Interviewing with the NSA

George Maschke writes "'Interviewing With an Intelligence Agency (or, A Funny thing Happened on the Way to Fort Meade)' is a humorous and entertaining account of one man's recent experience seeking employment with the National Security Agency (NSA). But this story, newly posted to the Federation of American Scientists website, is also one with a serious message. Written under the pseudonym 'Ralph J. Perro,' it includes discussion of the job interview, psychological testing, polygraph, and background investigation. It will be of interest to anyone contemplating employment with a federal intelligence agency."

Comments

This is a repost that needs to be said....

[Greenisus]

Lie detectors are not effective. This is just being used to scare people into thinking they can't lie. I really wish more people knew more about psychology....

Re:This is a repost that needs to be said....

[George Maschke]

You're right. Polygraph "testing" is a pseudoscientific fraud that is in the same league as phrenology and graphology. You'll find a thorough debunking of it on the website AntiPolygraph.org.

Re:This is a repost that needs to be said....

[ajs318]

Defeating a polygraph test is easy. If you can obtain access to a polygraph machine of your own, you can prove this nicely.

The polygraph machine records various parameters {heart rate, skin resistance, movement} on a moving paper roll as a series of questions are asked. These questions fall in three distinct groups. Now let's pretend we have two candidates: Honest Jane {goody two-shoes} and Harry the Knife {low-life}, going for the same test, and look at their answers to each type of question.

1. Control Questions - e.g. "Have you ever lost your temper?" Jane answers truthfully "yes", Harry answers mendaciously "no".
2. Test Questions - e.g. "Have you ever stolen anything?" Jane answers truthfully "no", Harry answers mendaciously "no".
3. Dummy Questions - e.g. "Is it Friday today?" Jane answers truthfully "yes", Harry answers truthfully "yes".

Dummy questions are not the same as control questions, because the answer to a dummy question is obvious - even Harry cannot lie. But with the control questions, the "wrong" answer is less obvious. But even Jane must have lost her temper at some stage. Harry is obviously lying. A lie in response to a dummy question will be found out straight away. So the dummy questions appear to provide the calibration data. In fact, it is the control questions which determine the truthfulness or otherwise of the subject, because Harry and Jane would answer them differently. Even the test questions, where Harry and Jane would give the same, desirable answer, are not much use.

Most of the test is in the structuring of the questions, and the machinery is a prop. If you ever have to take a Polygraph test, this is what to do {assuming you don't have access to drugs, either stimulants or sedatives, which would balls up the result}. Imagine Jane and Harry. Run each question by both of them in your mind, and see how they would answer. Then say what Jane would say. You will pass the test with flying colours.

My experience

[olympus_coder]

I interviewed with the NSA. Government jobs have their advantages... In any case, as someone who wanted to get a Ph.D. and be a scientist (an now I'm about 3/4 of the way there), I decided against it because anything I ever wanted to publish after the fact would be subject to government sensors due to the clearance you must have to work for the NSA. I declined the initial offer they made (the one before they do the expensive background investigation) so I didn't go through any of the harder core testing or background checks.

When I interviewed (~1999, pre 9/11) it was pretty relaxed. None of the cloak and dagger, don't admit your are interviewing with the NSA stuff.... I pretty much called and explained I needed a flight for an NSA interview, a hotel for the interview, etc. No one ever told me not to talk about it, or keep it secret.

I did, however, sign a non-disclosure agreement that said I wouldn't talk about what I saw inside their facility (and thus I'm not), but that makes perfect sense.

And for a security clearance ...

[taniwha]

and here's a somwhat older story about the perils of applying for a security clearance from risks.d http://yarchive.net/risks/mongrel.html

actual story about a co-worker who went nsa

[Frymaster]

the guy who does the equivalent of my job in the american office left for a position with the nsa a couple of months ago. he went throught the whole rigamarole of interviews and such and got accepted.

here's the interesting (or frightening) part. two weeks before he left for his new job, i had to send a bunch o sensitive data to some management type. so i called up our spook-to-be and said "point me to yr public key so i can send you this data pgp'd and yout can pass it on." his response?

"i don't have a public key. that cryptography stuff is a waste of time."

good luck national security association in all your future endeavors!

Re:"The everything graph"

[pdhenry]

The Everything Graph would be called a pangraph, not just a polygraph...

Re:Psychological test

[dr_canak]

No,

while the MMPI may no *longer* have credibility, as someone has already pointed out, the MMPI-II has tremendous volumes of research validating its clinical scales.

However, unlike a previous poster suggested, the MMPI-II in no way suggests a probability that a mental illness is present or not. The MMPI-II very simply looked at the response patterns of different groups of patients, and looks at the degree to which your response patterns matched theirs.

So for example, a question like "I like mechanics magazines" does not in and of itself, having anything to do with a psychotic disorder. However, if 85% of schizophrenic patients endorsed this item as "True" and 10% of Depressed patients report "True", then an endorsement of "true" puts you closer to the Schizophrenic camp and further from the Depression camp. But that is just *1* item. The MMPI-II has hundreds of items, so it's a very good identifier of pathology. Not perfect, but very good.

Over the years, there has been tons of psychometric research trying to further figure out what items can help discriminate one group of clinical patients from another, or distinguish clinical patients from "normal" persons.

hth, and yes IAALP (I am a licensed psychologist) ;-)

jeff

Re:I have to agree

[Copperhead]

I know HOW they conduct their evaluation process. I simply wanted to know why I was refused.

FYI, if you're denied a security clearance, you have the right to appeal. The Defense Office of Hearings and Appeals takes care of this.

According to about.com, "If you are denied a security clearance, or an assignment to a sensitive position or a position of trust, or your current clearance or access is revoked, you have the right to appeal the adjudicative decision. Under such circumstances you will be provided a statement on the reason(s) why you are ineligible for the clearance and the procedures for filing an appeal. If you believe the information gathered about you during the investigation is misleading or inaccurate, you will be given the opportunity to correct or clarify the situation."