Abstract: | Three‐ to nine‐year‐old children (n=144) interacted with a photographer and were interviewed about the event either a week or a month later. The informativeness and accuracy of information provided following either open‐ended or direct rapport building were compared. Children in the open‐ended rapport‐building condition provided more accurate reports than children in the direct rapport‐building condition after both short and long delays. Open‐ended rapport‐building led the three‐ to four‐year‐olds to report more errors in response to the first recall question about the event, but they went on to provide more accurate reports in the rest of the interview than counterparts in the direct rapport‐building condition. These results suggest that forensic interviewers should attempt to establish rapport with children using an open‐ended style. Published in 2004 by John Wiley & Sons, Ltd. |