Take this test to see if you can tell a criminal from a law-abiding citizen.

Reminds me of the Blackadder episode "The Witchsmeller Pursuivant":

"Its the 'air, he's guilty!"

WalterMitty wrote:Reminds me of the Blackadder episode "The Witchsmeller Pursuivant":

"Its the 'air, he's guilty!"

But here in the quiz you can actually check whether you sniffed it right.

There was another one a couple of years ago showing the more aggressive ice hockey players tended to have wider faces, reflecting higher testosterone. http://www.economist.com/node/11959198

This recent paper about facial symmetry and socioeconomic status is interesting too.

http://www.sciencedirect.com/science/ar ... 7X11000773

30% correct.
Second try I got 70% correct, but three people were also in my first try.

Got 80% on my first try.
Basically my criteria was "happy face = not a criminal"
The rationale is that the photos of criminals are probably mugshots, and that criminals aren't happy when caught

70%

Fallible wrote:70%

So was Lombroso right?

I was going to say that this is guff, then I got 100% right on the first try!

But I still think it's guff!

mizvekov wrote:Got 80% on my first try.
Basically my criteria was "happy face = not a criminal"
The rationale is that the photos of criminals are probably mugshots, and that criminals aren't happy when caught

similar logic on second try but still caught a smiling rapist!.

everyone looks criminal, weird asymmetric faces.

Biowatch wrote:There was another one a couple of years ago showing the more aggressive ice hockey players tended to have wider faces, reflecting higher testosterone. http://www.economist.com/node/11959198

This recent paper about facial symmetry and socioeconomic status is interesting too.

http://www.sciencedirect.com/science/ar ... 7X11000773

I will correlate this with my scale later to the photos given, but the correlation factor is not given. When I am free of course.

This might help a lot in crime investigation, keeping probability on the side of law.

Spearthrower wrote:I was going to say that this is guff, then I got 100% right on the first try!

But I still think it's guff!

You think that you got a 100% by chance?

Dickens wrote:

Spearthrower wrote:I was going to say that this is guff, then I got 100% right on the first try!

But I still think it's guff!

You think that you got a 100% by chance?

Pretty much, yes. Of course, I did look for clues that would suggest someone having been arrested, as in 'sour expression', but as it's a random selection, i doubt I'd score 100% again.

Just to check that, I did it again and got 50% - one of the faces was the same as my first attempt, but this time I decided incorrectly he was a criminal.

There's an obvious test to be run here. Have entirely blank mannequin faces and maintain the binary 'criminal' / 'not criminal' selections, and see how well people do. My expectation is that there's no statistical significance to this when it's just a binary selection.

There was a period about 30 years ago when customs officials relied heavily on 'experience' and the belief they could detect smugglers. This was widely believed. Then some scientists came along and spoiled the party, setting up real life scenarios with smugglers, to see how well the officials performed at detecting them.

The result was indistinguishable from random chance. In fact, it appeared that the main criterion by which officials decided someone was a likely smuggler was prejudice. Skin colour. hair style, clothes etc. Yet a clean cut man in a suit is just as likely to be a smuggler. Certainly, if I was a smuggler and I knew the officials were relying on their 'instinct', I would dress that way, and get through without any suspicion.

Needless to say, customs no longer rely on 'experience'. I believe that random checks are more the order of the day.

One guy claims to get a 90% result using Edenic Phrenology

http://excavatingeden.com/2012/09/09/ap ... -not-quiz/

80% on my first attempt, what do I win?

HughMcB wrote:80% on my first attempt, what do I win?

Have you checked out "Edenic Phrenology"? Impressive stuff.

Dickens wrote:

Spearthrower wrote:I was going to say that this is guff, then I got 100% right on the first try!

But I still think it's guff!

You think that you got a 100% by chance?

Be aware that even if there was some valid process behind the test you linked to, the authors of the study (where the pictures are taken from) go to some lengths to explain why Lombroso's theory was "wrong and fallacious". So trying to link this finding to Lombroso is baffling.

Secondly, the pictures in this study weren't blindly selected, so whilst searching for pictures of "rapists" or "murderers" who fell within their search criteria, it's entirely possible that the selector picked the image that looked more "criminal-y" than another. This would explain why nobody can distinguish between different types of criminals, and can't even distinguish between violent and non-violent criminals. In other words, it's plausible that biological factors that increase aggression could lead to that person being more likely to commit an assault and for those biological factors to have an observable physiological effect, however, it seems much less likely to me that there is a general "criminal gene" which creates observable characteristics and can equally cause someone to become either a murderer, or fraudster, or rapist, or thief, etc.

This is why, I think, the authors of the study try to make it clear that they aren't discussing the Lombrosian concept of the "born criminal", and instead are only discussing the accuracy of stereotypes. This result is confounded by (what seems to be) poor randomisation of sample stimuli, but if it proves to be a valid result, it only tells us that criminals look different from non-criminals. This could be explained by tired eyes, more wrinkles, worse skin, etc, which are characteristics common among people who are a bit down on their luck, struggling with money issues, holding down a job, etc - i.e. the sociological factors that lead to a person being more likely to become a criminal. This would explain why people cannot distinguish between various forms of crime, and avoids the unparsimonious assumption that there is a kind of "criminal gene".

Lance wrote:There was a period about 30 years ago when customs officials relied heavily on 'experience' and the belief they could detect smugglers. This was widely believed. Then some scientists came along and spoiled the party, setting up real life scenarios with smugglers, to see how well the officials performed at detecting them.

The result was indistinguishable from random chance.

What you call real life scenarios are in fact fake scenarios. Suppose that the officials could detect a smuggler because he is nervous. Obviously, you "real life" smugglers are not going to be nervous since they are not risking jail.

Lance wrote: In fact, it appeared that the main criterion by which officials decided someone was a likely smuggler was prejudice. Skin colour. hair style, clothes etc. Yet a clean cut man in a suit is just as likely to be a smuggler. Certainly, if I was a smuggler and I knew the officials were relying on their 'instinct', I would dress that way, and get through without any suspicion.

Needless to say, customs no longer rely on 'experience'. I believe that random checks are more the order of the day.

Do you suggest that crime rates do not depend on skin color? I don't know about smuggling, but black commit 7-8 times more murders per capita than whites. This is about as much difference as between men and women.

http://www.colorofcrime.com/

Suppose that you perform an act of an extreme prejudice and arrest all black men. Assuming that 15% of population is black and 85% - white you will arrest only 7.5% of the population. But the arrested will include a half of all murderers. This will well above statistically random result.