Re: What Are the Race Deniers Denying?
Don Dale (dale@princeton.edu)
31 Oct 96 13:45:20 -0400
>> If you're particularty interested in human variation, try a book like
>> R.C. Lewontin's Human Variation, which shows that the variation within
>> human subgroups is much greater than that among the subgroups. That is,
>> the average genetic difference between any two human subpopulations is
>> less than the average difference between two members of the same
>> subpopulation.
>
>This is an extremely important point and one which I hope the people
>following this thread will think carefully about. If within group
>variation is greater than between group then there is SOMETHING WRONG
>WITH THE WAY THE "GROUPS" ARE CONSTRUCTED.
This is false. Consider that the time it takes for a man to run a mile is
a random variable with a huge variance. Likewise, the time it takes for a
woman to run a mile is also a random variable with a huge variance and a
slightly higher mean than that for the men.
Now, there's a runner at the starting line, and you want to predict his or
her time. Do you want to know the gender of the runner? Does this
information give you any predictive power at all? Of course it does.
Granted, there might be other pieces of information you might wish to have
in order to better predict the runner's finishing time. Some of these
pieces of information might even be more "important" (in the sense of
improving the accuracy of your prediction) than the gender of the runner.
But gender still matters, even though the variables have large variances.
Don
|