Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Brain-Scan Lie Detection Rejected By Brooklyn Court

timothy posted more than 4 years ago | from the precedent-can-be-based-on-principals-though dept.

The Courts 197

blair1q writes "A judge in Brooklyn has excluded Functional Magnetic Resonance Imaging (fMRI) lie-detector evidence from a trial there. However, the decision will not set a precedent, as it was made without even conducting a hearing on the method's validity, but on the principle, argued by the defense, that 'juries are supposed to decide the credibility of the witness, and fMRI lie detection, even if it could be proven completely accurate, infringes on that right.' That principle can be tested in later hearings, such as one scheduled for May 13, 2010, in Tennessee; in this case, the defense wants to use fMRI evidence it has already collected to prove its client is innocent. fMRI has been shown to be 76-90% accurate. That number seems significantly larger than the rate of false convictions."

cancel ×

197 comments

Sorry! There are no comments related to the filter you selected.

The statistical value of "resonable dobut" = 0.1% (4, Informative)

LostCluster (625375) | more than 4 years ago | (#32116442)

The Supreme Court has given science the legal definition that a "beyond a reasonable doubt" equates to 99.9% certainty... a system that is wrong 10% of the time or more needs at to at least be much times more accurate before it's going to be trusted. You're only allowed one blooper in 1000 by this standard. Nice tech, but it's not there yet.

Re:The statistical value of "resonable dobut" = 0. (3, Insightful)

Shakrai (717556) | more than 4 years ago | (#32116466)

Nice tech, but it's not there yet.

What happens to the right to remain silent when it is there? The British have already gutted this right. Not that hard to envision the same happening here.....

Re:The statistical value of "resonable dobut" = 0. (1, Interesting)

Anonymous Coward | more than 4 years ago | (#32116662)

What happens to the right to remain silent when it is there? The British have already gutted this right. Not that hard to envision the same happening here.....

It's a lie detector not a mind reader. If you look into this and other lie detectors it becomes quickly obvious that no application can ever approach 99.9% accuracy. Humans and the reasons they lie are too varied to ever be categorized. Lie detectors are part interrogation and part snake oil and there is a 0% chance they can be made reliable.

Food for thought: Scientologists currently use more sensitive lie detectors than the FBI.

Re:The statistical value of "resonable dobut" = 0. (2, Funny)

WrongSizeGlass (838941) | more than 4 years ago | (#32116750)

It's a lie detector not a mind reader.

Yeah! Do you know how expensive it is getting Kreskin [wikipedia.org] to testify? Plus, I bet he's booked for months in advance.

The right to remain silent (2, Informative)

mangu (126918) | more than 4 years ago | (#32117416)

The right to remain silent is meant to make sure no one can be forced to speak under torture. It does not mean you have a right to keep authorities from getting evidence against you. This means, for instance, that police can force you to take a breath or blood alcohol test.

If it weren't for the right to remain silent, the police could tell you "say you are drunk or I will break all your teeth" but it would be meaningless for them to say "breath here with a BAC of 0.20 or I will break your teeth".

Re:The right to remain silent (1)

sexconker (1179573) | more than 4 years ago | (#32117474)

No person shall be held to answer for a capital, or otherwise infamous crime, unless on a presentment or indictment of a Grand Jury, except in cases arising in the land or naval forces, or in the Militia, when in actual service in time of War or public danger; nor shall any person be subject for the same offence to be twice put in jeopardy of life or limb; [b]nor shall be compelled in any criminal case to be a witness against himself[/b], nor be deprived of life, liberty, or property, without due process of law; nor shall private property be taken for public use, without just compensation.

You're wrong.

Re:The right to remain silent (3, Informative)

Monkeyman334 (205694) | more than 4 years ago | (#32117628)

Gosh. Legal issues are frustrating to discuss on Slashdot. People don't have the right to remain silent. You have the right to not incriminate yourself in court. That means if you are the target of an inevstigation you can be given immunity and forced to talk. If you are a witness and not the target you can be forced to talk. You can be forced to take a breathalizer test because it's not testimonial. None of this has anything to do with this case.

Re:The right to remain silent (1)

mog007 (677810) | more than 4 years ago | (#32117862)

Mandatory breathalyzer tests are, in the US, unconstitutional. The fifth amendment protects against self incrimination, which means testimony OR evidence. The state cannot compel you to provide evidence against yourself. With breathalyzers, if you refuse to blow, you probably lose your license, but you can't be jailed for it.

Re:The statistical value of "resonable dobut" = 0. (0)

Anonymous Coward | more than 4 years ago | (#32116514)

Nice tech, but it's not there yet.

Nice tech? I'm going to have to grab a tinfoil hat before firing up bit torrent now...or firing up in general. :/

It is better than a jury of Bobs (2, Insightful)

linzeal (197905) | more than 4 years ago | (#32116574)

If it is better than what we have for false convictions than why prefer human prejudice/error over machine error. It seems to me one of those is far more likely to improve than the other and I'm not talking about Homo Sapien's ability to use critical thinking skills when confronted with conflicting emotive/subjective versions of events.

Also it is would only be a portion of the evidence which the case depends upon currently for it to reach a verdict. Anyone who compares this to lie detector machines does not understand that lie detectors are little better than a coin toss. This when properly used has been shown to top out at 90%.

Re:It is better than a jury of Bobs (4, Insightful)

Shakrai (717556) | more than 4 years ago | (#32116610)

If it is better than what we have for false convictions than why prefer human prejudice/error over machine error.

Because one of the reasons we have a jury system is to provide a check and balance on the ability of the government to lock people up. Jury nullification may be a bad word in the modern legal system but it's still there.

