Interviewing child witnesses

Abstract

The general aim of this thesis was to examine interviewing techniques for child witnesses. Specifically, the effectiveness of the cognitive interview with children was investigated. In addition, the self-reported practice of police officers who interview child witnesses was explored. The principal objective of Study I was to explore how the cognitive interview affects children's recall after a long delay. In Study II the effects of the cognitive interview on the validity of a reliability assessment technique, the Reality Monitoring technique, was assessed. Study III examined whether the cognitive interview would be improved by physical reinstatement of sensations. Study VI investigated the self-reported practice of police officers who interview child witnesses in the United Kingdom , the Netherlands , and Sweden . In study I, 10-11-year-old children (20 girls and 29 boys) saw a film. Half of the children were interviewed after seven days and half after six months. At each test session, half were interviewed according to a cognitive interview (CI), and half according to a structured interview (SI). The children in the CI condition recalled significantly more correct information than the children in the SI condition, both after seven days and after six months. Study II examined whether the Reality Monitoring framework is a valid method for assessing the reliability of statements obtained from a CI. Fifty-eight 10-11-year-old children (27 girls and 31 boys) participated. One-third watched a film and were interviewed according to a CI and two-thirds made up a story and were interviewed according to either a CI or a SI. The CI statements based on observed events contained more visual, affective, spatial and temporal information compared to CI statements based on imagined events. The CI statements based on imagined events did not differ from the SI statements based on imagined events. Considerable developmental work is needed to turn the Reality Monitoring technique into a valid reliability assessment technique. In study III, 6-7-year-old children (38 girls and 37 boys) were presented with a smell, a song and a taste while watching a live event. Each child was allocated to one of five interview types; the CI , the SI , the CI while physically reinstating either the same smell, sound or taste as was present during the live event. No significant differences were found between the interview types. In study IV, 230 police officers (159 women and 71 men) from the United Kingdom ( n = 59), the Netherlands ( n = 49), and Sweden ( n = 123) completed a questionnaire on how to interview child witnesses. Significant differences were found between the three countries. Despite possible discrepancies between actual and self-reported practice, the results support that adequate training being available to practitioners is crucial. In sum, this thesis demonstrates that the cognitive interview can be used successfully with children, both after shorter and longer delays. The results also indicate that children's recall is sensitive to interviewer interference. Finally, practitioners and researchers are advised to work together to increase the benefits of future research on interviews with children.