Those lie detectors again!

[SKIRRID5]

For my sins, I sometimes switch on Radio 5 for the news headlines. This morning I heard part of an item about these damned gadgets. Harrow Council has been using them to detect benefit fraud, and their spokesman proudly named the large sum they’d saved as a result. The interviewer (an intensely irritating, facetious idiot called Nicky Campbell) pointed out that some people wouldn’t speak English, expressing amazement at the idea of, as he put it, “A Gujerati lie detector”. Clearly he’s one of those who think lie detectors literally read your thoughts.
The main news bulletin informed us that more of these things are to be used by councils. The annoying thing is, they’ll get some results, for exactly the same reason as fringe therapy patients “feel better”, because they’ll believe the machines do work.
Now, lie detector evidence is not accepted in US courts, I believe. Is it here, can someone tell me? If not, surely demanding that someone take a test infringes their legal rights. A reasonably sensible newspaper (perhaps the Indie?) should mount a campaign telling people to refuse such tests. I might write to them.
A thought: I hope this won’t bring you-know-who out of the woodwork to say that the councils are to be applauded!

[Admin]

This is bad news for genuine claimants who will get nervous when they're informed that they're being monitored and great news for benefit cheats who know that this technology just doesn't work: false negatives will serve to protect them from being investigated.

Despite being promoted as 'just another tool' to help against benefit fraud, I'm sure their results will be given more credence than is warranted.

They have been piloted and the government claim that of those who have been picked up as 'a risk' by these voice stress analysers some have subsequently been found to be guilty of benefit fraud.

That may seem as being justification for spending money on them but I'd be interested to know whether if a random-number generator had been used to select claimants for investigation the results would have been much, if any, different.

Normal polygraphs are ~70% accurate and these voice stress analysers are said to be less accurate. They're not exactly a great help if they're hardly more accurate than the 50% accuracy of flipping a coin!

[Electric Angel]

I do not have a correct reference, but I do remember when the lie detector first was developed in the United States, they decided to test it on a person who had admitted murder.

He passed the test with flying colours, when asked how he'd managed to do so, his response was "Easy...I lied"

(It was on one episode of QI)

[SKIRRID5]

There's a good article here:

http://commentisfree.guardian.co.uk/ian_williams/2007/04/politicians_need_their_bumps_f_1.html

I hadn't realised these were not standard polygraphs, but that doesn't make much difference, because as far as I can see the voice thingies are equally unreliable.
As the author of the article points out, huge sums are unclaimed. We never see headlines in the tabloids about that though, do we?

[Tomolac]

in the police force lie detectors are used more as an interrogation tool as opposed to a actual lie detector, working on the person they are interrogating believes they can tell when they are lying. but if you tell them you know how the machines work before hand then they likely wont bother as the desired effect will be gone. voice detectors suffer the same set back, and are completely useless over a phone as an interviewer has absolutely no way to determine what factors are causing the "stress" triggers in the callers voice, say having just stubbed your toe or whatnot but will work on the same principal to trick people into thinking that they will know when your lying.
so stop giving the game away, if you tell people about it you'll ruin it all and those millions will be down the drain :tongue::cheesy:

[Admin]

I haven't seen anything written about just how easy these things are too fool.

They seem to get their baseline measurement by talking to the person on the phone and recording them and then comparing their voice recorded then to it later on when the person is answering questions about their circumstances.

The idea being that when we lie, the pitch of our voice rises etc.

To fool this thing, all you'd have to do is tense yourself up whenever you're on the phone to them (the buttock clenching technique would probably do) and when you're answering questions about your claim, relax yourself.

How long will it take for those people who are in the benefit culture (and who routinely make false claims) to learn about how to fool this thing?

As I said earlier in the thread, if you are a benefit cheat, this move by the government will actually help you defraud the system.

I wonder if John Hutton MP has heard of the Doctrine of Unintended Consequences.

They're not too big on evidence-based thinking New Labour, are they? :undecided:

[SKIRRID5]

Have you ever met a compulsive liar? I have, and I suspect there are quite a few of them about. Will they show stress when lying, or do they temporarily believe what they’re saying? It’s not quite the same thing, but there are also lots of people whose brains virtually block out unwelcome information. The world must be the way they want it to be, and to an extent they convince themselves that it is. They’d fool the machine and its operator, because they fool themselves.

