Information Gain of the DLST Information Gain of the Directed Lie Screening Test

We adapted the Information Gain (IG) analysis procedures described by Honts and Schweinle (2009) to the Directed Lie Screening Test (DLST; Handler, Nelson & Blalock, 2008), a comparison question technique (CQT) variant, based on the Test for Espionage and Sabotage (TES; Research Division Staff, 1995a; 1995b), evaluated using the Empirical Scoring System (ESS). IG provides an index of the increase in information gained by using a test result over predicting the base rate of the target event. The DLST is used in a number of screening settings, including public safety applicant screening and post-conviction sex offender monitoring. We found that the DLST/ESS combination provided significant increases in IG over unassisted professional lie catchers across a range of base rates from 0.01 through 0.94 for deceptive outcomes and 0.07 through 0.99 for truthful outcomes. For NSR results, the IG exceeded 5% across the base rate range of 0.06 through 0.99. For SR results IG exceeded 5% across the base rate range of 0.02 through 0.93. Our analyses provide empirical support to the hypothesis that adding the polygraph component to a screening process adds incremental validity. The findings also remind us of the relative limitations of unassisted professional lie catchers in veracity assessment compared to using polygraph testing. These findings should be of interest to program managers, risk assessors and other consumers of polygraph testing.