Narrative elaboration: A test of effectiveness of a recall improvement procedure with low-SES preschool children.

Abstract

Preschool children are at high risk for child abuse, and low-SES children are disproportionately represented in reported abuse cases. Preschoolers' accounts of past events are typically vague and incomplete. Eliciting more complete accounts of a past event usually requires further questioning, but some research demonstrates that preschoolers are more vulnerable to suggestion from misleading questions than older children. Designed to increase the amount of detail recalled without using specific (i.e. potentially leading) questioning, previous research has found that the Narrative Elaboration technique (NE) increased the accurate completeness of recall narratives of European-American elementary school children. This dissertation tested the effectiveness of NE with middle- and low-SES preschool children. 49 low-SES and 50 middle-SES four- and five-year-olds were randomly assigned to the NE or control condition. Children participated in a staged event designed to simulate many characteristics of events about which children must often testify. One week later, NE group children were trained to facilitate recall by organizing events into forensically relevant categories (e.g. participants, setting, actions), each represented by a generic, non-biasing drawing on a visual cue card. Control group children were instructed to try their best when recalling events, but were not trained in the memory strategy nor presented with the cards. One day later, interviewers asked children to recount the staged event using (1) Free Recall, (2) Cued Recall (with cards), and Probed Recall (open-ended questions). Analyses of variance revealed that NE-trained children from both preschool SES levels recalled significantly greater amounts of accurate information about the staged event, without producing more errors, than control group children, particularly in Cued Recall. In addition, low-SES preschool children generated more errors than middle-SES preschool children, regardless of treatment condition, particularly in Probed Recall. This research demonstrates that recall of preschool children from a variety of SES backgrounds can be improved using techniques based in developmental research. Improving the accuracy of young children's testimony could facilitate assessment, investigation, and prosecution of child abuse, protecting children from real abuse, and protecting innocent adults from false accusation. Implications for future research is discussed.