For more than a century, social scientists have found in studies that human memory is unreliable during high-stress moments like crimes and violent interactions. Peer-reviewed clinical studies in the last three decades have confirmed that eyewitness identifications are often unreliable and that the format of traditional police lineups has contributed to wrongful convictions by creating an atmosphere ripe for misidentification. More than 75 percent of the 206 people exonerated by DNA testing in the United States were convicted – at least in part – based on eyewitness misidentification.
But last year, a non-scientific report from several Illinois police departments threw confusion into pending reforms across the country by claiming that accepted reform models didn’t work. A new report this month in the leading journal Law and Human Behavior has begun to clear up this fog of confusion.
The new report was written by a “blue-ribbon” panel of social scientists, including Nobel Laureate Daniel Kahneman of Princeton and Harvard professor and author Daniel Schachter. The panel finds that the Illinois report was fatally flawed and that its results should be viewed with extreme caution. The Illinois report’s format “has devastating consequences for assessing the real-world implications of this particular study,” the panel found. “The design guaranteed that most outcomes would be difficult or impossible to interpret. The only way to sort this out is by conducting further studies…”
The reforms in question are “sequential double-blind” lineups – in which suspects and “fillers” are presented to an eyewitness one at a time, instead of all at once. The procedure is called “double blind” because neither the police official nor the eyewitness knows which lineup member is the actual suspect. Research has shown that in traditional lineups, witnesses tend to compare members and choose the one bearing the closest resemblance to the perpetrator. And if an official knows which member is the suspect, they can sometimes give intentional or unintentional clues to the witness.
The Illinois report’s fatal flaw was the combination of two variables, said James Doyle, director of the John Jay College of Criminal Justice's Center for Modern Forensic Practice, which coordinated the panel review.
Comparing group and individual lineups, while at the same time using some administrators who knew suspects and some who didn't, was like comparing "apples to dolphins," Doyle told the Chicago Tribune. "Just putting the two of them together doesn't make it a scientific review. You've changed two variables at once. You can't do that."
When the Illinois report was released during 2006 legislative sessions around the country, opponents of identification reform used the report’s faulty conclusions to derail bills in several states. Ultimately, the Illinois report confused many lawmakers and led them to table important eyewitness reforms that were making progress around the country.
Some states – including
– saw through the distortion and confusion caused by the Illinois report and nonetheless passed important eyewitness reform bills in recent months. And just last month,
lawmakers approved sweeping reforms that will require all police departments across the state to begin using “sequential double-blind” lineups.
This month’s scientific review, invalidating the Illinois report and calling for more reliable field research, should pave the way for more reforms to move forward in several states over the next few months.
Law & Human Behavior:
Studying Eyewitness Identifications In the Field
The new report’s publication has sparked a vibrant discussion among scientists.
to the new report on the Law & Human Behavior website.
Best police lineup format not yet ID’d
A field study in Hennepin County, Minnesota
– which was conducted with scientific rigor and did not have the limitations of the Illinois report – found that the “sequential double-blind” reforms lead to more accurate eyewitness identifications.
Read about more than
150 wrongful convictions
in which eyewitness identification played a part.