Witness for the prosecution? The effect of confessions on eyewitness testimony

January 28, 2009

What is it with false confession? It seems crazy on the face of it, to take the blame for a crime you didn't commit. Yet experts have found that while some innocent confessors are mentally disturbed attention seekers, or taking the fall for someone else, most innocents who confess do so under stress, as an act of surrender, during a high stakes, high-pressure police interrogation. According to one estimate, fully a quarter of convictions later overturned by DNA evidence involved a false confession.

Confessions are powerful evidence in court. But in our legal system, confessions are supposed to be just one more piece of evidence, weighed along with fingerprints and alibis and eyewitness testimony. Are judges and juries capable of weighing evidence fairly when they're faced with someone proclaiming: I did it? Or is a confession so potent that it trumps everything else?

Legal experts are obviously very interested in this question, and psychologists have been working to piece together an answer. In a recent study, psychologists Lisa Hasel of Iowa State University and Saul Kassin of John Jay College wanted to see if eyewitnesses would be swayed enough by a confession to change their mind about their memory of the crime. In these experiments, the researchers staged a crime scene to explore the mind of the eyewitness and the integrity of eyewitness testimony.

A small group of students were sitting in a university laboratory, waiting for the experiments to begin. At some point, a person walked in, picked up a laptop from the desk, and walked out of the room. A few minutes later, the research assistant entered the room and announced, with obvious distress, that her laptop was missing. The group of students were the eyewitnesses and were asked to help solve the crime. The students were first asked to identify the thief from a line-up (however, unbeknownst to them, the actual thief was not in the lineup) and rate the confidence of their answers. The students returned two days later, to continue helping with the investigation. When they returned, they were told either that all of the suspects denied involvement or that a specific suspect confessed to the crime. The students were then to reconsider their original identification and rate how confident they were.

The results, reported in the January issue of the journal Psychological Science, a journal of the Association for Psychological Science, are a bit unsettling. None of the men in the lineup was in fact the real thief, so the only reliable testimony came from those who fingered no one. That is, the vast majority of study volunteers identified an innocent man as the criminal, and many did so with confidence. That's disturbing in itself, but it gets worse. While few were persuaded by claims of innocence - that happens all the time - a disturbing number changed their mind when a suspect confessed. An astonishing 60 percent who had fingered one suspect flip-flopped when a different man confessed. Even those who had been very sure of their original identification experienced a steep drop in confidence. When asked to explain their change of heart, most said they had been mistaken earlier, that their memories had fooled them.

How about the rare volunteers who (correctly) refused to finger any one of the original six in the lineup? These witnesses would seem to be especially cautious about falsely accusing someone of a crime. Yet fully half of these eyewitnesses also changed their minds when told that a specific suspect had confessed.

These findings have very serious implications for our legal system and suggest that investigators need to be aware of how confessions may influence eyewitness testimony. The authors conclude, "Once informed of a confession, an eyewitness is forever tainted."

Source: Association for Psychological Science