Wrongful Convictions, Cognitive Bias, and Forensic Science

In May 30, 1997, a man shot at Sergeant Gregory Gallagher with the sergeant’s gun and fired at an area resident, Benjamin Pitre. The man then forced himself into the house of a local resident, Bonnie Lacy, and, when fleeing the residence, he left behind a sweatshirt, a mug he had drunk water from, and the gun. After a few days, Sergeant Gallagher identified the suspect in a photo lineup, but the other witnesses (Benjamin Pitre and Bonnie Lacy) did not. When the suspect, Stephen Cowan, was subsequently put into a live lineup, Sergeant Gallager and Mr. Pitre identified him. Mrs. Lacy did not. The thumbprint the shooter left on the water mug generated no match in the computerized national fingerprint database in which Cowan’s fingerprint was stored; however, during the trial, two Boston, Massachusetts, Police Department fingerprint analysts testified that Cowan’s thumbprint was a match. This testimony, along with the eyewitness identifications, led to the jury decision that Cowan was guilty of all charges on June 30, 1998. Cowan was sentenced to 35 to 50 years in prison, and the decision was upheld by the Appeals Court of Massachusetts on October 12, 2001.1

Cowan’s request to have the mug, baseball cap, and sweatshirt tested for DNA was approved on May 22, 2003, by a Suffolk County Superior Court judge, and the tests proved that the DNA from all three items came from one person—and that the person was not Cowan. Two days into a new 2004 trial, the Suffolk County district attorney stated that the original thumbprint match that contributed to Cowan’s conviction was a mistake. All charges were subsequently dismissed; the two original fingerprint examiners were placed on administrative leave; and the Boston Police Department Latent Fingerprint Unit was temporarily shut down.2