Re:It is better than a jury of Bobs (4, Informative)

linzeal (197905) | more than 4 years ago | (#32116948)

Well perhaps we need a check on the jury's seemingly endlessly ability to be lured into convicting innocent people out of motley assortment of prejudices and 'gut-feelings'. You can't tell me the fact that across the world minority conviction rates are higher for the same crimes because of anything but bigotry. Don't even start with the idea of ever person having the right to appeal, which can take years if not decades to overturn a wrongful conviction.

If this technology can get to 9X.x% accuracy in the near future I would like it to at least be used to prevent people from having to go to trial or have charges brought against them in the first place. District Attorneys have far too much power right now to prosecute people charge them with 5 crimes that can result in years in prison and 10's pf thousands of dollars for petty crimes like vandalism and than have them plea down to 30 days in jail and a 1000 dollar fine.

My friend in his last year of college was walking home from his art studio down town with all of his art supplies in a backpack near a bank downtown. A rent a cop came out of nowhere and started accusing him of vandalizing the property and started manhandling him and moments later the real cops showed up and arrested him. Why, well he had marker pens in his backpack that were the same type used to write anti-capitalist graffiti on the bank's ATM and someone had superglued the deposit door shut. He was held over night was forced to call his parents for bail, had all of his art supplies confiscated, charged with 3 misdemeanors that could of landed him in jail for 2 years, fought all the charges with a private attorney his parents paid for, lost all the charges got 6 months in jail and a 3500 dollar fine, appealed the conviction, found evidence that another bank in the area after he was arrested had the same graffiti done to it while he was with his parents 100 miles away, went to trial and the judge spent 3 days grilling him on his political activities before overturning the convictions. He got back his art bag, everything was either broken in half or torn and it smelled like feet because it had been locked up with his shoes for over a year at this point. He told me it took over 25,000 dollars to prove he was innocent and if he did not have well off parents he would of been in jail.

Public defenders in this country on average handle over 500 cases a year whereas a DA handles half that, something has to change or more and more of us are going to be going to jail as innocent men. In NYC the average case load of a public defender is 720 cases a year.

Re:It is better than a jury of Bobs (1)

Lehk228 (705449) | more than 4 years ago | (#32118208)

Public defenders in this country on average handle over 500 cases a year whereas a DA handles half that, something has to change or more and more of us are going to be going to jail as innocent men. In NYC the average case load of a public defender is 720 cases a year. seems to me, a constitutional amendment requiring every jurisdiction to provide the same level of funding, staffing, and facilities access to the public defenders office as they do to the prosecution.

Re:It is better than a jury of Bobs (1)

Darkness404 (1287218) | more than 4 years ago | (#32116684)

why prefer human prejudice/error over machine error.

Because humans are more accurate than machines in a lot of tasks. No human exists in a vacuum, a machine does. Feed a decently educated person false facts and they will reject them. Feed a machine false facts and it will believe them.

Also, brain-scans and lie detector tests should not be used because they also subvert the 5th amendment. The constitution says

"nor shall be compelled in any criminal case to be a witness against himself"

Because of that, one should have the right to withhold information that may harm them. With a human jury that is easy enough to do, but with a machine, it isn't. And many times the end result is the same just without the witness being harmed for victimless crimes. For example, if the person says I was studying for a test, when they were really smoking pot, it is a lie, but the end result is the same, they were at home, but a true lie detector could call the entire true story into question.

Re:It is better than a jury of Bobs (2, Informative)

linzeal (197905) | more than 4 years ago | (#32117070)

What if they choose to have the fMRI done? You can still have lie detectors admitted in some jurisdictions so why shouldn't the defense be allowed to use this tool? No one is forcing anyone to do the fMRI in this case.

Re:It is better than a jury of Bobs (1)

osgeek (239988) | more than 4 years ago | (#32117436)

The fifth amendment wouldn't seem to be a license to lie. In order to take the fifth, people say, "I take the fifth". They don't just commit perjury.

I would think that you'd invoke the fifth amendment before even taking the scan or when agreeing to the questions that the prosecution is allowed to ask.

Re:It is better than a jury of Bobs (0)

Anonymous Coward | more than 4 years ago | (#32117500)

Feed a decently educated person false facts and they will reject them. Feed a machine false facts and it will believe them...

Yeah, sure, that's why FOX news went the way of the Dodo so quickly.

Re:It is better than a jury of Bobs (0)

Anonymous Coward | more than 4 years ago | (#32117196)

If it is better than what we have for false convictions than why prefer human prejudice/error over machine error.

Humans would prefer that certain mistakes are made by humans.

It's said that sociologists can indicate with a very high degree of certainty whether someone will become a criminal based on a large set of data about his background and how he grew up. If that accuracy is 90%, should we periodically lock up, or even arrest and thoroughly interrogate, all who fit the profile?

Re:The statistical value of "resonable dobut" = 0. (0)

Anonymous Coward | more than 4 years ago | (#32116702)

The Supreme Court has given science the legal definition that a "beyond a reasonable doubt" equates to 99.9% certainty... a system that is wrong 10% of the time or more needs at to at least be much times more accurate before it's going to be trusted. You're only allowed one blooper in 1000 by this standard. Nice tech, but it's not there yet.

Yes, if it's the only piece of evidence you have. What if I have 3 tests, totally uncorrelated, each of which is 90% accurate and all agree your client is guilty? The chances if innocence in that case meet the 0.1% threshold.

Reasonable doubt should apply to the case as a whole, not a single test. Or otherwise, they can strike any and all eyewitness testimony now, because human observers are notorious for errors.

Re:The statistical value of "resonable dobut" = 0. (1)

geekoid (135745) | more than 4 years ago | (#32116734)

Ironically, personal testimony is worse then 90%. A lot worse.