[SKIRRID5]

Here's another thing - suppose the subject is Chinese or Japanese. We use tone for expression, don't we? But in Chinese tone functions differently. What seems to us the same word can be another word, if spoken with another tone. There aren't I suppose, all that many Chinese speakers, percentage wise, in our lovely country, but still...
My great expertise in this field is based on viewing many crazy Hong Kong films!

[Cuddles]

Here's another thing - suppose the subject is Chinese or Japanese. We use tone for expression, don't we? But in Chinese tone functions differently. What seems to us the same word can be another word, if spoken with another tone. There aren't I suppose, all that many Chinese speakers, percentage wise, in our lovely country, but still...
My great expertise in this field is based on viewing many crazy Hong Kong films!

Good point. Even though Chinese speakers don't make up a particularly large group in Britain, there are pleny of other people who speak other languages as their native language. Even various different English dialects have very different tones of voice. I really can't see this being in any way useful.

[Tomolac]

but the thing is that even in the areas where a poly graph is still used it is not used to determine if a person is telling the truth or not, its used merly as an aide, the people that use them know they can't tell didly squat about if the person is telling the truth and so the information supplied by it is never relied upon. I think you'll find the only people who think its reliable are the people on Trisha. ;D

[SKIRRID5]

but the thing is that even in the areas where a poly graph is still used it is not used to determine if a person is telling the truth or not, its used merly as an aide, the people that use them know they can't tell didly squat about if the person is telling the truth and so the information supplied by it is never relied upon. I think you'll find the only people who think its reliable are the people on Trisha. ;D

I'm not quite with you, I'm afraid. Operators know it doesn't work, but it's used as an aid? How does it aid them if it doesn't work? Few people think it reliable? In fact I'd guess vast numbers of "ordinary people" believe the things work, and they take a refusal by a suspect to take the test, as a virtual admission of guilt.

[Tomolac]

as said in american polygraph is inadmissable as evidense, and is much the same over here. its basically just used to increase the stress levels of the suspect, to make them slip up, along the same lines as when you work harder because your boss is watching you, or any other interogation technique. but apparently it can tend to end up that if they get the results they wanted then they will use them as supporting eviedense, but if they don't then its ignored, it carries no weight on its own but a confession brought about in part by a polygraph is still a confession. of course if you know its bunk then you won't be intimidated by it.

[SKIRRID5]

"If they get the results they wanted". I don't like the sound of that. Surely the result they should want is the truth. And that phrase about increasing stress levels, I like even less. Once the law allows that, we're on a slippery slope. How is such a procedure different from the absurd attempt to con Colin Stagg into a confession? A confession to a crime which, we now know, he didn't commit.
Surely we have to remember that in a civilised country you are supposedly innocent until proved guilty. The burden of proof is on the police. That does not, or should not, mean that they start with a belief that such-and-such is the case, and then use underhanded methods to make their belief sound plausible.

[Tomolac]

well since it can't really be used in court then they can only use it to try and get a confession, wouldn't be any point in telling them the polygraph shows their telling the truth, as you said its innocent until proven guilty, so why would they bother building a case proving their innocents, what would be the point?
and all integration/interview techniques are design to stress, confuse or intimidate the suspect, even down to the old good cop/bad cop cliché. how exactly do you think they can get a confession or information out of someone who is guilty? what kind of a slope is it your imagining here? the Rachel Nickell murder case shows that even a confession alone won't guarantee a guilty verdict, so where is this slope?
the innocent until proven guilty system is set up so that even if the cop thinks their guilty, they still have to prove it first. and lie detectors are about as effective as the good cop/bad cop bit, its just that fewer people are currently aware of the lie detector technique.

[SKIRRID5]

Regarding the Nickell case, it's not that simple. The "confession" wasn't sufficient, not because no confession would be, but because this one was thrown out on account of the way it was obtained. Anyway, he didn't actually confess, did he? The lady cop (who ought to be deeply ashamed of herself) tried to make him say he killed Nickell, but he wouldn't. If he had made a regular statement at the outset, saying he did the murder, I assume that would count.

[ZERO]

I agree with SKIRRID5.
The slippery slope of coerced confessions is one we don't want to go down.

[Tomolac]

fair enough, I dont really know anything about the case as you might have been able to tell :tongue:
but as you said it was the way it was obtained, so if a polygraph was part of the confession then they would mention it, if it wasn't then they wouldn't bother.
well if we can't go down that slope then we had better just hope only honest people commit crimes from now on........

[ZERO]

well if we can't go down that slope then we had better just hope only honest people commit crimes from now on........
We had also better hope people innocent of any wrongdoing are not arrested and coerced into a confession.

