How well adults can detect if children are lying or reporting misinformation is no better than the odds of chance, reports a new Cornell study. The findings have implications for physical and sexual abuse investigations, which often rely heavily on children's eyewitness reports.
Past research has repeatedly shown that adults are also poor at detecting whether or not other adults are lying.
"Our research suggests this lackluster performance extends to [interpreting] statements made by preschool-aged children," said Stephen Ceci, professor of human development in the College of Human Ecology and co-author of the study with Yi Shao, Ph.D. '10, assistant professor of psychology at Oklahoma City University.
The study, published in Applied Cognitive Psychology (25:1), is the first to assess adults' ability to detect the credibility of children's statements when they were telling the truth, deliberately lying or misreporting due to misleading suggestions.
Children may make inaccurate statements intentionally, Ceci said, if they have been coached to lie, for example, or unintentionally, perhaps due to suggestive questioning that alters the child's memory. In fact, suggestive interviews in investigations are a primary reason for children's unintentional misreporting, he said.
In the study, 129 college students assessed the credibility of statements made by 24 preschool children after a game of Simon Says. One-third of the children had been urged to deliberately misreport what they did during the game; one-third were asked suggestive questions designed to alter their memory of the events (e.g., they were asked, 'Dori [a stuffed animal] touched your knees, didn't she?' even though that did not happen during the game). The remaining third answered neutral questions.
The two types of inaccurate statements were perceived differently, the researchers found. Misinformation that resulted from misleading interviews was more readily detected than outright lies. But while the adults could accurately detect the truth at a rate greater than odds of chance, their ability to detect outright lies and misinformation from leading questioning was at a rate less than the odds of chance. The adults were most confident of their ratings of truth-telling children.
"Humans are inclined to believe what others tell them; they exhibit a truth bias," Ceci explained. This is a two-edged sword that suggests jurors will believe young children's accurate statements, but they will also tend to believe their inaccurate statements."
Misleading interviews are a well-recognized source of memory error yet persist in courtrooms, police departments and in therapeutic settings. In the study, children were subjected to only one event where an experimenter repeated a suggestive question up to three times.
"The effect of the suggestive questioning upon the children was substantial," Ceci said. "During debriefing the misled children generally maintained that their inaccurate answers were correct. They appeared to have incorporated the misinformation into their memory.
"Further research is needed to explore the parameters of these findings," he added. "We had expected that the misled children would be the hardest to detect because their memories had been altered, whereas the lying children were expected to display facial signs of deception that were detectable.
"Because the results were not in line with our expectations, we have several hunches that we plan to probe in subsequent research before these results are ready for translation to practitioners in the legal arena," Ceci said.
The study was supported by the College of Human Ecology.
Karene Booker is an extension support specialist in the Department of Human Development.