Re:The statistical value of "resonable dobut" = 0. (1)

jgagnon (1663075) | more than 4 years ago | (#32116830)

I think you're at least 10% wrong. :p

Re:The statistical value of "resonable dobut" = 0. (1)

idontgno (624372) | more than 4 years ago | (#32116970)

Ironically, personal testimony is worse then 90%. A lot worse.

Are you personally testifying to this? If we are to believe you, you're saying there's a 90% chance you're wrong. So we can't believe you.

Damn you for introducing a Liar's Paradox to a perfectly good Slashdot discussion.

Re:The statistical value of "resonable dobut" = 0. (2, Insightful)

timeOday (582209) | more than 4 years ago | (#32116950)

You're only allowed one blooper in 1000 by this standard. Nice tech, but it's not there yet.

What you say would be correct if people were being convicted only on the basis of this fMRI lie detector test. In practice, how you get to a 1/1000 error rate is by combining several less reliable sources. For example, convicting somebody on the basis of one witness is crazy, but convicting them on the basis of 10 witnesses is reasonable. (Given certain assertions of statistical independence etc).

Re:The statistical value of "resonable dobut" = 0. (5, Insightful)

westlake (615356) | more than 4 years ago | (#32117066)

The Supreme Court has given science the legal definition that a "beyond a reasonable doubt" equates to 99.9% certainty

Citation needed.

"Proof beyond a reasonable doubt" is not a formula or a slogan, to be sold, like Ivory soap, as "99 and 44/100% pure."

It only means, that in the light of all the evidence presented, the jury can in good conscience say that the defendant's guilt has been proven to their satisfaction and that any questions which remain will not alter their decision.

Re:The statistical value of "resonable dobut" = 0. (1)

Conspiracy_Of_Doves (236787) | more than 4 years ago | (#32117090)

Just how is a jury supposed to to determine exactly how likely it is that a defendant is guilty, down to a tenth of a percent?

Re:The statistical value of "resonable dobut" = 0. (1)

BobMcD (601576) | more than 4 years ago | (#32118110)

With a cluster of twelve of the most powerful super computers [wikipedia.org] known to date.

But that isn't relevant here. (1)

ElizabethGreene (1185405) | more than 4 years ago | (#32117798)

Scientific evidence is not evaluated on reasonable doubt, but by the Daubert standard [wikipedia.org] . Part of this includes "general acceptance by the scientific community". IMHO and IANAL, this is too new to be generally accepted.

Quick, everyone go read the Truth Machine. :D

Because they are unreliable. (4, Interesting)

Mekkah (1651935) | more than 4 years ago | (#32116512)

They don't tell you shit, I've taken a couple of them for government work and it depends on the person, and their ability to concentrate. You can easily get false positives and easily beat it if you have the right mindset.

Re:Because they are unreliable. (1, Informative)

Anonymous Coward | more than 4 years ago | (#32116632)

Yeah, just ask the salmon [wired.com]

Re:Because they are unreliable. (5, Insightful)

IICV (652597) | more than 4 years ago | (#32116764)

Absolutely no lie detector will ever be able to tell the difference between "this is true" and "at this moment, the suspect believes that this is true". With some mental training (or, you know, schizophrenia) it's entirely possible to temporarily convince yourself that the sky is purple, and there's basically no way any machine will be able to pick up on it in the foreseeable future.

No judgment should ever rely on "the machine says he thinks its true, therefore he is guilty" - no matter how accurate the machine is.

Re:Because they are unreliable. (3, Insightful)

characterZer0 (138196) | more than 4 years ago | (#32116854)

Instead, we rely on "these 12 guys who have been struggling to stay awake during the proceedings think it's true, therefore he is guilty."

Re:Because they are unreliable. (2, Insightful)

Conspiracy_Of_Doves (236787) | more than 4 years ago | (#32117128)

"these 12 guys who have been struggling to stay awake during the proceedings, and just want to go home, think it's true, therefore he is guilty."

FIFY

Re:Because they are unreliable. (1)

tool462 (677306) | more than 4 years ago | (#32117572)

Even if the fMRI lie detectors were allowed, you'd still be relying on those 12 angry men to determine guilt or innocence. fMRI results would lull them into a false sense of certainty, giving them yet one more reason to shut off their brain entirely.

"The machine says he's lying, so he must be lying. Who cares about the rest of the evidence?"

This would be particularly bad in the he-said-she-said types of cases, where there isn't much other evidence besides personal testimony.

Re:Because they are unreliable. (1)

ObsessiveMathsFreak (773371) | more than 4 years ago | (#32118080)

When my freedom's on the line, I'll still take that over a machine test any day of the week.

Re:Because they are unreliable. (2, Insightful)

Ichijo (607641) | more than 4 years ago | (#32118034)

Absolutely no X will ever be able to Y

Where have I heard that before?

Re:Because they are unreliable. (1)

carp3_noct3m (1185697) | more than 4 years ago | (#32116774)

Exactly, I've found that when getting boxed, the biggest issue is when you have some fresh college grad who thinks hes the judge and jury, and if he or she doesnt like you and you flub up at all, sweat a bit too much, wear something they don't like, say an off the wall remark, they get real aggressive. The solution is to find the oldest saltiest bastard around, he can weed through bullshit faster than anyone, you want him to box you. Polygraphs are suedo-science, and everyone knows it.

Re:Because they are unreliable. (1)

akkornel (1800252) | more than 4 years ago | (#32116796)

They don't tell you shit...

That's because they're looking at the wrong end.

Re:Because they are unreliable. (2, Insightful)

Cassini2 (956052) | more than 4 years ago | (#32116848)

More importantly, lie detectors can only tell what the subject believes to be true. Given the number of people in America that believe that the world is flat, that Elvis is alive, that George Bush masterminded the 9/11 bombings, that Oswald didn't kill Kennedy, or that off-sea oil-rigs pose no risks, I think it is safe to say: "All sorts of people in America believe things that are not true."

This is a huge problem for witnesses at accidents. 5 different witnesses will give the cops 5 different stories, and then when the case gets to trial, 5 additional slightly different stories. People remember things in different ways, and some believe strange things. It is a fact of life.

Re:Because they are unreliable. (5, Informative)

Shadow of Eternity (795165) | more than 4 years ago | (#32116916)

Wrong again. Polygraphs can only tell when the subject is showing physical signs of stress through pulse, blood ox, temperature readings, and galvanic skin response.

Maybe I am stressed, I had a bad burrito and I'm terrified that the next fart won't be silent or dry.

An FMRI "lie detector" only shows you what parts of the brain are active on the assumption that certain parts lighting up mean someone is thinking too much and thus making it up.

Maybe my brain IS active, that one girl had some really nice breasts and I wonder what they look like underneath all that clothing...

Re:Because they are unreliable. (1)

SpeZek (970136) | more than 4 years ago | (#32117330)

Maybe I am stressed, I had a bad burrito and I'm terrified that the next fart won't be silent or dry.

Clenching your sphincter at the right times will fool a lie detector. It raises your blood pressure, pulse, core temperature, and your conductivity while you sweat under the stress.


Spies and intelligent criminals laugh at lie detection. It's trivial to control your body when you know how it works.

Re:Because they are unreliable. (1)

Shadow of Eternity (795165) | more than 4 years ago | (#32117680)

I'm not a spy or a criminal, and it was still trivially easy to convince a cheap polygraph that I was a female angel with green wings and Dr. Scholls footwear.

Re:Because they are unreliable. (1)

BobMcD (601576) | more than 4 years ago | (#32118138)

I'm not a spy or a criminal, and it was still trivially easy to convince a cheap polygraph that I was a female angel with green wings and Dr. Scholls footwear.

It doesn't count when you actually are a female angel with green wings and Dr Scholls footwear. How is the machine supposed to know those don't exist?

Re:Because they are unreliable. (1)

osgeek (239988) | more than 4 years ago | (#32117472)

Good point. I see a high conviction rate for horny ADHD slashdotters on the horizon.

Re:Because they are unreliable. (1)

Ichijo (607641) | more than 4 years ago | (#32118072)

An FMRI "lie detector" only shows you what parts of the brain are active on the assumption that certain parts lighting up mean someone is thinking too much and thus making it up.

Close. First the "reciting from memory" and "making stuff up" parts of the brain are mapped, and then you can tell, based on which part of the brain lights up, whether the person is lying or telling the truth.

Re:Because they are unreliable. (1)

timeOday (582209) | more than 4 years ago | (#32117024)

That's why they're called lie detectors instead of false belief detectors. If people had the faith in them that you seem to think they do, then they would be trying to "prove" the existence of god by wiring up a devout believer and simply asking them.

Re:Because they are unreliable. (1)

Korey Kaczor (1345661) | more than 4 years ago | (#32116910)

Just wait until every job one applies for requires these. It's going to happen.

Look at how a lot of businesses already require mumbo-jumbo psychology tests for even 10$/hr wages; once this is cheaper it's going to be adopted for all jobs except flipping burgers or mowing yards.

Re:Because they are unreliable. (1)

Verteiron (224042) | more than 4 years ago | (#32117348)

Grant even proved this in an episode of Mythbusters. He successfully tricked the fMRI specialist into thinking he'd stolen the wrong item. He DID get nailed by the traditional lie detector, though.

Re:Because they are unreliable. (1)

Kazymyr (190114) | more than 4 years ago | (#32117656)

Not to mention that fMRI can show that a dead salmon can have feelings.
http://www.wired.com/wiredscience/2009/09/fmrisalmon/ [wired.com]
After seeing this it will take a LOT to convince me there is anything at all valid about the method.

Not Very Accurate (4, Insightful)

WrongSizeGlass (838941) | more than 4 years ago | (#32116518)

fMRI has been shown to be 76-90% accurate

That's certainly better than a weatherman but not good enough to convict someone.

Re:Not Very Accurate (1)

El_Muerte_TDS (592157) | more than 4 years ago | (#32116818)

Indeed, juries are much more reliable.

Re:Not Very Accurate (1)

ElectricTurtle (1171201) | more than 4 years ago | (#32116870)

Not to mention that comparing its accuracy rate with false convictions as an argument to use it to try to get more convictions doesn't really follow. If a jury acts on the evidence from an fMRI that falls within the 10-24% and consequently produces another false conviction, what, exactly, has been improved?

Re:Not Very Accurate (0)

Jah-Wren Ryel (80510) | more than 4 years ago | (#32116872)

That's certainly better than a weatherman but not good enough to convict someone.

Except, the point is to exonerate someone. It doesn't prove he's innocent, but it adds to the weight of the defense's argument.

I think might be quite useful to admit the results and then instruct the jury about how results might be wrong, like if the guy is a pathological liar and is able to believe his own lies then the MRI will probably believe him too. It might cause the jury to think to wonder, if he might be a pathological liar, perhaps he's been able to fool them too.

Re:Not Very Accurate (1)

BJ_Covert_Action (1499847) | more than 4 years ago | (#32117124)

It doesn't prove he's innocent,

Why the hell is anyone trying to prove he is innocent? Shouldn't he need to be proven guilty while being presumed to be innocent?

Re:Not Very Accurate (1)

Jah-Wren Ryel (80510) | more than 4 years ago | (#32117290)

Why the hell is anyone trying to prove he is innocent? Shouldn't he need to be proven guilty while being presumed to be innocent?

