ITHACA, N.Y. -- When preschool children were asked weekly about whether a fictitious event had ever happened to them, more than half the 3- and 4-year-old children by the tenth week reported that it had and provided cogent details, according to a Cornell University study.

Even more surprising, however, is that more than one-quarter of the children could not be convinced the event never occurred when the researchers and their parents explained that the events never occurred. Furthermore, professionals who specialize in interviewing children could not distinguish between children telling false or true accounts when they were shown videotapes of the children's "recollections."

"When young children, ages 3 and 4, are questioned by neutral interviewers, they do very well. They recall events with 90 percent accuracy," said Stephen J. Ceci, Ph.D., the Helen L. Carr Professor of Psychology in Cornell's College of Human Ecology who led the study.

"However, when children are repeatedly interviewed over the course of weeks and months with misleading suggestions -- which sometimes occurs in forensic cases -- many come to remember the false events as true and provide detailed and coherent narratives about these false events," said Ceci, whose interviewers asked children, for example: "Think real hard. Did you ever get your hand caught in a mousetrap and go to the hospital to get it off?"

"So compelling did the children's narratives appear that we suspected that some of the children had come to truly believe they had experienced the fictitious events. Neither parents nor researchers were able to convince 27 percent of the children that the events never happened," said Ceci, a well-known research developmental psychologist who has been studying the suggestibility of children's memories for more than a decade.

With Cornell colleague Mary LynCrotteau Huffman, Ph.D., Ceci reported his findings on how suggestible the memories of preschool children are in the July issue of the Journal of the American Academy of Child and Adolescent Psychiatry (36:7, July 1997).

"These false beliefs or false memories appear to arise when children forget the basis of an event's familiarity," said Ceci, co-author of the book, Jeopardy in the Courtroom: A Scientific Analysis of Children's Testimony (American Psychological Association, 1996). "Young children are especially prone to what we call 'source amnesia.'"

Ceci suggests that when children are asked to think periodically about a fictitious event, they imagine a fictitious scenario, initially rejecting its authenticity because it is unfamiliar. Weeks later, when asked about it again, however, they may falsely accept the event's validity because it is now familiar as a result of having imagined the scenario earlier.

"Consequently, it is exceedingly, devilishly difficult for professionals to tell fact from fiction when a child has been repeatedly suggestively interviewed over a long period of time," Ceci said. "These children frequently display none of the indicators of lying or tricking; they look and act the way children do when they are trying to be accurate and honest."

These findings have important implications for cases involving young children, including those related to child abuse and sexual child abuse because in some of these trials children are interviewed many times over the course of weeks, months, or even years. In fact, the average child in the courtroom is interviewed formally 3.5 to 11 times before a court appearance and many more times informally.

"When suggestive interviewing techniques are used, as in these studies and in real cases, they lead to high levels of correct disclosure when the child actually experienced the event in question," Ceci pointed out. "However, the problem is that they also lead to high levels of false assent when the event was not experienced."

The studies were supported, in part, by a grant from the National Science Foundation.