A Pox on Polygraphs

In the first major U.S. government report on polygraphs since 1983, a panel of the National Academy of Sciences this week said the government should not use the so-called lie detectors to see if an employee poses a security risk. The study, commissioned by the Department of Energy (DOE) in the wake of the Wen Ho Lee affair (Science, 15 September 2000, p. 1851), says that although the devices can be a help to criminal investigators, they are too crude to screen out possible spies.

---

CREDIT: FBI FILE PHOTO

---

The panel, led by statistician Stephen Fienberg of Carnegie Mellon University in Pittsburgh, Pennsylvania, notes that lie detection research "has not progressed ... in the manner of a typical scientific field." It calls for expanded tests of the polygraph and other "indicators of deception."

 The panel has briefed DOE and other agencies that now test thousands of employees. Fienberg didn't say if any was planning a change in policy, but polygraph critics say the course is clear. Physicist Alan Zelicoff of DOE's Sandia National Laboratories in Albuquerque, New Mexico, says the panel has come to a "very strong conclusion. ... As a screening test, [the polygraph] has now been tossed onto the ash heap of history."