Re: Crime creation

[ltrwritr@moscow.com (Mark Rounds)]

Folks

I spent more time than I should have checking this out but being I geek, I
kinda do this for fun anyway.  Time for a quick little lesson in science.
The premise is that a statisical corelation "proves" a given hypothesis.  

Saying, for example, (is that too many comma's?) that because a single
corelation occurs, it proves anything is treading on thin ice.  If you say
for example, that, since black americans score lower on IQ tests (which is a
very true statement) when looking at the national averages, it proves that
blacks are less intelligent the national norm.  The connection isn't proven.
What it proves is the black scored lower on standardized IQ tests.  There
could be a plethora of reasons why this occurred.  There could be bias in
the tester who give the tests, the schools on the areas the blacks are in
could be poorer and hence the children being tested score less well, perhaps
the teachers of a given area are of a different eithnic background and can't
relate as well to the students.  it could mean that the test writers
included their own cultural biases (which we all do) and hence those of
their own background scored higher and on and on.  

For true science to occur, you must construct an objective test  for each
hypothesis and mitigating factor.  As you work through all of the potential
mitigating factors, the hypothesis closest to the truth slowly becomes
evident and guess what?  In the process of constructing tests and completing
them, you actually learn something, change the hypothesis and test some
more.  It isn't as easy or as black and white as TV and the media make it
sound.  Good science takes time.

The studies named take a simple statisical corelation and use it as proof
something exists.  The corelations do nothing of the kind and rather are
interesting statistical facts that bear a more complete examination.  It
should have been handed back to them in peer review with the comments,
"interesting hypothesis, but where is the proof?".  It has happened to me
when I have attempted to publish prematurely and I expect that it happens to
just about every researcher once in a while.  Its part of learning how to do
science.  

The fact that it did get published doesn't give it great validity either.
Its only when the little drop of science that you add to the race's pool of
knowledge connects with others, spawning other research that proves valid,
is there a connection.  A breakthrough may happen once in a life time but
good science requires a life time of dedication.

Mark Rounds