At the bottom of that slope is torture.

We have to be careful to not allow people's rights to be abused. With all the pressure tactics police use it can be a fine line remembering that sometimes accused people really are innocent.

Convictions at any cost a better society does not make.

[Lord Muck oGentry]

For my sins, I sometimes switch on Radio 5 for the news headlines. This morning I heard part of an item about these damned gadgets. Harrow Council has been using them to detect benefit fraud, and their spokesman proudly named the large sum they’d saved as a result. The interviewer (an intensely irritating, facetious idiot called Nicky Campbell) pointed out that some people wouldn’t speak English, expressing amazement at the idea of, as he put it, “A Gujerati lie detector”. Clearly he’s one of those who think lie detectors literally read your thoughts.
The main news bulletin informed us that more of these things are to be used by councils. The annoying thing is, they’ll get some results, for exactly the same reason as fringe therapy patients “feel better”, because they’ll believe the machines do work.
Now, lie detector evidence is not accepted in US courts, I believe. Is it here, can someone tell me? If not, surely demanding that someone take a test infringes their legal rights. A reasonably sensible newspaper (perhaps the Indie?) should mount a campaign telling people to refuse such tests. I might write to them.
A thought: I hope this won’t bring you-know-who out of the woodwork to say that the councils are to be applauded!

Here is the website of the company marketing this stuff:
http://www.digilog.org/dgfaq2.html

Perhaps wisely, they avoid calling it a lie-detector. But they do so in such a mealy-mouthed way that the average journalist or reader could be forgiven for thinking that that is exactly what it is supposed to be.

Have a look too at this report:
http://www.timesonline.co.uk/tol/news/politics/article3387260.ece

The technology is used not to make decisions on entitlement but, in conjunction with the operator's judgement, to pick out cases for further scrutiny.

I see two dangers here. The first is that the technology may be entirely parasitical on the operator's good sense in detecting suspect cases—so, even where there is cause for suspicion, the taxpayer is paying for rubbish.
The second is that even if the system causes no more than delay in settling innocent claims, the people affected may be no more than one benefit payment away from destitution or eviction. The higher the rate of false positives and the lower the true rate of dishonest claims, the greater the danger.

[DrS]

And if they're set targets ..... :undecided:

[Lord Muck oGentry]

And if they're set targets ..... :undecided:

Of course there will be targets! " What gets measured gets done"!

And when books are to be cooked, they will be cooked at Gas Mark Umpteen for as many years as it takes. ;D

[davron]

Such stupitity, using lie detectors.

When you use polygraph or variants you need to run a set of tests to calibrate against the persons known values. Mothers name, education, religon, you need to evaluate against known truths etc. It takes a good half hour to get a correct reading. Things that can give false readings are sweating, dehydration, irritation caused by questions, uncomfortable chair and need of toilet facilities.:sad:

To cheat is easy... clench stomach and breath very shallow. when telling lie relax stomach, breath deeply and sigh, relax neck and readings all change.O0

The chances of actually geting true result are nil on a quick machine with an opperater with no skills. What is happening is fear and they would probably all confess to being Bin Ladens mother if you asked them.

The stupidity of small mind officials always leves me stunned and full of awe.???


What we need to do is put them in the chair and ask about expence accounts!O0

Davron

[Mulder]

Even if you could design a foolproof, working lie detector, it wouldn't detect confabulation. People who think they've been abducted by aliens would pass the test easily.

[davron]

Standard training in the police/military is 6 months to use a machine. Alongside this you have to work for a year beside an experianced opperator to gain the understanding of the small things. Studies show that there is a 25% difference in understanding and readings between established and new readers.

I do not see a pen pusher spending that amount of money on the training for the use of such a machine. The 25% difference in understanding of the small things is also a problem. Those with mental disorders and those who are 'flying saucer man' cannot be detected. So it looks like 25% of the readings will be wrong. Given level of people on Prozac in community that takes a big cut as well, so its now down to the area of guessing games.

Davron

[FarSideOfTheMoon]

I believe MRI has shown some potential for being able to tell when the brain is lying. However this is clearly impractical for general usage.

With a sci-fi hat on, I think it is possible there will be scanners in the future that might have this capability. There was something on the news a couple of nights ago about how the same technology used to produce deep space images minus the distortion, are being used to sharpen MRI images which have been affected by patient movement.

Hand held brain scanners anyone?

[Tony Williams]

I think that there will always be a percentage of people who can lie without detection however good systems get, because it comes as naturally to them as telling the truth.

