 | market - home |
|
| market - home |
|
 |
 |
News & Events:
|
Posted by David L. Martin on July 18, 2002 at 10:01:20:
In Reply to: Am I missing something here? posted by steve jones on July 17, 2002 at 17:57:57:
Good! So how do we go about establishing which
is "larger"? Perhaps this example will clarify
the problem. Are the people of Cameron County,
Texas larger than the people of Miami-Dade
County, Florida? Presumably we’ll need to
measure some people, no? Now, once we have
measured everyone, what are our criteria?
Well, we might simply ask, how many 7+ foot
people are in Cameron County vs. Dade County?
I think you can see that such a comparison might
lead to objections. We are talking about a
population of about 350,000 vs. over 2 million.
There might well be far more 7+ foot tall people
in Dade County, even if the mean, variance, and
shape of the distributions are exactly the same.
It seems counterintuitive to use that as our
criterion for comparison. Well, what about mean
size? That seems more meaningful. Yep, I think
it is. But it also suffers from problems. What
if Cameron County has a much higher percentage of
children than Dade County? That could pull the
mean down. Or what if there is more variance in
body size in Cameron County than in Dade County,
even if the means are the same? That could
produce far more large people in Cameron County,
as a percentage of the total. Well, what about
the percentage of people in the population
7+ feet tall? That might work out. Yeah,
maybe, but the problem is, what if there are
only 2 or 3 7+ foot people in Cameron
County? We don’t have much data to make a
comparison. It is particularly tricky to compare
the larger size classes because, as in any
population, we will find that the numbers tail
off slowly in both cases. There is not a
clear-cut “maximum size” in either case. There
almost never is, Steve, in any population.
You see, asking the question, “Is this larger
than that?” isn’t as simple as it might seem when
we apply it to populations. That is exactly why I have
avoided making simplistic statements like “atrox
is larger than adamanteus,” or “adamanteus is
larger than atrox.” (I seem to recollect reading
that “there are absolutely no simple answers.”)
But notice that no matter how we frame the
problem, we seem to come back to the same kind
of data. Size-class distributions. Am I missing
something? You tell me. How do you propose we
compare the sizes of the two species, if not by
measuring snakes and building up size-class
distributions? And what criteria would you
choose?
