(Untitled)

[novapsyche]

Obama: 'We tortured some folks' -- The folks over at DailyKos aren't amused.

HOMICIDE: Medical examiner says NYPD chokehold killed Staten Island dad Eric Garner | more

Lethal drugs injected 15 times in botched Arizona execution -- "'The people who were in charge of putting him to death really did not know what they were doing,' said Austin Sarat

Comments

[dionysus1999]

One of the big problems is how results of studies are published. Negative results are very important, but they make uninteresting reading, so they are rarely published. Exciting new results need proper caveats, though does anyone remember caveats?

The social sciences are especially bad at pretending one study (swallow) makes a summer.

[novapsyche]

See, I never even considered what they described as the "file-drawer problem." I recognized it intuitively once I got the gist, but as someone who studied social science, I never would have thought of just filing experiments that did not confirm what I was seeking, or that failed to replicate another person's findings. That thought just never occurred to me.

I understand about the disincentives to publishing secondarily, which is one of the reasons I highlighted the article. It's good that others have tried a method to counteract these disincentives.

P.S. I felt from the start that you'd be one of the people most interested in that article. :)