Kellerman did two, ahem, "studies." IIRC: In the first, he looked at just deaths and got "43 times" and in the second, he looked at deaths and injuries and got "22 times." Both were small scale studies (in Seattle and Nashville, IIRC).Used to be 43 times more likely .... what's changed?
Both excluded all forms of self protection with a gun that didn't end in a bullet entering a human body. Given that the vast majority of self protection incidents do not involve any shots (and still more involve warning shots or misses), he excludes the majority of self defense incidents. It's impossible to draw conclusions about the danger/benefit of self defense with a gun when excluding the majority of self defense incidents. The fatal flaw with Kellerman's work is that it assumes that self defense with a gun requires actually putting a bullet into someone.
Further, Kellerman also looked at only houses where shooting occurred, which doesn't give a real picture of the danger of owning a gun -- what about the millions of homes with guns where no shootings occur (including those where self-defense with a gun occurred with no shots fired). As flawed as Kellerman’s studies are, it is even more wrong for anti’s to cite his bogus numbers as some sort of indication of the dangers of gun ownership.