Using serving policemen and policewomen with real-life witnesses/victims of crime, this study explored the relative merits of three training methods in investigative interviewing. Police personnel of five or more years' experience were trained in either the enhanced Cognitive Interview technique (CI), Conversational Management (CM), or a combination of both (CI+CM). The efficacy of these three forms of training was compared with an untrained (control) group of police, in a before-after design. Both the nature of the interviewers; behaviour (number, rate and type of questions asked) and the relative amounts of relevant information elicited before and after training were looked at. Twenty eight police were randomly allocated to one of the four conditions of training. It was found that the interview-behaviour of the CI-only trained officers changes much more than any of the other trained or control groups between pre- and post-training phases (p < .001). These findings offer support for the ecological validity of the Cognitive Interview as a powerful investigative interviewing technique in eliciting potentially relevant information and strongly implies its superiority to Conversational Management in this regard. The finding that CI+CM did not produce the best result overall is discussed. The marked commonality found in the pattern of questioning in all four pre-trained groups offers insight into the 'standard police interview' which, while often used as a control condition, has remained largely unarticulated in published studies.
