DoDPI Banner for Research Papers

Abstract

DOHM, T. E. and IACONO, W.G. Design and pilot of a polygraph field validation study. July 1993, Report no. DoDPI93-R-0006. Department of Defense Polygraph Institute, Ft. McClellan, AL 36205.

This study was conducted to determine whether panels of individuals could accurately determine ground truth by reviewing case material contained in the files of individuals who underwent confession-verified polygraph tests. Panels were presented with all case material collected prior to the conduct of the polygraph exam. To determine how well panels could decide ground truth, the reliability with which panelists arrived at decisions of suspect-guilt and innocence was determined, and the validity of their decision was evaluated by tests. The study was based on selected specific-incident criminal investigative files requested from the military. Files were sanitized (removing all identifying information) with case material placed in chronological order and abstracted. Twenty-seven confession- verified multiple suspect criminal investigative files were then selected and reviewed by three six-member panels consisting of lay persons, police investigators, and attorneys. The results indicated that panelists, regardless of their professional backgrounds, could not decide which individuals committed crimes with high reliability. The decisions of the panelists were not accurate: their decision did not match the ground truth criterion. These results indicated that it is unlikely that panels can be used to establish ground truth with any degree of confidence. Field studies of polygraph validity must use some other procedure to establish ground truth.

Key-words: panel study, validity, ground truth, confession- verified polygraph tests, reliability, field polygraph testing

Director's Foreword

Historically and ever increasingly, there is debate about the value of research results obtained in the laboratory setting and the generalization of those results to field application. When performing studies related to psychophysiological detection of deception (PDD), the academic, scientific and PDD practitioner debate focuses on the problem of determining "ground truth."

Published literature suggests expert panels should be able to accurately determine ground truth by reviewing case materials in the files of individuals who underwent confession-verified polygraph tests. The panel decisions would be based on investigative information absent the results of any PDD examination.

During this study, selected confession verified, multiple suspect, specific-incident criminal investigative files were reviewed by three six-member panels of lay persons, police investigators, and attorneys to determine how well they could decide ground truth, the reliability of their decisions of suspect guilt and innocence, and the validity of their decisions based on ground truth established by confessions after PDD examinations. The results reflect that panelists could not reliably or accurately determine which individuals committed the crimes in question. These results indicate that it is unlikely that panels can be used to establish ground truth with any degree of confidence and further, field studies of PDD validity must use some other procedure to establish ground truth.

Michael H. Capps

Director