Message-ID: <495A5AA4.3010300@pricom.com.au>
Date: 2008-12-30T17:30:12Z
From: Philip Rhoades
Subject: Additive/multiplicative estimations?
In-Reply-To: <alpine.LFD.1.10.0806201117150.30072@gannet.stats.ox.ac.uk>
People,
If Test 1 estimates the likelihood of Population A of being Type Z as
60% and if a completely separate and unrelated Test 2 estimates the
likelihood of Population A of being Type Z as 70%, is the likelihood
that Population A being Type Z raised above 70% ?
My gut feeling is that if the results of the two tests are greater than
50% and the two tests are unrelated, the actual likelihood should be
more than the mean of the two tests.
Thanks,
Phil.
--
Philip Rhoades
GPO Box 3411
Sydney NSW 2001
Australia
E-mail: phil at pricom.com.au