Wednesday, 2 October 2013

Why Eyewitnesses are so unreliable


Image: sodwee.com

A Face in the Crowd
(click here to Tweet this post)

During the Wimbledon tennis tournament last year, a friend of my husband, Simon, was idly watching television. There were various court-side shots and in one split second, Simon knew he recognised one of the faces in the crowd. He rewound the footage and took a photograph and sure enough, he’d captured my husband as a tiny speck in the distance, surrounded by hundreds of faces.  We got the photograph a few months later in a Christmas card!
 
I’d say that was pretty remarkable, but some people are good at identifying faces. Others are not so good. In fact, according to police statistics, the average person on the street is fairly useless when it comes to being a witness, but research indicates that in court, eyewitness accounts provide the most persuasive evidence and are believed 80% of the time. An eyewitness statement to a jury is very compelling - I was there and I saw what happened is very believable when there’s no apparent vested interest other than to tell the truth.
 
Image: theblaze.com
Faulty Feedback

In 1975 a team of researchers broadcast a TV crime simulation which generated 2,145 callers. Only 14.7% were accurate in their identification of the perpetrator. In 2009, of 235 people exonerated through DNA evidence in UK, more than 75% of them had already been charged and sent to prison, largely due to an eyewitness. It’s a high fallibility rate.

Weapon Focus and the fear caused by violence are further reasons for not remembering faces. In a crime where a weapon is involved, our focus is automatically drawn to the area of most threat. For example, bank employees who have been face to face with robbers can describe the weapon in great detail, but rarely the person holding it. Witnesses are more likely to remember details if the situation is non-violent.

For those who do remember, certain details we recall tend to be more accurate than others. Gender, for example, is usually 100% accurate. Characteristics of speech or accent is next at 84%. In terms of accuracy, a person's build scores around 57%, but height only 44%; age - even less at 38%. We're noticably poor at telling how old someone is!


Image: LM Otero/AP
There’s also the ‘misinformation effect’ where the words used to help someone remember can influence their statements. Tests showed that witnesses asked to estimate the speed of a car when it ‘smashed into’ another were invariably higher than when the question involved the words ‘hit each other’.

How we remember

Many people believe that memory works something like a videotape. They think storing information is like recording - and remembering is like playing it back. Memory, however, does not work in this way. It is a feature of human memory that we don't store information exactly as it is presented to us. Rather, we extract from information the gist, or underlying meaning, depending on those aspects which are most relevant to us. In other words, people store information in the way that makes the most sense to them. We distort unfamiliar or unconsciously ‘unacceptable’ information in order to make it ‘fit in’ with our existing knowledge. This can, therefore, result in further distorting eyewitness testimony.

Looks like catching an incident on your camera (then face recognition technology or a beady eye, like Simon's!) is the only way to be sure - and of course, this is becoming more and more likely with mobile phones at our fingertips. (click here to Tweet this post)