Is SEO the best way to improve eyewitness recognition? Putting people who DO NOT look the same in police lines can improve accuracy by 10%, a remarkable study reveals
- U.S. and UK investigators recruited 19,732 volunteers to play the role of witnesses
- Each participant was shown a video of a faked crime, followed by a sketch
- The team tried to line up with people who were very similar to the suspect
- Less similar sequences were found to correctly detect perp.
- But they did not magnify wrong things in lines with a suspicion of just innocence
The accuracy of identity marches could be improved by 10 percent by including people who do not all look the same as the suspect in police ranks, a study found.
U.S. and UK investigators hired more than 19,000 participants to play the role of witnesses in a line-up after witnessing a recorded crime.
They found that people were more likely to identify the perpetrator correctly if they were photographed with dissimilar faces, rather than resembling the suspect.
However, they were not at the same time more likely to mislead an innocent suspect in a line of sight where the real perpetrator was not present.
The approach differs from the usual method, in that the ‘fillers’ that make up the rest of a line are selected to be broadly similar to the descriptions of the suspects or are witnesses.
The accuracy of identity marches could be improved by 10 percent by including people who don’t all look the same in police lineups, a study has revealed
‘In practice, police tend to err on the side of building face-like nets for their line,’ said the paper’s author and psychologist John Wixted of the University of San California Diego.
‘What our study shows is that, against intuition, it is better to choose fillers that are unlike face.’
‘By doing so he protects the innocent to the same extent and helps witnesses to identify the culprit correctly more often. ‘
In their study, Dr. Wixted and his colleagues played a fake video of a crime showing a white male stealing an office laptop, to 19,732 volunteers.
After viewing the photos, each participant was shown six photos, one of which was a ‘suspect’ – and either the perpetrator or the innocent.
The other five images of police line fillers, each similar, were to varying degrees (as previously estimated by a different set of volunteers), to the suspect.
The team found that taking ‘fill-in’ images of people unlike the suspect improved the ability of witnesses to select the perpetrator when committing the crime. in line.
At the same time, however, this did not increase the likelihood of the wrong witness identifying an innocent suspect when the perpetrator was not part of the identity march – which the team called ‘false alarms’.
‘Eye misinformation has contributed to many false convictions, which are compounded by DNA evidence,’ said the paper’s author and psychologist Melissa Colloff from the University of Birmingham.
‘If a perpetrator is not identified when involved, perpetrators may be acquitted of further offenses.’
‘While many useful reforms have been introduced to protect the innocent, sometimes these provide protection for the guilty.’
The findings, she said, show that one police line can be made more effective for everyone, ‘increasing the likelihood of a perpetrator being identified, without increasing the an innocent suspicion is likely to be aroused. ‘
The full results of the study were published in the journal Proceedings of the National Academy of Sciences.