’t as much DNA left behind as there would be in blood, saliva, or semen, yet “experts” have used contact DNA to put people in prison. One example is the case of Mayer Herskovic, who was convicted in 2017 for allegedly taking part in a multi-person attack against a young, gay black man in Brooklyn in 2013. Experts testified that trace amounts of Herskovic’s DNA was found on one of the victim’s shoes. To come to this conclusion, the crime lab had to take a tiny sample of the DNA and replicate it to get a useable sample. This involved copying the DNA and then copying the copies thousands of times, and then running that data through the computer to get statistics claiming it was highly likely that the tiny amount of DNA on the shoe belonged to Herskovic.
The problem is that the technique, which is no longer used by the crime lab, didn’t find the victim’s DNA on his own shoe, but also found the DNA of someone else. And because the community where the crime occurred was largely homogenous, the DNA results could have matched someone other than Herskovic. Herskovic’s conviction was overturned in 2018.
911 Call Analysis
Police and prosecutors sometimes train in 911 call analysis, which claims to detect a murderer in a phone call by listening to their speech patterns, tone, word choice, and other factors when calling authorities, ProPublica reported.
The technique was designed by Tracy Harpster, a retired deputy police chief with limited homicide investigation experience and little scientific background. He used a small study to create the technique for his master’s thesis, and it was soon accepted by law enforcement nationwide. The technique includes a checklist of how many “errors” a person commits during a 911 call, which Harpster said gives away their true involvement. Asking “huh?” in response to a question from a dispatcher is one indication of guilt, as is saying “please” by itself. If a person doesn’t seem urgent enough, that is also an indication of guilt, Harpster concluded.
Studies from the FBI in 2020 and 2022, however, warned against the technique, saying its findings could not be corroborated and that it may simply create bias against the accused.
The technique was used to convict Jessica Logan of killing her baby in 2019. Logan’s baby had died during the night and she called 911 frantic. She was crying and screaming and could barely speak to the 911 operator. Detective Eric Matthews determined her agony was fake and that she had actually killed her own baby.
Matthews had taken a two-day course on 911 call analysis, and used what he learned to get Logan convicted. Many are questioning her conviction.
We’ve all seen images of experts comparing the lines of one bullet, referred to as rifling marks, to the lines on another bullet. This is firearm analysis, the belief that every gun in the world contains minute imperfections that distinguish it from every other gun on the planet, just like DNA.
But there’s no evidence this is true. It is unknown whether marks in the gun’s barrel leave unique marks on bullets that no other gun can recreate, Radly Balko wrote last month. He also raised other questions regarding firearm analysis.
“We don’t know if a gun’s firing pin and ejection mechanism leave unique marks on a casing,” he wrote. “We don’t know the frequency with which specific marks may appear among the entire population of fired bullets. We also don’t know how to account for the likelihood that a gun will leave different sorts of marks on bullets over the course of its life as a functioning gun as the ridges and grooves inside the barrel wear down — an issue that would presumably affect older cold cases.”
Yet firearm analysis is used to convict thousands, even though it is a subjective technique.
Bloodstain Pattern Analysis
Everybody’s favorite TV serial killer, “Dexter,” was a blood spatter analyst by day, yet his “scientific” technique isn’t all that scientific. Experts in the field analyze blood drops, spatters, and trails at crime scenes to reconstruct what happened, yet ProPublica reported that it has never been definitively proven or quantified.
In 2009, the National Academy of Sciences published a landmark report casting doubt on bloodstain analysis and other types of forensics, including bitemark analysis and other so-called pattern matching forensics. The report stated that “the uncertainties associated with bloodstain-pattern analysis are enormous,” and that experts’ conclusions were subjective, rather than scientific.
Faulty bloodstain analysis was used to convict Joe Bryan in the 1980s, with an analyst that had been trained on the technique just four months before Mickey Bryan was found dead. That training was just 40 hours long. Nonetheless, the “expert” testified that small reddish-brown flecks found on a flashlight in the trunk of Joe’s car were blood consistent with Mickey’s blood type (and half the population). The expert went on to make claims that had no basis in bloodstain analysis, alleging the killer must have cleaned up in the bathroom before leaving the house, even though no blood was found in the bathroom.