And of course, there are those who are "method actors" and can convince themselves that whatever they say is true, so they lie with complete sincerity. I can recall a certain recent Prime Minister who seemed to have that knack...

[tolman]

If I were in charge of benefit fraud investigation, I'd be tempted to have "We may use voice stress analysis" added to the "We may record this call" preamble for people calling in, even if I had no intention of using it.

Even if the technique is almost (or even complete) bollocks, it does provide some potentially useful cover for situations where there have been tip-offs about someone committing fraud.
People who think a machine might have caught them out may be a little less likely to wonder if their neighbour had done the good-citizen thing.

[Admin]

These things simply rely on what's known as the 'bogus pipeline technique'. This is where people have a tendency to answer questions or give responses more truthfully when they believe that they're being monitored by some form of 'lie detector'.

Of course this relies on the belief that people have that these bogus devices actually work.

If a fraudulent claimant knows that these 'lie detectors' don't actually work then it has no effect on them; and indeed, if the operator believes the machine works too and they don't get indications of lying they may well not single out such people for investigation. i.e. these machines could actually help fraudulent claimants by producing false negatives.

And you can bet your bottom dollar that the very people who will cotton on to the fact that these machines don't work will be the fraudulent claimants.

[polomint38]

I think that there will always be a percentage of people who can lie without detection however good systems get, because it comes as naturally to them as telling the truth.

I know a few people where lying is more natural than the truth

[lazerustheduck]

If I were in charge of benefit fraud investigation, I'd be tempted to have "We may use voice stress analysis" added to the "We may record this call" preamble for people calling in, even if I had no intention of using it.

Even if the technique is almost (or even complete) bollocks, it does provide some potentially useful cover for situations where there have been tip-offs about someone committing fraud.
People who think a machine might have caught them out may be a little less likely to wonder if their neighbour had done the good-citizen thing.It would provide no worry for the professional (and there are plenty) benefit cheats and extreme worry for the genuine claimants.
I'm genuinely ill and can't work and got suspended from incapacity benefit for 6 months while they investigated my case, while I do know some one working 2 jobs, getting incapacity benefit and claiming carers allowance for both his wife and himself (Apparently she's a full time carer for him and he's a full time carer for her?) He's been under investigation several times but he knows all the tricks for dodging it, me on the other hand had to nearly starve for 6 months because I was told I couldn't even get income support (which later turned out to be a lie, but you trust what yor told at the benefit office).
All this is designed to do is get rid of the easy claims whether guilty or not. I can't even get DLA for my case even thought I'm virtually housbound because I've got a medical condition that's easily dismissed by the government and I get treated as the cheat.

[tolman]

If someone actually has 2 jobs, what tricks can they have for explaining them away, if the investigators know where the jobs are, unless his employers are covering for him?
Sounds like someone who knows where he's working should maybe make a phone call.

With incapacity benefit, if everyone has periodic medical reviews to decide whether they still qualify and at what level, then that seems to be the point at which most problems can occur, unless a person is seen (or maliciously reported) doing things that they aren't supposed to be able to do.

[lazerustheduck]

If someone actually has 2 jobs, what tricks can they have for explaining them away, if the investigators know where the jobs are, unless his employers are covering for him?
Sounds like someone who knows where he's working should maybe make a phone call.

With incapacity benefit, if everyone has periodic medical reviews to decide whether they still qualify and at what level, then that seems to be the point at which most problems can occur, unless a person is seen (or maliciously reported) doing things that they aren't supposed to be able to do.He seems to have no problem with investigation, he (or his boss) always seem know ahead of time when he's going to get investigated and modifies his behaviour accordingly and yes the main job he has people cover for him and his second job as a handyman rakes him in quite a bit. I have reported him but again he seems to know easily when they are on his tail.
People don't need to be reported doing anything to get taken off benefit just an unfavourable report from a government doctor is enough to do it. The last doctor that I had to go and see reported that hard skin on my feet indicated that I walked a lot. Clever trick he's got since he never even asked me to take off my shoes.

[tolman]

It would certainly seem a risky thing for an investigator to tip off a suspect.

[Dubious Dick]

Has anyone contacted Harrow Council to point out the problems with using this type of equipment? Is anyone aware of other Councils who have picked up on this? Anyone contacted the benefits minister (or whatever title they go under)?

DD

[davron]

Hey, the best thing is two guys who use this stuff for junk TV are going to the law because their findings did not match. Ha! I told everyone the traing has to be very long for any degree of full result.