Posted in Science Etcetera on October 2nd, 2009
Ardipithecus ramidus, “Ardi” for short, is the human race’s new oldest potential ancestor at 4.4 million years. They walked on two legs, but were also well adapted to life in the trees (HT Rachelle).
Ardipithecus ramidus
Although parents teach their kids that even white lies are not acceptable, 78 percent of parents lie to their kids because it’s easier.
More than half of babies born today should live to see the year 2100.
The discovery that named cows give more milk than unnamed, empty beer bottles are better for smashing over someone’s head than full, and a brassiere that can be converted into two gas masks are a few of the winners of the 2009 Ignoble Awards.
Genes that are beneficial to females, but detrimental to males in cichlid fish create an interesting conflict of the sexes in the species.
A Gene that Camoflauges Females (Left) Makes the Males Stand Out (Right)
Researchers find that water shortages in Southeast North America are the result of growing populations, not global warming.
The flu-fighting drug Tamiflu is being found in rivers downstream from sewage treatment plants in Japan, leading to fears that, as the drug is consumed by birds, resistant strains of bird flu will emerge.
People possessing a gun are 4.5 times more likely to be shot during an assault than those without one.
Magic Tricks/Science Facts
Leave a Comment
Does the gun study seem weird to you? The selected ~650 people *who had already been shot* and then determined that those who had a gun were more likely to have been shot. But where is the cause & effect? Perhaps these people lived in an area where they felt they might get shot, and got a gun because of that.
I don’t understand the control — they called up other random people and asked them if they had a gun. And then when they didn’t, this meant they were less likely to get shot?
Seems fishy to me that their original sample set is 100% people who were shot.
I really find no fault with the gun study. There probably is validity to their findings, though they don’t claim a reason.
My personal opinion is that most likely those who own guns aren’t necessarily willing to use them on another person, in which case hesitation is dangerous.
That’s not what I hoped you’d say Chriggy (RE: finding no fault).
But yes, hesitation bad. The Samurai were definitely on to something with “once you draw your blade, you must draw blood”.
It seems like if you were aimed at a door you could just shoot someone when they opened it, but I’m thinking most people are trying to talk it out or something??
I’d love to see a more detailed study of home invasions.
I think it is flawed. They are comparing people who got shot and seeing if they have a gun to people who had not been shot at all. Which doesn’t even make sense to me. Please explain that.
What they really need to do is take a random set of people, and look at which group was shot more.
Not everyone who thwarts someone with a gun is going to report it to the police.
Hell, if I shot someone, and he ran away, I might not even call it in, because I don’t want a police report to be used against me in a liability suit. I would assume he would assume I called the police and not come back because I’d already shot him the first time. It’s a tough call.
This also cannot measure the effect of a burglar deciding not to enter your place in the first place (and possibly shoot you) who is deterred by a place with high gun ownership.
Instead, since those people didn’t get shot, they aren’t counted in group A (but are counted in the control group?!).
brain..melting..
^by “place with high gun ownership”, I mean “neighborhood with high gun ownership”.
Basically, if someone breaks in and you point a gun at them and get shot, you get counted in the first group. But if you don’t get shot, then you don’t necessarily get counted in the 2nd group. I’m sure this can be adjusted for, but to my knee-jerk analysis it seems like it would mess things up.
First of all, the study makes no claims about cause and effect.
Second of all, I don’t see where home invasion plays into this, since they only talk about assault in general.
Though the main moral I’m drawing out of this is the modern samurai moral: If you pull out a gun, you better be willing and able to pull the trigger.
Since Clint brought this to my attention…
Well, the article is a cut-n-paste from the UPenn press release. I tried to find the actual study, but the full study isn’t available w/o a AJPH subscription. The abstract is less informative than the article.
Historically, medical studies of gun violence have not had a good track record of producing useful results. That doesn’t mean this one is flawed, but I do see some possible issues.
1) This type of study only looks at correlation, not causation. The classic example – people with diet soda in their refrigerator are 4.5x more likely to be medically obese (made up stat). In this case, there is no way for the study methodology to identify if possession of the gun caused an increase in the incidence of being shot OR if people who have a higher incidence of being shot (as a population, say gang members) are motivated to a higher rate of gun possession.
2) Their two population samples are measured using different techniques. The gun injured were measured by official reports; the random controls were self-reporting in a survey. These two methods have different biases for reporting possession of a gun at the time. For example, a random citizen is much more likely to lie about possessing a gun (esp an illegal one) than a police officer or ME is about the discovery of a gun at a crime scene. Maybe the study tries to address these biases? Won’t know until I can see it.
3) The spurious comparison to smoking/lung cancer and drinking/car crashes completely ignores the fact that gun possession and incident rate of assault has a plausible positive feedback system at play. No one takes up smoking because they hear about a lot of people being sick w/ cancer and no one (well, we hope) drives drunk more often because they hear about a lot of drunken car crashes. (The whole idea of anti-smoking and anti-drinking PSAs is that there is a negative feedback cycle to these relationships.) The opposite is likely true for gun possession – when people feel at risk to being victimized by violence, gun possession MAY increase. I’m not a statistician, so I can’t validate whether that positive feedback cycle exists, but it’s plausible. Comparing a negative feedback cycle to a positive feedback cycle is misleading in the context of a press release like this.
Anyway, my (more than) $0.02.
Cheers,
Daniel
Follow-up:
For a rather intensive analysis of how “public health” style gun violence studies have issues, you can check out this 2002 article in Reason (or just google Gary Kleck)…
http://reason.com/archives/1997/04/01/public-health-pot-shots/
No idea why I thought that was from 2002, it’s obviously from 1997. Must have been thinking of one of the other references I was reviewing. My bad!
Good stuff, Daniel! Very interesting.