Fingerprint/Shoeprint/Tire Tread Analysis
Connecting a suspect to a crime by using a fingerprint is one of the oldest investigation techniques in the book, since we’re constantly told that no two fingerprints are identical. Investigators, however, rarely find clear, full fingerprints at a crime scene. Instead, they usually find partial prints, which can be smudged or have been dragged along a surface.
Here is when law enforcement turn to fingerprint experts, who often make assumptions about the prints found at the crime scene.
Dustin Phillips, a defense attorney in Oklahoma, pointed out the issues with this kind of analysis, noting that fingerprint experts can’t even agree on how many points of comparison must be found in order for two prints to match.
“Furthermore, even if full, complete fingerprint was lifted from a crime scene, even if the print could be proven beyond all doubt to belong to a suspect, the print itself does not mean the suspect had anything to do with the crime. It only proves that he or she was in that location at some point in time, possibly days or even weeks before the crime occurred,” Phillips wrote.
The same can be said for shoeprints and tire treads. It cannot be determined when the shoeprints or tire tracks were made, and therefore experts can’t definitively say a suspect left those particular tracks during the crime.
Microscopic Hair/Fiber Analysis
In 2015, the FBI and Justice Department acknowledged that in more than 95% of trials they reviewed, agents used the flawed method of hair analysis.
Hair and fiber analysis involve experts comparing one hair or fiber to another, and claiming there is a match or not. Ed Pilkington at The Guardian reported that “consensus by real experts is more straightforward than ever: there is nothing that can credibly be said, by FBI-approved analysts or anyone else, about [ ] the frequency with which particular characteristics of hair are distributed in the human population.”
Fiber analysis is similar, with investigators unable to determine that a fiber found on a victim or at a crime scene could be definitively linked to a suspect, particularly given the rise of synthetic fibers, which can be found in many homes.
Handwriting analysis is used to determine if two documents were written by the same person. Like many of the entries on this list, an analyst’s report is entirely subjective, and different experts can come to different conclusions.
“The technique of comparing known writings with questioned documents appears to be entirely subjective and entirely lacking in controlling standards,” a federal court ruled in United States v. Saelee (2001). The court in that case noted that in one test, “the true positive accuracy rate of laypersons was the same as that of handwriting examiners; both groups were correct 52 percent of the time.” This means that a person without any training in handwriting analysis was able to correctly identify matching documents as often as trained experts.
BONUS: Legitimate Forensic Techniques Used Wrongly
Video footage can be extremely helpful in a trial, showing jurors exactly what a suspect did, but in 2008, video evidence was used to send the wrong man to prison.
Back then a white, slender-built male was robbing convenience stores in Killeen and Copperas Cove, Texas. In video of one incident, the suspect can be seen leaving the store and the clerk quickly follows after to lock the door. One can see in the video that the clerk and the perpetrator were about the same height. The clerk was 5-foot-six, meaning the perpetrator was likely on the shorter side.
Someone called a tip line to accuse George Powell of the crime. The problem was that Powell is more than 6 feet tall. But at trial, prosecutors called in a man with crime reconstruction experience to claim that the suspect in the video was Powell and that he was just slouching. The man had never done any kind of height estimation prior to this case, and wasn’t even trained on the program he used to claim Powell was the man in the video. Powell was convicted and is still fighting for his exoneration.
Location data can be used typically to give a range of where someone is at a given time. It’s created when a person’s cell phone “pings” off a nearby cellphone tower. If a person is traveling between towers, investigators can map their general direction.
In the case of Quinton Tellis, who was charged with murdering 19-year-old Jessica Chambers by burning her alive in 2014, location data was used incorrectly. An intelligence analyst testifying for the prosecution found one of Chambers’ cell phone pings to be off by a certain distance, so he then shifted all of the data for Chambers and Tellis by the same distance. He claimed this proved that Chambers’ and Tellis’ phones were together leading up to her death.
But another expert looked at this evidence and found that the last phone call Chambers made pinged off a cellphone tower that didn’t cover Tellis’ house, meaning she couldn’t have been there. Tellis was tried twice for Chambers’ murder due to a mistrial, and the charges were eventually dropped.