Home Statistics and Visualization Win-Vector Blog Worry about correctness and repeatability, not p-values