A poston Psychology Today says that people in a study can tell criminals by sight from non-criminals.
(The link has the details).
But the study doesn’t seem to go into detail on why. Is the study true? Can people really tell criminals by sight? I tried the test but just barely beat 50%.
Bernie Madoff?
Maybe there are subtle clues like skin color or tattoos or facial hair or accents that make it “obvious”. Or… If you saw Mel Gibsons mug shot, would you say “criminal”?
I am very skeptical. And strike another blow against the Jury system.
I’m going to call BS on that. I’m pretty sure that very few of the bad guys and gals in the movies that look the part on screen have much more than assault or DUI on their rap sheets in spite of being a serial killer in the current blockbuster.
The link provided by the OP showed pics of non-famous white males without scars, tattoos, or facial hair.
Physiognomy has recently resurfaced as a number of studies have found that people have the ability to make accurate predictions about other people based solely on their facial appearance.
Certainly not Politically Correct, but true nonetheless:
I have no opinion on the paper cited but Satoshi Kanazawa is an embarrassment to the field of evolutionary psychology and anything he writes about should be taken with a large grain of salt.
Assuming the test presented photos in a controlled manner, and that your score is typical of most people taking the test, then the answer is probably yes, people can tell criminals from non-criminals, within some margin of error.
I tend to agree and Razib Khan of GNXP has been saying this about Kanazawa for years. However, in this particular instance, he appears to be summarizing the science accurately. You can assess criminality, openness to experience, homosexuality, agreeableness, religiosity, intelligence, extraversion, neuroticism, likability, and conscientiousness to some degree simply by looking at facial photographs.
Well first of all EVERY ADULT is technically a criminal, everyone alive has broken laws 
Second what is their definition of a criminal? What about pot smokers? Underage drinkers? Someone who skinny dipped decades ago? A career bank robber? We need specifics.
But I will concede that criminal subcultures often intentionally adopt facial expressions and body language that is instantly recognized by those used to it, but again this is intentional on their part not a subconcious tell. Someone is going to ask for a cite now
Why don’t you read the article, then? It explains exactly what crimes the men in the photos have committed.
Read the first link in the OP. Those questions are answered, with more details besides.
This claim makes it sound like what people are really seeing is psychopaths:
A man who is just as likely to sell some weed as murder someone is pretty abnormal even among criminals.
GROAN
I made a poll out of this, so we can check for ourselves.
The test uses pictures of convicted criminals and people who haven’t been tried or convicted. The prisoners might look “different” because they are showing the stress of having been caught, tried, convicted and imprisoned.
The other guys might look (and feel) innocent because they have not yet been caught or tried. Or maybe they have not yet committed any crimes.
The most violent, criminal minded people that I know can turn on the charm and could con anyone. If they get cornered, they put on an entirely different face!!
Hmm, appearance is one thing, I spotted the violent offenders pretty easily - except for one.There is a certain attidue in which they are looking at the photographer, a tightness around the mouth and eyes.
I certainly would not want to put money on it. Any convicted offender will likely have an undiscovered criminal past so if there were any truth to this then the idead of identifying the type of offence is going to be a very hit and miss affair - there is just too much noise in the background.
There is also a likelihood that an offender may have a present conviction for one offence, but also a history of other offences. This particular offending age group will almost certainly have a long history, since reoffending in males 13-20 is well over 90% in the UK.
Those involved with hard drugs will generally have certain aspects, hair, teeth, fingernals that give it away.
I’ve seen enough of them to know that there is no reliable appearance indicator though, body language does reveal a great deal more when you are dealing with career criminals.
Other indicators are red flags, such as very poor literacy and numeracy and concrete thinking patterns - these are especially pronounced in offenders.
What does that mean? They don’t change their minds?
I’m a huge believer in the validity of evolutionary psychology which is why I come down especially hard on those who do it so poorly.
[ol]
[li]Satoshi summarizes the results of a single paper and presents it as truth which is is a shoddy but common scientific journalism practice. Often, replications of outlier papers find significantly less effect sizes due to regression to the mean and survivorship bias. By vastly overstating the confidence of the result, he is misleading readers.[/li][li]The studies were done with 44 & 36 participants respectively which is a relatively paltry sample.[/li][li]The studies were published in “The Journal of Social, Evolutionary, and Cultural Psychology” which is a 5 year old, online journal of which Satoshi is an Associate Editor, and doesn’t even have an impact factor. IOW, I’m not even able to assess the validity of peer review process that this paper went through.[/li][li]The effect size found in the study was that non-criminals were rated as a 4 on a 1-7 scale of criminality and criminals were rated as a 4.5 While the results were statistically significant, the actual magnitude of the different is rather small. In the study, people ranked 72% of criminals as criminals but also 53% of non-criminals as criminals.[/li][li]Correlation does not imply causation and the study does not discount other, plausible explanations. For example, we may just be good at detecting people’s socio-economic class due to facial features, hairstyle and dress and SES is correlated with criminality. Also, while the study tries to control for mugshot vs non-mugshot, I don’t think they did enough to exclude this as an explanation.[/li][li]Moving on from the paper itself, Satoshi makes a number of huge, unsupported leaps. His assertion that criminals do not specialize is a laughably broad overreach from the data he actually presents and his rapist hypothesis is the worst kind of just-so story there is.[/li][li]Satoshi’s assertion that the participants could not tell what type of criminal is unsupported by the data. The study, in actuality did not have enough distinguishing range to tell whether this assertion is true or not.[/li][li]Satoshi consistently demonstrates a misunderstanding of basic statistical methodology, experiment design and scientific literacy that is embarrassing.[/li][/ol]
In short, the paper is perfectly fine for what it is, it may very well be that this hypothesis is true but a single paper is far from enough to establish this either way. Satoshi’s extrapolations from this paper are dangerously false and should be denounced with vehemence.
Concrete thinkers focus on experiences, abstract thinkers focus on ideas. I don’t know if that makes any sense, or if I’m explaining it well, but abstract thinkers are more likely to generalize and take skills and observations from one situation and apply them to others.
There’s a Chinese folktale in which a farmer believes his neighbor’s son has stolen his hoe. Looking at the boy, he says to himself, “He looks so guilty. He will grow up to be a dangerous robber.”
Before he can confront the child, the farmer remembers that he’d left the hoe in the field the night before. He finds it. It wasn’t stolen at all. He looks at the boy again, and says, “The boy looks totally innocent. He will grow into an honest man.” And yet the boy’s expression hadn’t changed at all.