A "classic" study of children's memories has shown they can be manipulated more than is currently believed, a conference at Leicester University heard last week.

The findings form some of the first rigorous evidence validating British methods of interviewing children in abuse cases. But they undermine experts who say they can tell if a child's story is true.

The conference focussed on children's allegations of abuse, and on false memory syndrome -- in which it is claimed adults are induced to have fictitious memories that they were abused as children.

Stephen Ceci, of Cornell University, described his "Sam Stone" study in which a man appears in a room of children and leaves after two minutes, having done nothing.

ÖйúAƬ

When questioned in a probing way about the visit, children who were told stories beforehand about how clumsy he was, claimed to have seen him break or spoil things. When asked very leading questions, another group also remembered events that had not happened. Seventy-two per cent of a group of three to four-year-olds exposed to both of these pressures claimed the events happened.

A specialist audience was unable to tell which of the children's videotaped accounts were true but were strongly convinced that they could.

ÖйúAƬ

Graham Davies, professor of psychology at Leicester University, said the study showed that all interviews of children should be conducted as early as possible and videotaped: "This underlines the wisdom of the legislation we have in the United Kingdom."