NCCA Banner for Research Papers


YANKEE, W. J. An exploratory study on the effectiveness of event- related stimuli as a control procedure in the psychophysiological detection of deception. October 1993, Report No. DoDPI93-R-0003. Department of Defense Polygraph Institute, Ft. McClellan, AL 36205.

To evaluate if event-related relevant questions could serve as control questions to discriminate between programmed guilty persons (PGP) and programmed innocent persons (PIP), rank order scoring was applied to the test results of forty individuals evenly and randomly assigned to innocent and guilty groups. The results showed that event-related relevant questions can be used as controls and will discriminate PGP from PIP.

Key-words: psychophysiological detection of deception, PDD, control question tests, event-related questions, event-related controls

Director's Foreword

This was the first attempt to develop a psychophysiological detection of deception (PDD) test that might correct for the several limitations and criticisms (see the text for details) that have been directed toward Control Question Tests. The approach in this study was based on the hypothesis that an innocent person will respond differently to a list of relevant questions regarding an event they were not involved in than a guilty person, who was involved in the event, would. (See the text for the complete rationale). As such, the testing format did not contain control, irrelevant, or symptomatic questions. Rather, all questions were relevant, event-related and associated with the case issues reviewed during the pretest interview. The case issues were based on an espionage and sabotage scenario.

The results of this exploratory study provide sufficient evidence to conclude that under the conditions of this study, an all relevant question test can discriminate between programmed guilty persons (PGP) and programmed innocent persons (PIP). The rank order scoring system used in this exploratory study is rather crude and based on subjective evaluative judgements of the physiological data in correlation with the test questions asked. In an attempt to improve on the diagnostic accuracy of this test, the data collection projects that are planned will be used to develop a data base for an artificial neural network (ANN) algorithm which, it is assumed, will be more objective and precise in evaluating the data for diagnostic purposes.

Michael H. Capps