Sydney researchers including from the Not Guilty project have confirmed direct eye contact may increase the perceived familiarity of a face and therefore the chances of a wrongful conviction.
People tend to overestimate how well they can recognise faces.
Direct eye contact may wrongly increase the perceived familiarity of a face, new research from the University of Sydney suggests.
The research provides the first evidence of a selection bias towards faces that make direct eye contact in a line-up task during eyewitness identification.
The paper – "Who is the usual suspect? Evidence of a selection bias toward faces that make direct eye contact in a lineup task" – was written by Dr Jessica Taubert, who did the study through the Forensic Psychology Lab; Dr Celine van Golde, the founder/director of Not Guilty: The Sydney Exoneration Project; and the McCaughey Chair of Psychology, Professor Frans Verstraten.
The study was published recently in the journal i-Perception.
In a series of experiments, the researchers from the University of Sydney’s School of Psychology showed that a person looked more familiar to participants when they were looking directly at them compared to when someone looked at them from a different angle.
“In line-up recognition tasks, the face looking directly at you is more likely to look familiar than faces looking away from you," said lead author Dr Taubert.
"This leads to more misidentification errors for direct gaze faces."
The human brain is highly sensitive to direct eye contact in general and the experiments suggest that people are biased towards selecting direct gaze faces when gaze direction differs within the set-up.
The bias disappears when the person to be identified closes their eyes, further indicating that eye gaze is indeed the pivotal factor.
Line-ups are often used in eyewitness identification, an important pillar of the judicial system. Suspects stand in one line while the person identifying them stands in front of them, meaning the gaze direction for every suspect differs slightly.
Eyewitness identification is often used because perception is the most reliable means of confirming a person’s identity when automated technology, such as fingerprints, is not available.
But results are prone to error because they rely on the abilities of a single person to identify someone by looking at them – a task that is much more difficult than we think.
Dr Celine van Golde said because people recognised "friends, family and familiar people so quickly and easily", confidence was inflated when faces were unfamiliar.
“People tend to overestimate how well they can recognise faces," Dr van Golde said.
This overconfidence was one reason why eye gaze had not been explored as a factor before, even though it is well-known to influence vision and memory.
The bias could be easily overcome by small changes to the line-up setting, which could lead to a lower misidentification rate in the future.
Listen to Dr van Golde's interview about false memory on the University of Sydney podcast series Open for Discussion.
Think back to a week ago, a month ago, a year ago. How much can you remember? Do you remember exactly what you did and everyone you saw? Memory is a tricky thing, so what happens when it’s all that stands between a person going to jail, or walking free?