Résumé
Child witness and victim interviewing practices were examined in a sample of interviews (N = 45) from Canadian police officers. Specifically, the interviews were coded for introductory behaviours (e.g., building rapport), inappropriate interview behaviours (e.g., interruptions), the type of questions asked, the type of response given by the child, the length of the child's response, the number of unique central and peripheral details given by the child, and the proportion of words spoken by the interviewer(s) and child. The lengths of the complete interview and the substantive phase of the interview were also recorded. Results showed that, on average, approximately 8% of all questions asked were open-ended invitations. Open-ended invitations resulted in the longest interviewee response, along with the greatest number of central details. The implications of these findings for reforming child interview practices and the need for training and feedback systems are discussed..