Statistical Intervals, Not Statistical Significance

Abstract Convincing practitioners of the inadequacy of significance testing can employ a two-step approach. First, explain the difference between statistical significance and practical importance. Then, at least in many situations, use an appropriate statistical interval to quantify the statistical uncertainty. By Gerry Hahn, Necip Doganaksoy and Bill Meeker