Showdown: DNA evidence vs. cognitive bias

Back in the 1980s, southern Alameda County in the East Bay was the hellmouth for serial murder. As a newspaper reporter covering the crime beat, I was reporting on at least three separate fiends prowling the suburbs and picking off young teenage girls at whim.

It was harder to stop them back then. Forensic DNA was still in its infancy. The historic evidentiary hearings in Oakland, California on the admissibility of DNA typing, with full-scale scientific battles tying up courtrooms for months on end, were still a few years away.

Tina Faelz and her mother Shirley

Fourteen-year-old Tina Faelz was one of the victims. In 1984, she was found dead with 44 stab wounds. She had taken a shortcut through a drainage culvert while walking home from school.

(As a side note, Tina had walked home that day because a group of girls was planning to beat her up if she rode the bus. Bullies tyrannized Foothill High School in suburban Pleasanton; on the same day as Tina’s murder, an alpha-male bully threw a football player into a dumpster and locked the lid.)

Detectives had no shortage of suspects. There was the mother’s violent boyfriend. There was the aforementioned school bully, whom someone had spotted near the crime scene. There was a man who was arrested shortly after Tina’s death for a similar assault in which the girl managed to escape.

What they lacked was hard evidence.

The case went cold for decades. It was finally cracked just a few years ago, thanks to the intersection of DNA science and a cop’s pregnancy. Detective Dana Savage couldn’t be on the streets due to her pregnancy, so she decided to take a gander at the vexing cold case.

Detective Savage was fairly certain that the culprit was one of two serial killers who’d been active in the region at the time; she just didn’t know which one. Based on the vigor of the attack, she figured the killer must have shed some blood, so all she needed was something to test for DNA. She struck gold with the victim’s purse, which had been found lodged in a nearby tree.

But when Savage got the call from the crime lab, she was in for a surprise. The culprit was not one of the serial killers. Nor was it any of the original suspects.

It was the 16-year-old classmate who’d been thrown into the school dumpster earlier in the day.

After killing Tina, Steven Carlson had dropped out of school and spent the next 30 years abusing meth and bouncing in and out of custody. When police came to talk to him, he started retching violently. He was tried and convicted, and is now serving a 16–to-life sentence.

It’s unfortunate that it took so long to catch the killer. But on the bright side, the Pleasanton police did things right: They kept their minds open and never fixated on the wrong person. That would have been far worse.

Barking up wrong trees

In other cases during that violent era, police sometimes got it tragically wrong. For example, when 8-year-old girl Cannie Bullock was raped and murdered in her home in nearby San Pablo, Detective Mark Harrison fixated relentlessly on William Flores, the sexually creepy guy next door, literally driving him to his grave. (If every creepy guy was a murderer there wouldn’t be many women left on the planet, or even many male cops if you believe the dismal statistics in the must-watch Netflix series Unbelievable.) Even after Flores self-immolated, the detective wouldn’t let him rest in peace. Once DNA technology became available, Harrison got a court order to dig up Flores’s body, certain the test results would clear the long-dormant case.

He was dead wrong. The DNA didn’t match that found on the little girl’s body.

(That case went cold for many years. Finally, DNA from a man convicted of sexual assault in Colorado was routinely entered into a database, which spit out a match. The killer, Joseph Cordova, was never a suspect in the girl’s killing, although he lived and worked in the area and had used drugs with the girl’s mother. He is now parked on California’s death row.)

But here’s the really bad news: Even with modern DNA technology’s miraculous crime-solving capabilities, fixations like Detective Harrison’s still lead police astray with some regularity. In particular, forensic science is no match for a priori stereotypes about the bad guys.

A case in point: The murder of elderly Leola Shreves in Yuba City, California.

The attack was frenzied. As detailed by San Francisco Chronicle reporter Matthias Gafni, the TV set was smashed and a door was ripped from its hinges. The 94-year-old victim had been tortured, strangled and beaten to a pulp. Her teeth were shattered, her jaw and back broken, and 17 of 24 ribs cracked. Her ears and scalp were nearly ripped from her skull.

