Archiver > GENEALOGY-DNA > 2010-11 > 1291094025

From: (John Chandler)
Subject: Re: [DNA] P value (was chances are, it's wrong)
Date: Tue, 30 Nov 2010 00:13:45 -0500
References: <F9C440A2-FC59-4A9E-AAAC-85DEE9D2FAB0@GMAIL.COM>, ,<COL115-W50D879F102DC3996D9D454A03A0@phx.gbl>, ,<>, ,<COL115-W1464B78AF0292D6AEFA183A03B0@phx.gbl>,<COL115-W5950BB2C58A31B4806036EA03B0@phx.gbl>,<>,<COL115-W45724B549DCDA5DD2EC4A0A03B0@phx.gbl> <COL115-W424C7732D1583F8960685CA03B0@phx.gbl> <> <><> <><>
In-Reply-To: <> (message from James Heald on Tue, 30Nov 2010 00:42:06 +0000)

James wrote:

> But you're absolutely right. A "95% CI" is *not* an interval that is
> 95% credible, it is *not* an interval for which
> P (theta within interval | data) = 95%

You've made two statements here as if they were equivalent, but they
are not. The first is true, but the second is false. Indeed, the
statement "P (theta within interval | data) = 95%" is *precisely* the
definition of the 95% CI. As I keep saying, the word "confidence" in
the expression "95% CI" has always referred to confidence that the
parameter value lies within the interval. It would be clearer if we
used rigorous punctuation, as in "95-percent-confidence interval", but
hyphens, alas, are going out of style. The point I made before is
that credibility is either an absolute trait (in which case the notion
"95% credible" is meaningless) or a relative trait (in which case the
notion "95% credible" is meaningless). Attempts to redefine
"credible" to mean something else are misguided at best. Another way
of putting it is that the entire 95% CI is "credible" by one popular
convention, but some parts of it are more credible than others.

John Chandler

This thread: