Intelligent tutoring for interviewing to detect deception: Can investigator be trained to elicit and detect accurate cues to deception?

The present study examined whether investigators can be trained to elicit and detect cues to deception. The study included two training conditions and a control condition. Participants in the virtual Human Intelligent Tutoring System (vHITS) conditions completed a training program for deception detection and investigative interviewing. The primary components of this training were one-on-one interaction with a virtual human and tutoring tailored to specific participant responses. Participants in the Computer Based Training (CBT) conditions completed a comparable training program for deception detection and investigative interviewing. However, this program provided a more passive training environment with no interaction between student and tutor. Participants in the control conditions received no training. The performance of all three conditions on a deception detection task and an investigative interviewing task were compared. Results showed significant within-group differences between pre-test and post-test performance for both training conditions on all measures. However, results did not show significant between-groups differences at post-test for the training conditions on any measure. Possible explanations of these results are discussed as well as implications for future research.