Yet Hal, isn't this exactly what we are discussing? This sanctification does not arise out of mathematical statistics or theory but out of areas of application and their practitioners who want or need a value for insertion into a program. In my experience many have no clue about type 2 errors, increase in sample size vs significance, or the reasoning behind the inherent lack of choice with 0.05.
------------------------------
Lee-Ann Hayek
Chief Mathematical Statistician;Senior Research Scientist
Smithsonian Institution
------------------------------