Interviewing witnesses: the effect of forced confabulation on event memory

After viewing a crime video, participants answered 16 answerable and 6 unanswerable questions. Those in the "voluntary guess" condition had a "don't know" response option; those in the "forced guess" condition did not. One week later the same questions were answered with a "don't know" option. In both experiments, information generated from forced confabulation was less likely remembered than information voluntarily self-generated. Further, when the same answer was given to an unanswerable question both times, the confidence expressed in the answer increased over time in both the forced and the voluntary guess conditions. Pressing eyewitnesses to answer questions, especially questions repeated thrice (Experiment 2), may not be an effective practice because it reliably increases intrusion errors but not correct recall.