Recently, the noise of life has interfered with the signal of statistics but I'm thrilled that this particular signal has endured long enough for me to catch it.
George, this is brilliant.
Susan Spruill, in her reply, compared your writing to Hemingway prose - economical and understated. This characteristic gives to your fable added utility as an approachable and enjoyable way to explain p-values to my colleagues who are not statistically inclined.
Thanks for sharing your insight.
Linda
------------------------------
Linda A. Landon, PhD, ELS
Research Consultant
PhD, Molecular Pharmacology
Graduate Certificate, Applied Statistics
Board-Certified Editor in the Life Sciences
Research Communiqué
Clear, Concise Statistics & Words
LandonPhD@ResearchCommunique.com573-797-4517
------------------------------
Original Message:
Sent: 04-09-2017 14:12
From: George Cobb
Subject: p-values: a fable
Once upon a time, there was a brilliant scientist who invented a way to tell the difference between signal and noise in messages from data. His method was not perfect, but it worked well provided the message had been carefully crafted in advance to fit a particular format.
Unfortunately, because the method could be purged of ideas and reduced to a single number, well-meaning journal editors were led to declare, "Our journal will publish only signal, never noise."
Understandably, well-meaning authors, under pressure to publish, were motivated to make their messages look to editors like signal, never noise.
Reality intruded. Because many messages had not been carefully crafted, their content was mostly noise.
Thus was born the industry of trying to make noise look like signal. (This last is a lie, of course, or at least an anachronism. The industry of making noise look like signal was invented by politicians thousands of years before Fisher. Statisticians merely fortified the hoax by attaching numbers.)
The moral: Teaching the p-value as an abstract mathematical construct stripped of its context mainly contributes to the noise in the channel of science. Detecting the signal requires thought. Study design matters.
George Cobb