Voice Units Found to Detect Stress.
The AFRL has been conducting research on two of the devices that have entered the marketplace in the last few years, Diogenes Lantern by the Diogenes Company and the Vericator by Integritek Systems Inc., to determine whether the technology lives up to the manufacturers' claims. The manufacturer of a third product, the National Institute for Truth Verification, opted out of the study.
One of the study's researchers explains that the products tested claim to detect deception-related stress in slightly different ways. The Vericator establishes a voice stress baseline using about seven parameters before the interviewer poses questions testing deception. The difference in stress level between the answers and the baseline is said to indicate deception.
With Diogenes Lantern, the examiner settles down the subject in a preinterview phase before asking the questions. He or she then examines the difference between the wave pattern of the low-stress voice response against that of the high-stress voice response to detect deception.
According to Darren M. Haddad, the program manager in charge of the research, tests have shown that these devices can identify how speech patterns change under stress. This conclusion appears in a forthcoming report. But detecting deception is another matter.
The AFRL tests used taped interviews of murder suspects who eventually confessed. The tests found that the two devices accurately indicated stress-induced changes in speech patterns, and these correlated with instances where the interviewees were later found to have been lying. (An important feature of the interviews was that the suspects felt themselves in jeopardy of loss of liberty or other punishment, which the manufacturers say is a prerequisite to getting accurate results.)
This first test was not conclusive, however, because it did not indicate whether the devices can differentiate between types of stress, such as that caused by lying and that caused by fear.
The National Institute of Justice, which funded the first round of research to determine whether the technology had law enforcement applications, is now giving Haddad and his team two more years to look at whether these units indicate deception. With the help of the NU, the Department of Defense Polygraph Institute, and the Walter Reed Army Medical Center, Haddad is trying to compile a database of recorded testimony that will help differentiate deception stress from other types of stress.
Regardless of the outcome of the next round of tests, Haddad doubts that voice stress analysis will replace polygraphs. But they might supplement them to provide investigators with more sources of information--or more inputs--along with traditional measures such as pulse, breathing rate, blood pressure level, and perspiration.
|Printer friendly Cite/link Email Feedback|
|Author:||GIPS, MICHAEL A.|
|Article Type:||Brief Article|
|Date:||Sep 1, 2001|
|Next Article:||Rule Proposes Bank Privacy Standards.|