J. Peter Rosenfeld, of the Institute for Neuroscience at Northwestern University, described his findings recently in the journal Psychophysiology. We spoke last week about how the method works -- and its potential limitations.
Explain how your test for terrorists works.
There is a brain wave called the P300 that appears in the EEG in response to stimuli that are meaningful and rare. If you're flashing a series of stimuli on a screen, pictures or words, and occasionally -- 20 percent or less of the time -- something meaningful occurs, you will produce a P300 in response to the meaningful item. If you're planning an act of terrorism in which you'd want to use a certain kind of bomb and I show you 10 kinds of bombs repeatedly in random order, the one that you're planning to use will evoke the P300. It's an automatic response when you recognize something. There are ways you can try to manipulate your response. These are called countermeasures. But we have counter-countermeasures.
What are some countermeasures and counter-countermeasures?
If you see some of the [stimuli] that are not meaningful and you secretively execute a response to those, they will become meaningful and will evoke P300s. Then you can't tell the meaningful from the non-meaningful. That would be the most typical way. But the way we set up this protocol, even though you will produce P300s [to non-meaningful stimuli], the big ones will be for the meaningful stimuli and we will catch you anyway.
Also, you have to be pretty consistent in your response to the irrelevant [stimuli], so that if you decide, ‘Whenever I see a weapon which is not the weapon I'm planning to do the crime with, I'm going to mentally imagine my mother's face,' you take a little extra time to remember what you're supposed to do. Your reaction time increases and we can pick up your attempt to use a countermeasure.
If you are attempting to use countermeasures to some stimuli, there will be a special brainwave that occurs in response to the other stimuli that you are not countering. [The brain wave] tells us you are trying to counter some of the stimuli. That one is called P900.
There are two techniques of using P300 to catch bad guys. There's what I call the old method, which is before 2008, and that method is also accurate. But unfortunately that method can be countered and there's not much you can do about it. That's why we developed the new one in 2008 and that one, so far, has resisted countermeasures. Between 80 and 90 percent of people, even if they're using countermeasures, will be caught.
Describe how you used students to test your technique.
We told [the students] we were in a mock act of terrorism planning. They were to study three pamphlets. One pamphlet described various weapons they might use to attack a city. Another described cities they might want to attack. A third pamphlet described [pros and cons of attacking different cities]. After studying the brochures, the students wrote a letter to their fictitious handler in the terrorist organization and made a recommendation about which city to attack, what weapon to use and what date to attack on. There was a control group. They got brochures about vacation spots and they typed a letter to their significant other or a parent or a friend as to which place they'd like to go on vacation. Then, everybody took a test on dealing with just the terrorist information.
What's your response to the piece in Wired that says that your method assumes too much knowledge on the part of the terrorist and the interrogator?
There is what's called a comparison question test. It doesn't test for concealed information, which they assert nobody ever has. It tests whether someone is lying or not. The comparison question test does not have credibility among the scientists as a test of deception. My response is that if you're going to use a test, you've got to start from somewhere [and] have some idea what they're going to do. You could just as easily use a test that science approves of.
What's the next step for your research?
We have spent the last two or three years trying every kind of countermeasure we can think of. We've really tried to beat ourselves. My sense is that we've got something that's pretty good now. The thing to do now is get into a field situation with real bad guys.
Image: J. Peter Rosenfeld
This post was originally published on Smartplanet.com