December 29, 2002

NEWLY VALIDATED TEST OF PRESCHOOLERS' SUGGESTIBILITY MAY PROVE USEFUL IN LEGAL INVESTIGATIONS
After age four and a half, measures of yield and shift predict children's susceptibility to influence

WASHINGTON --Although young children can be suggestible, which can confound legal cases hinging on their eyewitness testimony, children vary in their degree of suggestibility. In response, psychologists have developed a new tool to assess aspects of suggestibility as well as individual differences among children. The research, published in the December issue of the Journal of Experimental Psychology: Applied, also helps to clear the misconception that all children are suggestible in the same ways, to the same degree, at all ages. The journal is published by the American Psychological Association (APA).

In a two-phase study of 50 preschoolers at Cornell University, Matthew H. Scullin, Ph.D.; Tomoe Kanaya, M.A.; and Stephen J. Ceci, Ph.D. validated the constructs measured by the Video Suggestibility Scale for Children (VSSC) that Scullin and Ceci published in 2001. They did so by comparing VSSC results with children's responses to suggestive questioning about staged events. The high correlation between the two approaches confirm the test's potential utility for lawyers, judges, caseworkers and psychologists involved with cases of abuse and neglect, child custody, and Persons in Need of Supervision actions. Investigators need, say the authors, to "identify children who are prone to making erroneous reports following suggestions. The VSSC can provide important information about when special interviewing precautions should be taken."

Scullin et al. split their participant pool between 25 children under and 25 children over age four and a half, because around that age, most children start to exhibit "theory of mind," the ability to understand that other people may have different thoughts from their own. Theory of mind helps children to grasp, for example, that people might try to manipulate them through leading questions. By studying preschoolers at both age levels, the researchers gained insight into the developmental timetable of suggestibility.

First, children were randomly assigned to take part in one of two actual events at their preschool. In one scenario, each of 25 children helped a stranger find her son's stuffed monkey outside the classroom. In the other, each of 25 children helped a stranger carry Play-Doh down the hall to another room; the stranger tripped and hurt her ankle, which was bandaged by a "nurse" who "happened" to pass by.

The researchers then interviewed the children every week for four weeks. Interviewers questioned the children about the event in which they took part, and also falsely suggested that they took part in the other event, describing it as if it really happened. Researchers used suggestive techniques as they mixed accurate and inaccurate information in their interviews: asking leading questions, reinforcing desired information, and applying indirect peer pressure ("Sally and Carol told me you were one of the kids who helped the nurse").

Scullin et al. tabulated the children's accuracy and inaccuracy in response to suggestive questions about both events (true and false), as well as how often they denied something the interviewer said.

A few weeks later, researchers tested the 50 children with the Video Suggestibility Scale for Children. First, the children watched a short video called "Billy's Birthday Party," individually or in small groups. Within days, a new interviewer checked in an open-ended manner what the children remembered about the video, and then asked 18 questions, 14 of which were leading. This phase of the study allowed researchers to see whether the children's performance in the earlier interviews corresponded with their VSSC scores, including their tendencies to "yield" (respond affirmatively to suggestive questions) and to "shift" (change answers in response to negative feedback).

The older children who, on the VSSC, tended to yield and shift more, had also been less accurate about Phase 1's true event (monkey or Play-Doh). Their VSSC Total Suggestibility measures also positively correlated with 10 of the 11 inaccuracy and suggestibility measures studied in Phase 1. (For example, an inaccurate response to yes-no leading questions about the true event, such as "Yes" to the question, "Did the lady have curly brown hair and a bright pink shirt?" when they had met a woman with straight black hair with a red streak in it, wearing a blue shirt.)

For the older children only, the Yield subscale was a significant predictor of whether a child assented to the suggested event in the free-recall portion of Phase 1's final interview. The higher-yielding, higher-shifting children also gave significantly more of both accurate and inaccurate answers about the suggested event, in both cases telling the interviewer what he or she wanted to hear (an important element of suggestibility when being interviewed). However, children, especially older ones, are less likely to shift their answers if the question is about a well-remembered detail.

There were no significant relationships among the critical variables for the younger children, meaning that at this point, researchers have validated the VSSC only for the older age group. Refining the test for younger ages could prove important, because, says Scullin, "Younger children have greater difficulty with reality monitoring than older children and adults."

The results, say the authors, provide "substantial support for the construct validity of the VSSC among children over the age of four years and six months." What this means as that Scullin et al. now feel confident that the VSSC measures what they hoped it would measure -- individual differences in suggestibility.

The authors conclude that the Video Suggestibility Scale for Children is a "valuable addition to the research on children's suggestibility, especially suggestibility in forensic contexts." The test could also help to train investigators, adds Scullin. "Hundreds of thousands of young children are interviewed every year for various reasons," he explains, "and many of the adults who interview them have no idea how easy it is to lead some children to say what the interviewer wants them to say."

Article: "Measurement of Individual Differences in Children's Suggestibility Across Situations," Matthew H. Scullin, Ph.D.; Tomoe Kanaya, M.A.; and Stephen J. Ceci, Ph.D., Cornell University; Journal of Experimental Psychology: Applied, Vol. 8, No. 4.

Full text of the article is available at http://www2.apa.org/releases/child_suggestibility.pdf

---American Psychological Association (APA)