Prove as in demonstrate without a doubt - like being seen by a million people on the jumbotron at a football game at the very same moment the victim was murdered over 200 miles away. That kind of proof it is not. Capiche kneejerk?

Re:Not Very Accurate (1)

john83 (923470) | more than 4 years ago | (#32117104)

It's certainly not good enough for a conviction. I'm not clear why that means it shouldn't be considered though. If a witness indicates that the perpetrator of a crime was blonde, that's weak evidence against a blonde defendant. It's admissible. Is the problem that juries are likely to give too much weight to these kinds of evidence? That doesn't seem to have been the judge's problem in this instance.

Re:Not Very Accurate (0)

Anonymous Coward | more than 4 years ago | (#32117178)

I'm 74% - 90% certain that I saw that defendant doing something that was 74% - 90% similar to the crime he's accused of.

It's the legal industrial-complex, duh! (1)

recharged95 (782975) | more than 4 years ago | (#32116548)

If you are in the business of law (or an attorney), it's about what you can prove with facts, not the truth.

Re:It's the legal industrial-complex, duh! (2, Insightful)

pdabbadabba (720526) | more than 4 years ago | (#32116738)

But of course the whole point is that facts, taken in total, tend to point the way to the larger truth at issue. How else would you have us do it?

Gaming the system (1)

Darkness404 (1287218) | more than 4 years ago | (#32116572)

All that lie detectors will do is encourage criminals to 'game' the system. Every system has flaws, lawyers make a living out of exploiting the flaws of the legal system, politicians careers exploiting the loopholes of the constitution, etc. All a lie detector test does, is encourage people who lie well. Some, if not most true criminals will replace the 'real' version of the crime in their mind with their invented version which if it solid enough will pass every lie detector test because the criminal thinks its true. So in the end, we have criminals gaming the system and innocent people who are stressed out or protecting their rights under the 5th amendment being considered lairs.

Re:Gaming the system (1)

linzeal (197905) | more than 4 years ago | (#32116616)

Do you have any studies that conclude or even allude to the possibility of gaming a fMRI ? From a layman's point of view from someone who has read over some of the literature it has more to do with the morphological differences of the patient's brain than the psychological/personal profile of the individual.

Re:Gaming the system (1)

Darkness404 (1287218) | more than 4 years ago | (#32116740)

For many criminals, they are not withholding information. They say exactly what they believe so it isn't 'lying' according to the scans. Just like if you asked someone 10 years ago if Pluto was a planet, they would all say yes because they weren't lying even though that information wasn't correct. Same thing with criminals, if they believe it strong enough it isn't a lie and if it gets verified as the truth they can manipulate the lie detector tests and if they are the only witness they will get off the hook easier.

Re:Gaming the system (2, Interesting)

linzeal (197905) | more than 4 years ago | (#32117030)

A fMRI is not a lie detector test it is more of a memory test, so unless the criminal is delusional or has those morphological differences I mentioned in my first post he is not going to get off.

Re:Gaming the system (0)

Anonymous Coward | more than 4 years ago | (#32117060)

Morphological images are captured as part of the fMRI scan, but they aren't the purpose of the scan. fMRI stands for functional MRI. Unlike traditional MRI, the intent is to have some measure of the brain while it functions. So fMRI analyses typically compare activity* during some baseline task and some task of interest. I suppose that for lie detection the baseline might be something like answering the question "Is your name XYZ?", while the task of interest might be something like answering the question, "Did you kill so and so?" You would then look at a region of interest, and see if there's any difference in the neural response between known truthful answers and the ones of interest.

One problem is that the signal to noise ratio is usually very low, so traditional research has relied on averaging over many repeated trials to obtain a solid measure of function-related activity. I'm not sure how that problem is addressed in lie detection. I mean, you can't really expect to ask a person the same question 100 times and elicit the same kind of thought process every time.

Regarding gaming the system: There are definite changes in fMRI activity associated with using different strategies to solve a particular problem. For example, you might be able to do add two numbers by counting on your fingers, by recalling an answer that you've memorized, or by imagining doing the sum as a mental image. No matter how you do it, you'd presumably arrive at the right answer. I'd say that this is an indication that the system can indeed be gamed, because people have control over their own hidden thoughts. Using an fMRI doesn't differ from a traditional lie detector in any deep sense, because you're still trying to detect the physiological changes associated with lying. In fact, you might be better able to control your thoughts, relative to things like how much you're sweating.

*: It just so happens that the blood becomes less oxygenated about six seconds after neural activity in a circumscribed area. Oxygenated blood differs from deoxygenated blood in how it responds to a magnetic field. Since MRI machines can induce and measure fluctuations in a magnetic field, they can indirectly allow analysts to compare neural activity.

Re:Gaming the system (1)

droopus (33472) | more than 4 years ago | (#32116820)

What makes you think "truth" has anything to do with the US justice system?

As any decent (cough) criminal attorney will tell you, it's not whether you did it or not, it's what the prosecutor can prove. And, with Grand Juries (now there's a fair concept...no defense..) indicting anything a prosecutor throws at them, prosecutors know that if they can prove falsehood (not hard) then their conviction rate goes up, they get elected to public office and...

Step 3: Profit!

Re:Gaming the system (0)

Anonymous Coward | more than 4 years ago | (#32117980)

"All that lie detectors will do is encourage criminals to 'game' the system."

Sociopaths always know how to "game" the system. There's plenty of other ways in which fake "lie detectors" are problematic that don't involve professional liars.

How do you fail a lie detector? (1)

Mekkah (1651935) | more than 4 years ago | (#32116628)

If you take it and pass, you are telling the truth, if you take it and fail, why would you take it if you were gonna fail, clearly you are in the 10% of bad results. Win-Win.

