Children better witnesses than previously thought, Swedish psychologist finds

Young children who witness crimes are good at recounting the events. It is true that they include fewer details than older children and adults, but what they say is more accurate. This has been shown in a study which Fredin has carried out on children in the age groups 8-9 and 11-12 as well as adults.

"Older children and adults have had more experiences and can therefore find it difficult to differentiate between reality and fiction," says Fredin. "Adults, however, are conscious that they may get things mixed up, while children in the 11-12 age group are often very certain, even when they are wrong."

In Sweden, where the study was carried out, there has not been any previous research on children and identity parades, but the general perception has been that young children are not particularly reliable, explains Fredin.

In another study, Fredin set up identity parades in line with the model recommended by Swedish police. Children aged 11-12 and adults were shown a film of a robbery. They were then asked to identify the robber from photographs.

The model is known as a sequential identity parade. The witnesses only get to see one photograph at a time, never all together. However, they can see each picture twice, and it is here the problem lies, according to Fredin.

"As the pictures are shown repeatedly, the witnesses' memory images are destroyed; in the end they believe that they recognise the culprit, yet the person might not even have been involved in the incident," she says, adding that 60 per cent of the children and 50 per cent of the adults pointed out innocent individuals.

In Fredin's view, consideration must be given to the fact that children are eager to please and may not always understand the consequences of pointing out an innocent person.

Fredin thinks that the value of an identity parade which yields such poor results can be questioned. In Sweden, unlike in the USA, a conviction cannot be secured on the evidence of one eyewitness only. But even in Sweden, eyewitnesses are seen to provide important help to an investigation.

The method which produced the best results in Fredin's study comes from Canada and is known as the elimination method.

"In this method, all the photographs are shown at the same time and the witness has to remove photos that do not depict the perpetrator one at a time. When there is one photograph left, the witness decides whether or not it is of the perpetrator," says Fredin.

Another finding in the thesis, which is perhaps not all that surprising, is that we pay more attention to people of the same age as us. Adults are generally better at recognising adult faces than children are, but one of Gunilla Fredin's studies showed that they did no better than children when asked to identify children's faces.

Fredin will defend her thesis on 16 September. The thesis is entitled Children as eyewitnesses: memory recall and face recognition.