Police quickly latched onto the next-door neighbor, a socially awkward video-game devotee. Michael Alexander aroused police suspicion in part due to his troubled past: He had been arrested at age 15 for threatening to kill a high school teacher and burn down the school after fighting with and choking another student.

Burdened with an intellectual disability, the 20-year-old was no match for the seasoned detectives who brought him in for questioning. When he denied ever being at his neighbor’s house, police lied to him, saying his fingerprints, shoe prints and DNA had all been found there. When he continued to profess his innocence, detectives suggested that maybe he had blacked out, and an alter ego named “Angry Mike” had committed the crime. Alexander’s naïve acceptance of the detectives’ ruses eventually led him to accede to their version of reality despite not having any recollection of it.

For anyone with expertise on false confessions, Alexander’s had all the classic hallmarks. It was replete with maybes and probabilities. The details did not match the evidence from the crime scene. And Alexander immediately recanted.

“Have you been looking for the real killer?” he later asked the detectives.

His question fell on deaf ears. He was arrested and charged with capital murder.

Unbeknownst to him at the time, there was indeed an abundance of real physical evidence – DNA, fingerprints and shoe prints. All of it excluded him and pointed to someone else.

Astonishingly, the identity of Shreve’s killer was in front of the detectives the entire time, but it took them six long years to realize it.

Armando Cuadras

On the night of the murder, a man named Armando Cuadras was found collapsed on the street just 300 yards away, drunk and badly injured. He was taken to the hospital by ambulance, but police failed to connect the two events. Cuadras, whose DNA was splattered all over the bloody crime scene, is now awaiting trial.

Mental blinders

Cognitive scientists have various names for the mental processes that cause people to prematurely focus on one solution to the exclusion of other possibilities. Tunnel vision. Myopia. Confirmation bias. In essence, the Yuba City police identified a suspect, based in part on their preconceived ideas about what a guilty person should look like, and in the process closed their minds to alternate possibilities.

Then, once all of the physical evidence came back and screamed out Alexander’s innocence, cognitive dissonance kicked in: It can be hard to abandon a firm belief even when confronted with irrefutable evidence that it is wrong. Cognitive dissonance was on florid display in the infamous case of the Central Park Five. As documented in the powerful Netflix series When They See Us, prosecutors still refuse to accept overwhelming evidence of the young men’s innocence. Such is the power of cognitive blinders. (My blog post on that astonishing case is HERE.)

Unfortunately, when police focus on the wrong person they not only destroy the suspect’s life, but also allow the real culprit to remain free, thereby endangering others in the community. There are myriad cases of very dangerous men who went on to rape and kill again after police investigators failed to diligently pursue all leads. (Again, let me plug the harrowing series Unbelievable.)

After almost two years in jail, Alexander was finally set free and the charges against him dismissed. But even with another suspect in custody and awaiting trial, police and prosecutors have stubbornly refused to concede that Alexander is innocent.

Which just goes to show, even the miracles of DNA typing are no match for minds that are rigidly shut.

* * * * *

FURTHER RESOURCES: The transcript of Michael Alexander's confession is available online, and is a good resource for teaching and learning about false confessions. Tina Faelz's killing is the subject of a true-crime book, Murder in Pleasanton, which includes back-story information not available elsewhere. If you are interested in diving deeper into the problem of cognitive biases in police investigations and how they can be avoided, a great resource is Criminal Investigative Failures, edited by D. Kim Rossmo. Two chapters I especially recommend are "Who Killed Stephanie Crowe," focusing on the appalling case that I've blogged about several times in which a 14-year-old boy was wrongfully arrested in his sister's murder, and "On the Horns of a Narrative," by my colleague David Stubbins and his brother, which focuses specifically on cognitive biases in criminal investigations.

A NOTE TO MY FAITHFUL SUBSCRIBERS: My apologies for the diminishing quantity of posts as of late. I'm working on a couple of larger writing projects. I also Tweet regularly on forensic psychology and criminology topics, so feel free to follow me on Twitter for more regular news and commentary.