Some solutions to the p-value communication problem– and why they won’t work [pdf] (stat.columbia.edu)