Thirty older (m age=71.73 years) and 20 young adults (m age=21.60 years) viewed a videotape of a simulated crime and were then interviewed with either a Cognitive Interview (CI) or a standard police interview (SI). The older participants were interviewed with either an SI, CI, or CI that was modified for older people (CI–M). No differences were found between the CI and CI–M. The CI elicited more information than the SI, without a reduction in accuracy rate. Moreover, the advantage of the CI over the SI was greater for the older than for the young participants. There were no overall age‐related differences. Results are discussed in terms of performance of older witnesses and implications for understanding how the CI functions.

Research has indicated that police may not receive enough training in interviewing cooperative witnesses, specifically in use of the Cognitive Interview (CI). Practically, for the CI to be effective in real-world investigations, police investigators must be trained by law enforcement trainers. We conducted a three-phase experiment to examine the feasibility of training experienced law enforcement trainers who would then train others to conduct the CI. We instructed FBI and local law enforcement trainers about the CI (Phase I); law enforcement trainers from both agencies (n = 4, 100% male, mean age = 50 years) instructed university students (n= 25, 59% female , mean age = 21 years) to conduct either the CI or a standard law enforcement interview (Phase II); the student interviewers then interviewed other student witnesses (n = 50, 73% female, mean age =22 years) who had watched a simulated crime (phase III). Compared to standard training, interviews conducted by those trained by CI-trained instructors contained more information and at a higher accuracy rate, and with fewer suggestive questions.