Statistics' Flimsy Foundation

[From
Fred Nickols (2010.03.21.0736 MST)]

I
picked this up on another list and thought CSGNet members might find it
interesting and useful.

“It’s
science’s dirtiest secret: The “scientific method” of testing
hypotheses by statistical analysis stands on a flimsy foundation”

http://www.sciencenews.org/view/feature/id/57091/title/Odds_Are,_Its_Wrong

Regards,

Fred Nickols, CPT

Managing
Partner

Distance Consulting LLC

fred@nickols.us

www.nickols.us

740-504-0000

“Assistance at a
Distance”SM

[Martin Taylor 2010.03.21.12.02]

[From
Fred Nickols (2010.03.21.0736 MST)]

I
picked this up on another list and thought CSGNet members might find it
interesting and useful.

“It’s
science’s
dirtiest secret: The “scientific method” of testing
hypotheses by statistical analysis stands on a flimsy foundation”

http://www.sciencenews.org/view/feature/id/57091/title/Odds_Are,_Its_Wrong

Thanks, Fred. It’s always nice to have these truths pointed out from
time to time. Since I was in grad school, I’ve never stopped wondering
why people continue to use (and worse, to believe in the implications
of) significance tests and frequentist approaches to probability. The
topic does get aired here from time to time. It never goes away. It
even shows up, hardy even disguised, in the perennial discussions about
the usefulness of moderate correlations.

Martin