This data collection, designed to improve the quality of children's testimony in court, evaluates how different types of interview formats affect the completeness and accuracy of children's recall performance. Specifically, the study assesses the impact of a "practice interview" about an event on the completeness and accuracy of later reports about a second, unrelated event. Three interview conditions were employed, and each condition consisted of both a practice interview and a target interview. The three conditions were RS, RC, and CC, where "R" represents a practice session with rapport-building only, "S" represents a target interview that contained all components of the standard interview procedure, and "C" represents either a practice or target interview that contained all components of the cognitive interview procedure. In rapport-building sessions, interviewers talked about school activities, family life, and favorite games with the child. In standard and cognitive interview sessions, the rapport-building sessions were followed by a request from the interviewer for the child to verbalize a narrative account of "what happened" during an event that had been previously staged by the experimenter. This narrative account was then followed by the interviewer's request for additional information about the event. Cognitive interviews also included several additional questions that were hypothesized to improve recall performance. The number of correct items recalled and the number of incorrect items generated were used to compare the performance of children in the three interview conditions.