Justice Not As Blind As Previously Thought 256
NotSoHeavyD3 writes "I doubt this is much of a surprise but apparently Cornell University did a study that seems to show you're more likely to get convicted if you're ugly. From the article: 'According to a Cornell University study, unattractive defendants are 22 percent more likely to be convicted than good-looking ones. And the unattractive also get slapped with harsher sentences — an average of 22 months longer in prison.'"
Re:Correlation is not causation (Score:5, Informative)
Read TFA please. The study was done with students at Cornell, who were asked to give their verdict after reading the closing arguments from the trial. The pictures of ugly and non-ugly people were inserted into these case studies, so that the same facts were presented as though they were about two different people.
The ugly might very well commit more crimes, but this study eliminates that as a confounding factor.
Re:Did they adjust for meth and crack use? (Score:5, Informative)
According to TFA, the researchers used theoretical juries of undergrads, and merely swapped the photo associated with them. I haven't seen the photos myself, but researchers usually use a distribution of attractive or unattractive photos that don't include "disfigured in a bar fight" and "barely cognizant heroin addict."
Of course, being a theoretical study on paper does mean that real-world influences could be much lower... or higher. For example, any signs of remorse in the courtroom, performance on the stand, etc might be much more significant to the overall judgement process. Or maybe the juries take real courtroom activity more seriously. Or maybe undergrads all just need to get laid.
Re:Lemme be the first... (Score:3, Informative)
Hilarious. You deliver a mild rant about people spouting off "correlation!=causation" without so much as reading the article to see if the researches took that into account, all the while your post makes it obvious that you yourself didn't so much as read the article and yet your get modded insightful and informative. Sometimes I think the mods don't even try.
The researchers didn't use real court room data, they created mock criminals, attaching different pictures to the same information about the case and compared the results from online volunteer juries. In other words, there are no actual criminals involved and both of your points are... well, pointless.
Re:Did they adjust for meth and crack use? (Score:3, Informative)
The result isn't actually that surprising. A similar result was mentioned in the book Blink [amazon.com] that was popular a while ago. The study mentioned in Blink showed that juries were very sensitive to the race of the accused, and that black defendants had a much higher conviction rate, even with very similar evidence.
For a beautiful example of how it works subconsciously, have a look at the Implicit Association Tests [harvard.edu] from Harvard.
Re:Female teachers and students (Score:3, Informative)
There are lots of studies on this.
Here is just one of them.
Holland, etc al. 2009
http://www.scientificjournals.org/journals2009/articles/1439.pdf [scientificjournals.org]
Female average sentence 11.60 years
Male average sentence 28.05 years
And if the victim and perpetrators are both male, the mean sentence is over 45 years. (this usually includes very long probationary periods ~20 years)
The mean sentence for first degree murder is currently around 34 years.
Re:Lemme be the first... (Score:4, Informative)
But since correlation isn't causation you'd have no idea as to whether or not the commenter is a "fucking moron."
If someone tags a story with "correlation!=causation" (c!=c) or attempts to use that phrase as an attack on the story's premise, that is prima facie evidence that the person in question is, indeed, a fucking moron (FM). Not all FMs parrot c!=c (PCC) at every opportunity, of course, but only FMs do so. The correlation between FM-ness and c!=c parroting is therefore quite large, certainly large enough to be significant.
In the presence of a significant correlation between X and Y, there are three causal possibilities: either X causes Y, Y causes X, or there exists a third factor, Z, which causes both X and Y. Let X = PCC and Y = FM, and break down the possibilities:
X ==> Y: PCC makes you an FM. This seems unlikely. We could test it, of course, but there's no causal mechanism.
Z ==> X,Y: certainly possible, but the universe of possible Zs is pretty large. No need to complicate the hypothesis -- Occam's Razor and all that -- when ...
Y ==> X: FMs are incapable of understanding statistics, since they're, well, FMs. And PCC depends on a profound lack of understanding of statistics. Ta-da! There's your causal mechanism and your significant correlation.
The remainder of the problem is left as an exercise for the reader. If you have trouble, there's a hint in my .sig, or see the TA during office hours.
Re:Correlation is not causation (Score:3, Informative)
I don't know if that was unintentionally stupid or intentionally funny.
http://www.merriam-webster.com/dictionary/butte [merriam-webster.com]