Re:How do you fail a lie detector? (1)

OrwellianLurker (1739950) | more than 4 years ago | (#32116756)

If you take it and pass, you are telling the truth, if you take it and fail, why would you take it if you were gonna fail, clearly you are in the 10% of bad results. Win-Win.

It should never be allowed unless the defense permits it.

To many people get (1)

geekoid (135745) | more than 4 years ago | (#32116714)

there science from magazine covers.

Newsflash, It's not really that good. In the best controlled enviroments, you might appraoch 90%. Maybe.

IN an actually court case? in a situation where you are under serious stress? no.

We have learned a lot about the brain in there last 15 year. a tremendous amount, but not enough to understand all the finer chemical interactions going on.

Even if it could be perfect,and it can't be, it would still violate the right to testify against yourself.

They court room is a horrible place for science.
Look at the stupid hysteria regarding toxic mold.

Re:To [sic] many people get (0)

Anonymous Coward | more than 4 years ago | (#32117100)

there science from magazine covers.

Newsflash, It's not really that good. In the best controlled enviroments, you might appraoch 90%. Maybe.

IN an actually court case? in a situation where you are under serious stress? no.

At least the magazine covers can spell and capitalize.

By the way, I love your sig. ~

76%-90% (1)

mevets (322601) | more than 4 years ago | (#32116716)

Without a headlong dive into probability here, isn't 50% the baseline [ie. cointoss]? If so, isn't 75% only 50% better than guessing and 90% only 80% better than guessing? A description of "fMRI lie detection is 50%-80% better than tossing a coin" doesn't seem so impressive.....

Re:76%-90% (1)

blueg3 (192743) | more than 4 years ago | (#32116930)

That's fairly disingenuous math -- you'd be saying that 100% accuracy is only twice as good as flipping a coin. For simple systems (two equally-probable states, like flipping a coin), 50% is a good baseline. It really depends on what their "accuracy" is measuring.

But yes, 75% accuracy is pretty terrible, and 90% accuracy might be good enough to help out an investigator, but not nearly good enough to present as factual.

Re:76%-90% (1)

Firethorn (177587) | more than 4 years ago | (#32117362)

But yes, 75% accuracy is pretty terrible,

For this application. To reach the 99.9% court standard, we generally apply a number of different tests, all of which are less than 99.9% likely to be correct.

I wonder if you stick a guy in the MRI again the next day would you get a true 75% chance to be right this time, or if somebody who can fool it the first time is likely to fool it again?

Heck, what about combining this with the more traditional lie detector tests? Mythbusters actually rated the traditional test better for their limited not very scientific tests. (No fools vs 1 fool).

Re:76%-90% (0)

Anonymous Coward | more than 4 years ago | (#32117246)

If your machine is right 75% of the time, then its only wrong 25% of the time, that's three good readings (hopefully) in every four questions. So far sounds like its 50% better when compared to two in four good readings.

But another way of saying it makes more sense in this case.

That's three good readings to every one error, instead of one good reading every error. So its three times "better" by my math than guessing, using the stated worst case accuracy.
If its accuracy is 90%, then that changes to nine good readings for every error instead of one, or about 9 times better.

Lie Detection (5, Informative)

Adrian Lopez (2615) | more than 4 years ago | (#32116810)

Calling these devices "lie detectors" is misleading at best. Until they invent a machine that can travel back in time and compare the suspect's claims against the facts, there can be no lie detectors. All a lie detector can do is make visible certain physiological responses that are more or less associated with lying. While sometimes these responses are explained by the fact that the person is lying, humans are complex enough that it should never be considered a trustworthy mechanism as far as the law is concerned. Juries are easily influenced by apparently scientific evidence, but lie detection is a questionable science at best and could prejudice juries against the defendant.

There's an episode of Penn & Teller's Bullshit that does a good job of putting the lie to the lie detector.

Re:Lie Detection (2, Interesting)

AtomicDevice (926814) | more than 4 years ago | (#32116990)

Very true.

Even the best "lie detector" could only prove what someone believed or remembered to have happened. Many studies have shown memories to be very open to manipulation, children have been convinced by their doctors that they were raped by their own parents (when they were not). People have been manipulated to believe that certain individuals (who look nothing like the real perpetrators) committed acts of violence against them.

Even without overt manipulation, eyewitness testimony is notoriously unreliable, and there is no reason to believe that even suspects who have not been coerced might come to believe falshoods about their own actions through some combination of internal and external pressure.

Furthermore, as one with personal experience with fMRI data analysis, the though of using fMRI to sentence someone to years in prison or worse is frightening. While over large sample groups certain types of analysis can be reliable, fMRI data is frought with noise, is very low resolution (both spatially and temporally), and due to the huge amount of pre-processing required to get any useful data, would be very vulnerable to manipulation by unscrupulous types (or merely accidental bad analysis by poor technicians or bad software).

Re:Lie Detection (2, Insightful)

Nematode (197503) | more than 4 years ago | (#32117646)

Another problem with "lie detectors," and a good reason that juries rarely ever hear about them, is that juries tend to give them undue deference. You can get a competent defense counsel to present evidence to a jury that they're not reliable, have a lot of false positives, etc etc....and at the end of the day, many jurors will look at it and still think "that's a lot of high-tech sciencey doohickamajigs right there, and this defendant is just trying to talk himself out of scientific proof! I mean, look at those knobs and needles."

It's kind of a good thing that juries are disposed to trust "sciencey" stuff, but not so good when they can't grasp what it really is they're being told about, or what the shortcomings are.

fMRI defintiely works (3, Informative)

Nidi62 (1525137) | more than 4 years ago | (#32116842)

I mean come on, fMRI has even been used to definitively prove that dead fish can think! (http://neuroskeptic.blogspot.com/2009/09/fmri-gets-slap-in-face-with-dead-fish.html) How much more scientific proof for its accuracy do you need?

