In article <393jq3$jpv@sndsu1.sinet.slb.com> dcd@se.houston.geoquest.slb.com (Dan Day) writes:
> In article <LAMBERT.94Oct28084049@silver.cs.umanitoba.ca> lambert@silver.cs.umanitoba.ca (Tim Lambert) writes:
> >> >Either you did not read the article carefully or you are wilfully
> >> >misrepresenting it. The study found that having a gun in the home was
> >> >not associated with any increased risk of non-gun homicide, only with
> >> >gun homicide.
> Note the above. Tim claims that the increased risk was *only* gun
> homicide. Now keep reading and watch Tim change his mind *twice*:
The only way I can make sense of Dan's bizarre posting is if he
interpreted the above to mean that gun ownership was not associated
with anything at all that might be associated with it such as handgun
ownership or overall homicide. The meaning is "of the things
mentioned in this sentence the only associated with gun ownership was
gun homicide." My apologies if you thought I was saying that gun
ownership is not associated with handgun ownership or overall homicide.
> My comment:
> >> So now we have the totally unremarkable finding that if you get shot
> >> in your own home, there's likely to be a gun in the home. And drowning
> >> victims are usually found near water. Big deal.
Dan misunderstands what I wrote.
> Tim continues:
> >Sigh. The study found that overall homicide was associated with gun
> >ownership, not just gun homicide.
I clarify things. That should clear things up....
> Oh, lookee. Now it's *overall* homicide and "not just gun homicide".
> As soon as you make up your mind, Tim, let us know.
Nope. Seems that once Dan gets hold of a wrong idea he has a hard
time letting go.
> Tim now makes an argument about how increased *overall* homicide
> might be explained:
> >There are two plausible mechanisms
> >to explain this:
> > 1. Guns make violence more lethal
> > 2. People at risk of homicide acquire guns for defence.
> So far so good -- now he switches gears yet again in mid-stream:
> >If 2. is true we would expect non-gun homicide to be just as strongly
> >associated with gun ownership as gun homicides are. It isn't, which
> >suggests that 1. is the more probable explanation.
> Now he's *back* to the claim that only gun homicides were strongly
> associated, which violates the very premise to his "plausible mechanisms"
> above. So we see Tim saying "to explain X, we consider mechanisms
> 1 and 2, but since X is *not* the case, number two can't be true."
> Huh?
Now Dan seems to have possessed himself of the notion that if gun
homicides are associated with gun ownership and non-gun homicides are
not, then it is impossible for overall homicides to be. Here are the
actual numbers from the paper.
Adjusted odds ratio associated (and 95% confidence interval) with gun
ownership:
Overall homicide 2.7 (1.6-4.4)
Gun homicide 4.8 (2.2-10.3)
Non-gun homicide 1.2 (0.5-2.7)
Note that the last one is not significantly different from 1.0, but the
first two are.
> As soon as you make up your mind, Tim, let us know. In the meantime,
> we'll just sit back and watch you disagree with yourself.
I think I'll just sit back and watch you try to remove your foot from
your mouth.
Tim