Thursday, February 28, 2013


Okay folks, time to get a little wonky regarding self-defense against violence.  Unless you have been living under a rock (or perhaps hiding out from the law after killing your neighbor) you are aware about the current hotly debated struggle between Second Amendment advocates and Gun Control zealots.

     In testimony to the Senate Judiciary Committee last week, the American Academy of Pediatrics cited some statistics.  You can read the entire testimony (7 page PDF) here

          "Individuals possessing a firearm are more than four times more likely to be shot during an assault than those who do not own one."
     This is stated as if it is an unequivocal fact.  So, let's dig into this a little deeper.  Tracking this statement back to the source, we find that the AAP first refers to their own Policy Statement on gun violence, entitled Firearm-Related Injuries Affecting the Pediatric Population. in which they state:

          "Evidence from Philadelphia suggests that firearm possession increases the risk of being shot in an assault. In a carefully conducted case-control study, Branas and colleagues found that people possessing a gun were more than 4 times more likely to be shot in an assault than those not in possession of a firearm."
     Here it is stated with seemingly less confidence.  Let's dig a little deeper.  If we track this back to the cited article we find a 2009 study entitled Investigating the Link Between Gun Possession and Gun Assault

     The investigators looked at 677 shooting incidents in Philadelphia and compared those individuals to a control group (matched for age, sex and race) of Philly residents who had not been assaulted.

     Of the 677 shooting victims, 5.92% (or a total of 40) possessed a gun at the time of the assault.  Simple math tells us that 637 (or 94%) of the shooting victims therefore did not possess a gun.  This can be restated as "16 shooting victims were not carrying a gun for every one that was carrying".   (Incidentally, of the 684 non-victims in the control group, 7.16% (a total of 49) possessed a gun at a comparable point in time.)        Analysis of other variables indicated statistically significant difference between the group who had been shot and the group that had not been shot in the following characteristics:  

         "At the time of shooting, case participants were also significantly more often involved with alcohol and drugs, outdoors, and closer to areas where more Blacks, Hispanics, and unemployed individuals resided. Case participants were also more likely to be located in areas with less income and more illicit drug trafficking." 
       From these data the investigators concluded that an individual is 4 times more likely to be shot if they possess a gun than if they don't possess a gun.  A thoughtful and skeptical person might now ask themselves, "How did they reach this astonishing conclusion?"  Simple, actually.  They performed some freakishly complicated statistical calisthenics (regression analysis) and mathematically cancelled out all of the other characteristics (such as hanging around out of doors around crime-ridden areas in destitute parts of town) as being "confounding variables".  And the so-called "control" group had absolutely nothing to do with their bizarre conclusions.        This is a fine example of utter junk science, and yet we find a supposedly authoritative advocacy group such as the American Academy of Pediatrics, citing this study as a justification for recommending restriction of the gun rights of law-abiding citizens.  This goes well beyond sloppy analysis of data and enters the realm of willfully misleading and despicable behavior.

  Shame on them!    

(In 2009, Gary Kleck of Gun Owners of America did a nice discussion of the Philadelphia study here.)


  1. So having not looked over the "studies" and not inclined to do so, because these types of studies are usually predestined to prove the point that the studiers are trying to make.

    I do have one question however. If in fact a person who faces an assailant and is able to shoot said assailant, and that same assailant was assailing with a firearm, does that assailant become one of the statistics "Individuals possessing a firearm are more than four times more likely to be shot during an assault than those who do not own one." Obviously the assailant was in possession of a firearm and said assailant was shot, maybe even killed wouldn't that qualify as being him/her as a person in possession of a firearm being shot ?

  2. Hey Dean-0,

    You would think so, wouldn't you. It just makes my head explode when I see willful misrepresentation of facts to serve an agenda, especially when it comes from an organization that should be objective, rational and scientific in its analysis. I posted this as just one example from one small area of what rational, decent people are up against in this country.

  3. There really is no limit to what Second Amendment crazies will do to disparage studies that disagree with their preconceived notions of the sanctity of firearms. Maybe you should take another class in statistics before pretending to know what a regression analysis means.

    1. Thanks professor. You may be right. Allow me to make a correction. They performed a regression analysis and purposefully omitted several key variables from that analysis. Are you trying to make the case that this study makes any sense or are you just trying to throw out a distraction? My point remains valid.

      And if you come back, leave the ad hominems at home.