Brain Scans Predict Which Criminals Are More Likely To Re-offend

ananyo writes "In a twist that evokes the dystopian science fiction of writer Philip K. Dick, neuroscientists have found a way to predict whether convicted felons are likely to commit crimes again from looking at their brain scans. Convicts showing low activity in a brain region associated with decision-making and action are more likely to be arrested again, and sooner. The researchers studied a group of 96 male prisoners just before their release. They used functional magnetic resonance imaging (fMRI) to scan the prisoners' brains during computer tasks in which subjects had to make quick decisions and inhibit impulsive reactions. The scans focused on activity in a section of the anterior cingulate cortex (ACC), a small region in the front of the brain involved in motor control and executive functioning. The researchers then followed the ex-convicts for four years to see how they fared. Among the subjects of the study, men who had lower ACC activity during the quick-decision tasks were more likely to be arrested again after getting out of prison, even after the researchers accounted for other risk factors such as age, drug and alcohol abuse and psychopathic traits."

Comments

Good technology

[rmdashrf]

Now let's first use it on our politicians.

Re:Good technology

[hedwards]

We have the power, it's just that when half of Americans vote for people promising to bring the government to the knees, you don't wind up with the best or the brightest being elected.

Which is strange, I would have thought voting for people looking to screw up the government would be just the ticket for effective and useful governance. Who'dathunk.

Re:Good technology

[Sulphur]

Now let's first use it on our politicians.

Why? We already know.

To calibrate the tests.

Re:Good technology

[gnasher719]

Now let's first use it on our politicians.

Totally misunderstanding what this is about.

There's a bit in your brain that tells you "don't do this, this is a bad idea" when you want to do something that is a bad idea. That bit usually stops you from getting into trouble. It will stop you from smacking your boss in the face if he upsets you, which tends to be a bad idea. It will stop you from smashing a car window and grabbing things on the seat. It will stop you from doing things that hurt you (in the end), including any badly planned crime. People where this bit of the brain is underdeveloped tend to do stupid things, including re-offending after getting out of jail and getting caught.

On the other hand, criminals of any kind who carefully plan what they are doing are not affected by this. They also tend to get caught less often.

Re:Good technology

[isorox]

> We have the power, it's just that when half of Americans vote for people promising to bring the government to the knees, you don't wind up with the best or the brightest being elected.

Or half the people hopelessly defending a corrupted system of horizontal of checks and balances government has the capabilities to heal themselves. The only solution to that is the vertical check: nullifying government when necessary.

When you have 95% of the country always voting for their team, based on the assertion that the other team is wrong, nothing changes.

Re:Good technology

[hedwards]

Considering that the GOP has promised specifically to burn the country down to make President Obama a 1 term President and that those politicians have all been reelected, I'd suggest that there is indeed a substantial number of voters that are in fact voting wrong.

If you really think the parties are identical, then you haven't been paying attention. Yes, on a few issues there is way too much agreement, but if you seriously can't see a difference on issues of taxation and social issues, you're one of those "low information voters" that I keep hearing about.

It's not a case where there's some debate, the GOP has been promising specifically to destroy the government.

Re:Good technology

[hedwards]

Not really, you bring spending back in line by a combination of spending cuts and tax increases. The party that's knows that is coincidentally not the party that's promising to bring the government to its knees.

As for 2007 levels, are you fucking serious? We had massive spending during the 2001-2007 period and no tax increases to pay for it. Cutting spending to that level would be ridiculous as there was a ton of waste at that point. To get things to sane levels, we need to have a combination of tax cuts and tax increases.

Just because you have a lot of welfare states that don't want to accept cuts to their things, doesn't make it any more true. These politicians aren't promising cuts to bring things back under control, they're trying to get cuts to kill programs they don't like, even though their districts aren't actually contributint their fair shares and the spending that they are OK with is of no use to most people.

Beware of sampling bias

[SirGarlon]

Does the more impulsive decision-making mean they're more likely to commit new crimes, or simply more likely to get caught?

Brain discrimination

[i kan reed]

It's not illegal to discriminate against people on the basis of their brain activity. Should it be? Can you judge someone on the basis of their biology? Is it really that person's fault anymore if a part of their body predelects them to wrongdoing? Where does liability start? Can you fix people? Should you?

Too many questions about really understanding the brain that our primitive moral system could begin to address.

Re:Targeted Rehab or Targeted Parole

[characterZer0]

You use this test as an excuse to keep certain people in jail for political reasons.

Re:Targeted Rehab or Targeted Parole

[geekmux]

You use this test as an excuse to keep certain people in jail for political reasons.

So, let me get this straight...you're going to continue to incarcerate me, for something I might do in the future?

Believe me, I'm not questioning whether this would actually happen or not. We've proven corruption knows no bounds.

I'm merely pointing out the "minor" issues with this concept, regardless of where our Rights have dissolved away.

Statistically speaking, a lot

[Lucas123]

According to a 2011 Pew study, more than 40 percent of ex-cons commit crimes within three years of their release and wind up back behind bars. As reported on BBC Radio in 2005, the recidivism rates for released prisoners in the U.S. is 60% compared with 50% in the United Kingdom. According to the Bureau of Justice Statistics of nearly 300,000 prisoners released in 15 states in 1994, 67.5% were rearrested within 3 years. A study of prisoners released in 1983 estimated 62.5%. In general, U.S. prisons offer very little to inmates that would keep them from repeating crimes once they're released. Perhaps we should rethink this strategy?

Re:Statistically speaking, a lot

[h4rr4r]

We also make it near impossible for those recently incarcerated to find work. Just try to get a non-shit job with a felony record. No wonder they go back to crime, you have to pay the bills somehow.

Re:Statistically speaking, a lot

[gnasher719]

Of the two strategies, the second seems better backed by numbers. The sharp dip in the crime rate from the later 1990s through the 2000s is thought to be due to much longer sentences being handed down to violent criminals starting in the late 1980s-early 1990s. Incarceration costs are up, but public harm is down. Spend money, get benefit.

I read it was due to abortion becoming much more acceptable, which led to fewer kids being born into a bad family situation that would eventually lead to criminal behaviour.

Re:Block brain scanners by memorizing a host file

[dyingtolive]

Speaking of people with lower ACC activity..

Re:They were all caught and imprisoned initially

[SirGarlon]

They're measuring re-arrest, not recidivism (committing new crimes). There is a difference: not all who commit crimes get arrested, and not all who get arrested have committed a crime. One of the key principles in _How to Lie with Statistics_ (excellent book) was the following:

1. Discover that you don't have data to support a desire finding (in this case, that we can predict who will re-commit crimes)
2. Provide data that supports some other finding instead (in this case, re-arrest rate)
3. Pretend there is no difference between your desired finding and your actual finding

4. ...

5. Profit!