Hello. I thought I'd put in two cents on this one.
I'm a computer scientist that uses statistics routinely in my work
to model the results of human behavior (software development).
One of my favorite sayings is: ``Figgers never lie, but liars figger.'' :-)
I've learned that *statisticians* are usually like mathematicians,
and are very circumspect when interpreting their results. However,
*social scientists* are much less cautious, because their main interest
is often advocacy. Journalists, and politicians scientists are
usually even more bold.
I understand that opinion surveys do have a rigorous sound methodology
based on statistics, but I don't trust my local TV station to follow that
methodology.
Once interpretation starts, we are no longer talking about statistics.
At best, it is reasoning in the domain of study. At worst,
it is pseudo-science rhetoric.
I like to think of statistics as merely a model of my ignorance
about a situation.
In His service,
Ed
-----------------------------------------------------------
Edward B. Allen, Ph.D.
Research Associate
Department of Computer Science and Engineering
Florida Atlantic University, Boca Raton, FL 33431, USA
Office: (561)367-3916 Home: (561)487-2445
Fax: (561)367-2800 Fax: (561)477-8300
***** Note new telephone area code *****
Internet: allene@cse.fau.edu