Re:Has the US become the type of nation from which you have ...
by Ygg2 on Monday June 10, 2013 @11:35AM
(#5855082)
So its choice 3.
Enough to know that it's larger than ten. And knowledge of shootings is like the worst possible criteria. How many people you knew that died of asthma?
Why not count the crime infested ghetto? They are also result of current politics in one way or other.
I'm guessing with enough applied statistic any conclusion you wish can be reached. Also even by that statistics you'll reach a conclusion that American has a lot of school killings compared to other countries of similar wealth.
Re:Has the US become the type of nation from which you have ...
by eli_gottlieb on Monday June 10, 2013 @11:31AM
(#5855053)
This is practically a lesson in the misuse of data. However you define the metric you're measuring, you should be comparing America to other countries. By any measure, America has more school shootings than most other countries.
So let me rephrase, how many school shooting can you count on your hands that you knew about? I'm guessing you won't make it pass your two hands.
I'm not going to count crime infested ghetto area school shootings that involve a 70% dropout and incarceration rate, where one student shoots another over some argument over a girl. Nor gang related violence in those same schools. They will do that in and out of school as much as anywhere else. It's not related to any particlar place or time.
I'm not going to count a depressed kid that goes into the school to commit suicide in the bathroom. Nor someone that purposely shoots someone else in the butt and gives up. Or attempts to murder one single person at school because they don't know were that other person lives and it's convinent... All those things have nothing to do with schools.
What I'm going to do is only count real school shootings that involve someone walking in and trying to kill as many people as possible. And weight that against all the non-school shooting events in relation to the the number of pupils, schools, and days. Guess what happens then?
Data should always be filtered, outliers removed, then normalized or standardized.