Monte Carlo Study of Criterion Validity of the Directed Lie Screening Test Using the Seven-Position, Three-Position and Empirical Scoring Systems

• NCJ Number: 243045
• Journal: Polygraph Volume: 41 Issue: 4 Dated: 2012 Pages: 241-251
• Author(s): Raymond Nelson
• Date Published: 2012
• Length: 11 pages
• Publication Format: Article
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America

Annotation

This article presents a study on the criterion validity of the Directed Lie Screening Test (DLST), using three different versions of the Monte Carlo model.

Abstract

Monte Carlo methods were used to calculate criterion accuracy profiles for DLST examinations scored with the seven-position, three-position, and ESS models. Multivariate ANOVAs were used to compare the results. Decision accuracy was significantly greater than chance, over 87 percent, for DLST exams scored with all three TDA models, and there were no significant differences in the unweighted accuracies achieved by the three scoring methods. The ESS model produced the fewest inconclusive results. Differences in inconclusive rates were significantly greater for the three-position model, and the difference was loaded on deceptive cases. Results suggest that the component weighting achieved by the seven-position and ESS models is more effective than the three-position model at extracting diagnostic information from the raw test data. (Published Abstract)