People make terrible eyewitnesses – but it turns out there's an exception
People are often poor eyewitnesses. Psychologists have been demonstrating this in experiments for years, but the justice system in the UK has been slow to catch on. There have been improvements, but lawyers and judges continue to rely regularly on eyewitnesses to convict or exonerate people accused of crimes – in some cases, they might argue they have little option.
Yet it might now be time for psychologists to initiate a different discussion. According to new findings in which I have been involved, it looks as though there is an exception.
One personality type appears much better at recognising what they saw at the scene of a crime than any other. If this is supported by further research, it might enable judges and juries to differentiate eyewitnesses who are likely to be right from the rest of us.
The eyewitness problem
It's not possible to put a number on how unreliable eyewitnesses are, but they certainly get things wrong. Unreliable testimonies have been cited as the biggest contributor to miscarriages of justice, causing three-quarters of all those in murder and rape cases, for example.
One of the main problems is that our memories are not like video tapes – they are open to influence. One leading light in this field has been the American professor Elizabeth Loftus. Her work showed in the 1980s that eyewitnesses' memories are affected if a weapon is used at the scene of the crime, for instance. They focus on the weapon and are poorer at recalling other details as a result.
More recently Loftus co-authored research that showed that eyewitness testimony can be influenced by the wording of a question. Participants were shown a film of a car accident and were then questioned about it. When asked questions with an indefinite article, like "did a car hit the pedestrian?", they were less certain than for questions with a definite article, like "did the car hit the pedestrian?".
The problem is that these latter lines of questioning led them to be less accurate in what they remembered. This highlights how easily eyewitness testimony, or any memory for that matter, can be manipulated by police and lawyers to gain support for their position.
These are just two examples of a body of literature that has consistently undermined the value of eyewitnesses. In recent years, advances in DNA evidence have proven psychologists right about eyewitnesses in a number of cases. Alas DNA is not always available, and courts sometimes have to choose between eyewitnesses or nothing.
Pragmatism aside, the UK justice system has tended to be wary of psychology because it sees it as a more subjective science than, say, biology. It is less likely to call psychologists as witnesses than in the US or Germany, for example.
Having said that, attitudes have been changing somewhat in recent years. The police and the courts have sought to reduce inaccuracies in eyewitness testimony using several methods. The first is to try to eliminate variables that the system has some control over – including both the kind of interview questions I mentioned previously, and also cautioning witnesses ahead of identity parades to take their time and not identify rashly.
The second method has been to look at individual differences between eyewitnesses. This can include testing their eyesight or hearing as part of the evidence, for instance.
On the back of our new research, personality might also be relevant in future. We recruited 80 participants, 40 men and 40 women. We showed them a film in which two men robbed a woman. We also gave them a common test for the five recognised dimensions of personality: extraversion, neuroticism, openness to experience, agreeableness and conscientiousness.
The participants were then given misleading information and told it was a summary of the film clip. We subsequently asked them to fill out a recognition sheet in which they had to agree or disagree with statements about what they had seen.
The results showed that individuals who were high in the personality trait of openness to experience made much more accurate eyewitnesses. This might be because such people are more critical of misinformation. They are more prone to analyse what they are told and disregard things that are irrelevant.
I should stress that our research was not perfect. We demonstrated a correlation rather than a cause, so we can only say that openness relates to correct recognition in a way that other personalty traits do not. Also, the research was conducted in a quiet space rather than the kind of chaotic environment where you might expect a crime to take place.
Future research will therefore need to try and rectify these problems. This might involve repeating the experiment with two groups of individuals, one of them low in openness and one of them high in openness. This would make it possible to directly compare recognition rates, thereby establishing cause and effect. In addition, this could be carried out in a "real-life" setting.
But with those caveats in mind, the findings point to a new way forward for the justice system in relation to eyewitnesses. Eyewitnesses could be given a personality test and the jury could be told whether they are likely to be accurate based on the results.
This might eventually mean that only the best eyewitnesses are called to witness boxes. If so, it could make a major difference to the prospects of the right people being found guilty of crimes.