And he did very well in excluding it (3, Insightful)

arisvega (1414195) | more than 4 years ago | (#32116856)

I wouldn't be sure to trust this brain energy pattern recognition for a verdict- as far as I am concerned, it is much better to cut loose someone that _might_ be guilty, than to convict someone that is not.

By how much could this swing a case? (1)

akkornel (1800252) | more than 4 years ago | (#32116880)

For those who are actually trial lawyers, in the United States, and not just in their heads, I am wondering: How possible is it to change a ruling (from guilty to innocent, or vice versa) on what one person says while on the stand?

Just a side comment... (1)

masterwit (1800118) | more than 4 years ago | (#32116884)

How is this labeled: "Your Rights Online" Can they scan my brain while I'm viewing Slashdot? - they would find that one amusing if such a system existed :)

(Category of 'Science' perhaps better?)

Humans +1 (2, Funny)

tpstigers (1075021) | more than 4 years ago | (#32116918)

This is a huge victory for humanity. What it really means is that the machines cannot tell when we're lying.

Is this dead salmon lying? (2, Informative)

wembley fraggle (78346) | more than 4 years ago | (#32117010)

http://www.wired.com/wiredscience/2009/09/fmrisalmon/ [wired.com]

fMRI is a fairly arcane art- it's nowhere near the "thought detector" most laypeople think it is. The actual practice is rife with the chance to show confirmation bias, given the kind of data filtering that goes on during the process. Check out the link above- scientists were able to show the reaction that a fish had to watching pictures of pleasant situations (babies, puppies, flowers, etc). The fish was dead at the time of the test, however. So, if fMRI can be used to show that a dead salmon has feelings, I'm not likely to trust it for a "lie detector".

Relevant Conduct..ever hear of it? (5, Informative)

droopus (33472) | more than 4 years ago | (#32117014)

While we're on the subject of the "law..."

Anyone ever hear of relevant conduct as the feds consider it? Basically it means that your custody can be affected by behavior for which the charges have been dismissed, or even charges for which you were acquitted.

I kid you not. Here's a true scenario:

A man gets caught with five grams of crack. (FIve year mandatory minimum in the feds, until the crack/powder disparity is corrected.) That's five years for about a sugar packet of rock. But when he is arrested,the cop says" hey you look like the guy that hosed down that McDonald's with an AK47, killing 35 schoolkids!"

Of course, you aren't, you go to trial, and after 30 seconds of deliberation, the jury acquits you of the mass murder charge. But you still go away for the crack.

Here's the kicker: when you go to prison, the Bureau of Prisons (BOP) considers the murders "relevant conduct" and sends you to a very nasty pen, puts extreme violence on your record, and puts a Public Safety Factor on you, because of the "relevant conduct"...of which you were acquitted. Due process? Hah!!

Don't believe me? [washingtontimes.com]

Want the official Government position? [ussc.gov]

The US justice system is a fucking travesty, and unfortunately, you don't realize that till you're neck deep in it.

Be careful out there.

Re:Relevant Conduct..ever hear of it? (2, Informative)

nomadic (141991) | more than 4 years ago | (#32117476)

Actually the story you link to says the sentencing judge can take into account charges the defendant was acquitted on, but implies that the ultimate jail sentence can't be extended past the maximum sentence for the charge they were convicted on.

Re:Relevant Conduct..ever hear of it? (1)

droopus (33472) | more than 4 years ago | (#32117674)

Indeed so, exceeding it would be an Apprendi issue.

But, relevant conduct can be the difference between spending time in a camp or a penitentiary. also, the BOP can slap a "greatest severity offense" Public Safety Factor on you, and that affects your whole life.

Sound fair, for something of which you were found not guilty?

Re:Relevant Conduct..ever hear of it? (1)

osgeek (239988) | more than 4 years ago | (#32117548)

One of the many reasons that I always vote to decrease the power of government. Lower taxes + no deficit spending == less government power.

The Interesting Question is... (3, Interesting)

cheesethegreat (132893) | more than 4 years ago | (#32117034)

Here is the interesting hypothetical:

Assume this lie-detector is right 80% of the time, and that its success/failures are randomly distributed (e.g. not associated with socio-economic background).
Assume also that false conviction rates are at 21% (so only 79% of convictions are correct), and that there is substantial evidence that this is not evenly distributed (e.g. that false convictions are associated with low socio-economic status).

Would you be willing to entirely replace the system of jury trials with trial by lie detector?

Re:The Interesting Question is... (2, Interesting)

izomiac (815208) | more than 4 years ago | (#32118144)

This is why I'm curious as to the appeal of lie detectors in courtrooms. If there was an actual physiological change associated with lying, then the whole concept of lawyers and courtrooms becomes obsolete. Asking "Have you committed a crime?" with confirmation from the lie detector would be all that's required. Heck, police could start doing it to all suspects with portable versions, and drop off the people who failed at the local prison. The only point of a judge would be to determine sentence length, which, with a way to know what actually happened and what the motivations were, would basically lead to that just being a checklist. The suspect knew it was illegal? Add 6 months to the sentence.

Now, returning from our trek in fantasy land, lie detectors cannot work. Some people are terrible liars, but you can tell that in a jury trial. Others legitimately believe what they say to be the truth. "Did you commit a crime?" being a prime example. Serious criminals simply do not think like ordinary folk, a fact that should be obvious since ordinary folk (>98-99%) do not commit serious crimes. OTOH, for the situational crimes, the criminal's perception of reality is what would be detected by the lie detector, which probably has little resemblance to actual reality. Situational criminals don't knowingly do something evil, so they aren't "lying", whereas people who have the capacity to knowingly be evil aren't going to be flustered enough to be detected on a lie detector test.

