49: War and p's

Dan and James discuss a forthcoming paper that's causing a bit of a stir by proposing that biobehavioral scientists should use a 0.005 p-value statistical significance threshold instead of 0.05.