Voices for Evolution
Warren vonRoeschlaub
kv07 at iastate.edu
Wed Jun 1 15:22:57 EST 1994
In article <2si99s$no7 at news.iastate.edu>, kv07 at iastate.edu (Warren
vonRoeschlaub) writes:
> They do? From the politics list (full version):
[...]
Okay, I got scads of e-mail about this, so to slow it down (ha
ha) the following clarifications:
The numbers were correlation coefficients. A CC of 1 or -1 would
correspond to perfect correlation -- that the two data sets are
somehow related. A negative correlation means that the more of one
exists, the less of the other and vice versa.
The confidence given is the largest confidence interval that
excludes the null hypothesis (that there is no correlation) which
corresponds to the possibility that the correlation is really 0.
The data was gathered on a state by state basis, and usually
adjusted to a percentage of the population. Ie, Southern Baptists
was really just the percentage of the population that were
registered members of a southern baptist church in 1990.
SATs were the state averages (or adjusted ACTs if there were no
SATs in that state) and "doctorates" were the number of doctorates
per million people given out in 1991.
Details can be found in my WWW homepage.
--
Warren Kurt | By virtue of being correct, the opinions expressed
vonRoeschlaub | above could not conceivably be those of ISU.
If Kurt says something weird, you might want to check
http://www.public.iastate.edu/~kv07/
More information about the Bioforum
mailing list