Classic statistics problem (5, Insightful)

cortesoft (1150075) | more than 4 years ago | (#32117040)

This sort of thinking (that if the accuracy rate is improved enough it will become a valid way of determining someones guilt) shows a fundamental misunderstanding of statistics. It is the same reason blanket drug testing doesn't work and medical screening can sometimes be a bad thing.

Let's imagine for a moment that this lie detector technology has been perfected to a 99.99% accuracy rate. Since the test is so accurate, we decide that whenever a crime is committed, we will just have everyone in the area take the lie-detector test, asking them the question "Did you commit the crime?". Clearly, when someone fails the test, they are 99.99% likely to be the criminal. Right?

Except no. In cases like this (where the average person is much much much more likely to NOT be the criminal, the error rate will overwhelm the actual guilty-rate. If we are testing everybody in an area, then we can suppose that each person we check has an average chance of being the criminal of about 1 in however many we are checking. If this number we are checking is very large, then we are CERTAINLY going to have quite a few people who are found to be guilty on the test but are actually innocent. It will pick out more innocent people than guilty people.

While this sort of statistical phenomena will not take place if we don't giving blanket tests to everyone and limit the test to people who we already believe are very likely to have committed the crime, we as a society have a very bad tendency to not understand the statistics and think we should just give everyone the test and let the results tell us who is guilty. If you doubt this, just look at how many people think we should have a DNA database that everyone needs to join (so we can just run any DNA found at a crime scene against it). This combines the birthday paradox with the statistics I described above to create a situation where we have a very real fear of false convictions, exacerbated by the fact that people who are relying on this evidence (juries) do not realize that even a test with 99.999% accuracy can have a very high false positive rate in these sorts of circumstances.

Check out http://en.wikipedia.org/wiki/Bayes'_theorem#Example_1:_Drug_testing [wikipedia.org] for more info on the math behind this.

Re:Classic statistics problem (2, Informative)

blueg3 (192743) | more than 4 years ago | (#32117120)

Of course, nobody actually suggested just administering a lie detector test to everyone in the area. They only suggested allowing the results of the lie detector to be admitted into evidence.

Re:Classic statistics problem (2, Insightful)

nextekcarl (1402899) | more than 4 years ago | (#32117592)

While I agree with your post I can see the slippery slope the OP is talking about. Where does allowing something like this lead? If it is so good, shouldn't it be mandatory, etc?

Was the witness a dead salmon? (3, Informative)

chill (34294) | more than 4 years ago | (#32117098)

Without proper correction, fMRI has been shown to detect brain activity in a dead fish [blogspot.com] . Next up, trial lawyers.

Why is everyone saying they're relying on it? (1)

TSRX (1129939) | more than 4 years ago | (#32117106)

Why can't the machine's readout just be another piece of evidence that the jury accounts for?

Not If I Can Help It (0)

Anonymous Coward | more than 4 years ago | (#32117130)

at my upcoming Crimes Against Humanity [whitehouse.org] trial . I NEVER lied about Weapons of Mass Demolition.

Yours In Peace,
W

Thoughts vs blood flow (1)

Theril (606664) | more than 4 years ago | (#32117340)

The whole field of brain imaging, be it fMRI, MEG or EEG has way too much methodological problems to be used as anything resembling "mind reading".

One big issue is whether the location of brain activation in fact says anything about the function that it supposedly causes. Even for the "known" localities (Broca's area etc.), the position varies a lot between subjects. Especially for seemingly homogeneous area such as the neocortex this is a big issue. And even for the better known localities, there aren't very clear theoretical accounts about the computation.

Also it's to be proven that the blood oxygen level in certain place of the brain really correlates that well with the computation that's done in that part. And that intensive computation in some part even tells very much about what's really being represented in the brain.

The fMRI (etc) studies are mostly correlational and any significant results usually require many test subjects. The accuracy of the results are vastly overrepresented in the public, and unfortunately often also in recent cognitive science/neurology/psychology.

http://en.wikipedia.org/wiki/Functional_neuroimaging#Critique_and_careful_interpretation [wikipedia.org]

Busted Already on Mythbusters (2, Interesting)

2obvious4u (871996) | more than 4 years ago | (#32117386)

Grant beat it by doing complex math in his head while they asked him questions. Couldn't find a video, but found summary. [mythbustersresults.com]

UPenn Goat (1)

tanderson92 (1636327) | more than 4 years ago | (#32117690)

This case illustrates very clearly the "Goat Problem", where an open and defined process is perhaps considered more important than the right answer. The "Goat Problem" is actually an ancient Chinese legend and the University of Pennsylvania's law school has this as their mascot[1]. Is it the right call? Perhaps, for accepting this kind of technology is a slippery slope down to the all-knowing goat. [1]: http://www.jstor.org/pss/3311295 [jstor.org]

Oh please (0)

Anonymous Coward | more than 4 years ago | (#32117790)

"fMRI has been shown to be 76-90% accurate"

Bullshit.

From the article-

"The estimates
from the two available laboratory datasets, ranging between 76 and over 90%
(Davatzikos et al., 2005; Kozel et al., 2005; Langleben et al., 2005), should be considered
a strong indication for more extensive testing rather than a focus of debate on whether
the upper limits of this range is sufficient for court evidence. In practice, the accuracy
of fMRI-based lie detection is likely to vary with questionnaire-type, countermeasures,
and other, hitherto unexplored variables. Though individual variability may diminish the
value of a standard fMRI template in clinical lie detection, individual response patterns
to a standard task could be used as a within-participant control prior to every liedetection
test (Hakun, Ruparel et al., in press)."

That does not say "accurate" and that we should put it in place immediately.

Load More Comments
Slashdot Login

Need an Account?

Forgot your password?