Huffington Post's Easy Fumble
A story at Huffington Post tied to the second mass shooting at Fort Hood makes a bad case (veterans bring the war home with them), and does a very bad job making it. Some analysis.
Huffington Post's breakthrough revelation that crimes happen where people live and sometimes people are veterans http://www.huffingtonpost.com/2014/04/07/map-veteran-murder_n_5092548.html?utm_hp_ref=tw …- A conversation between
a veteran and the authortwo veterans about the piece. 
.@mileskathleen's exhaustive research: U.S. troops are Americans, and sometimes Americans kill other Americans http://huff.to/1mVlSqs
@AlexHortonTX TBF, I raised concerns about this piece with eds and they were very responsive, made edits to reflect lack of PTSD links- @maxjrosenthal How about the graph that links murder with more deadly tours? Is that a correlation, or was it a larger population of troops?
- @maxjrosenthal Wouldn't that theory prove a delayed reaction? Hard to be deployed in Sunni Triangle in 2006 and them kill at home in 2006.
- @AlexHortonTX Talked w/ them about lack of analysis re: causes/circumstances. I agree re: the graph but they're honestly trying.
- @maxjrosenthal And the violence clustered around bases. Jesus, man. Fear mongering sensationalism is not a good look. @mileskathleen
- @AlexHortonTX I don't think the next story we run on this topic will look similar to this one, I'll put it that way.
- @maxjrosenthal Really glad they're trying. But why was it even run when the fourth graf negates the entire idea of the graph/weak data?
- @AlexHortonTX The 4th graf was a suggestion of mine that they took. They believed the data was presented neutrally when they ran it...
- @AlexHortonTX Obviously we have a different take and they very were open to it, that's all I'm trying to say. They want to get better.
- An analysis of the bad social science practices in the story
Wow, somebody get @HuffingtonPost a social scientist if they're going to run alarmist pieces such as this: http://www.huffingtonpost.com/2014/04/07/map-veteran-murder_n_5092548.html?utm_hp_ref=tw …
@stcolumbia @HuffingtonPost so 194 homicides out of 2.5 million deployed men and women?- @stcolumbia @HuffingtonPost so 194 homicides out of 2.5 million deployed men and women?
- @stcolumbia @HuffingtonPost and the numbers of decreased as those wars have wound down? Someone call Eliot Ness
- @NeilBhatiya @HuffingtonPost Wow, the killers mainly targeted people close to them? How unlike all other homicide patterns!
- @stcolumbia @HuffingtonPost and they occurred near military bases? like, where soldiers live?
- @stcolumbia Fourth graf alone describing no link between PTSD and violence is enough to kill the piece.
- @stcolumbia Based on back of envelope math, the homicide rate for Iraq/Af veterans is 7.76 per 100K; its 12 per 100K for the U.S. as a whole
- @stcolumbia @AthertonKD actually it would be even less because that 194 # is cumulative over a decade or so
- @stcolumbia @AthertonKD damn my liberal arts degree
