Quote:
Originally Posted by Platapus
" We enrolled 677 case participants that had been shot in an assault and 684 population-based control participants within Philadelphia, PA, from 2003 to 2006. We adjusted odds ratios for confounding variables.
Results. After adjustment, individuals...."
Without knowing what adjustments and what the confounding variables, it is difficult to determine the accuracy of the results.
|
Further to what Sky just posted,
We adjusted odds ratios for confounding variables should also be explained...
Confound variables are pretty much things which affected the results and offer an alternate explanation other then the proposed one. All forms of psychology research use various methods to avoid or at least limit confound variables as much as possible.
In this case I would imagine that they made sure that the ratios of crime types (mugging, rape, home invasions) was balanced between the two groups to avoid skewing the results. This would be done as certain types of crime are probably more likely to incite violent behavior then others.
Statistical analysis is also usually a pretty solid research method provided everything was done on the up and up by the research team. Statistics themselves cannot be molded provided they were gathered and calculated properly, but you can mold how they are presented to the public, as they do not understand statistics.
As for the results I am not surprised by the results. Firearms will always escalate the probability of violent action, not suppress it.
Quote:
|
Originally Posted by Freiwillige
The probability of the bad guy getting shot by an armed citizen goes up 100% verses an unarmed victim.
|
I'm sorry but that statement is incorrect. The probability of someone getting shot goes up dramatically if both parties are armed sure. But chances are the bad guy will be the one who does the shooting first.