> From: "Kevin Parish" <email@example.com>
> > the only difference between base rate and individuating information is
> > which comes first.
Not quite. A personal diagnosis (about me) where 85% of those who get
this diagnosis have the common form of the disease and 15% the rare
form, is, whether I get it first or second, a base rate.
A personal test that says I have the common form, a test that is
right 80% of the time and wrong 20% of the time, is still information
about me, whether I get it first or second (although you can think of
the reliability of the test as another base rate).
But it is true that our tendency to think of the diagnosis with the
85/15 odds as being about THEM and the one with the 80/20 odds as being
about ME is rather irrational. In the long run, both probabilities
would affect me (i.e., if I had to do this every day of the week.
> > Subjects' estimates would define a base rate, or, in Koehler's terms, a
> > (subjective) prior probability
> Here is a fine contradiction. If Kohler has defined a base rate as the
> actual population probability of it being one thing or another then to
> say this (above piece) is merely a case of not agreeing with Koehler's
> definition. If Koehler has defined base rate wrongly then that does not
> make findings or theorised material any less valid in the context of
> his (possibly inaccurate) definition
Koehler's definition is fine. What the disagreement is about is, in
some cases, the task (what if the reliability of the test is different
for the common and rare variants? what if the data were presented as
frequencies rather than percentages, and so on.
Stevan Harnad firstname.lastname@example.org
Professor of Psychology email@example.com
Director, phone: +44 1703 592582
Cognitive Sciences Centre fax: +44 1703 594597
Department of Psychology http://www.cogsci.soton.ac.uk/~harnad/
University of Southampton http://www.princeton.edu/~harnad/
Highfield, Southampton ftp://ftp.princeton.edu/pub/harnad/
SO17 1BJ UNITED KINGDOM ftp://cogsci.soton.ac.uk/pub/harnad/
This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:22 GMT