Re: Race

Paul J Brantingham (branting@GAS.UUG.ARIZONA.EDU)
Sat, 5 Nov 1994 08:53:57 -0700

I am sorry for not being able to ignore Rushton's most recent plug for
his book (Could we get a price on that next time?). But, I have a genuine
question, not necessarily for Rushton, that derives from a simple problem
that I can't get out of my head:

I would really like to know about the statistical massaging being done to
get around the problem of comparing population mean values when the
within population variation is greater than between population variation
(putting aside for the moment the difficulty in defining the populations
in the first place). Rushton's frequent disclaimers (on the Net) on this
issue are never accompanied by an explanation of how he gets around this
basic problem. Any statistics text stresses that such analyses
are always suspect, regardless of the statistic you use. If this basic
flaw can't be overcome, then why pursue the rest of the scientific (sic)

Politics are pulling up the boot-straps, are they not?

Again, I apologize to those members of the Net who just want to forget
the issue.

Jeff Brantingham.