“Science has somewhat of a problem with sloppy use of statistics. Because many fields and researchers have gotten stuck with a tunnel vision on p values, a lot of research cannot be replicated. Because of the hunt for p values below an arbitrary boundary, researchers are more prone to publish positive results rather than neutral or negative results.
This has led to publication bias and an inflation of effect sizes. Because research offer a great flexibility, researchers who are more concerned about their next grant than solving the mysteries of reality can fudge the analysis by not adjusting for multiple testing or trying to adjust for different various factors until they get the result that they want.”