Re: The Bell Curve : Reaction time and IQ
(davidwss@freenet.edmonton.ab.ca)
7 Apr 1995 18:36:28 GMT
Paul Bernhardt (pb1487@u.cc.utah.edu) wrote:
: Hugh LaMaster (lamaster@viking.arc.nasa.gov) wrote:
: : In article <3lplc3$7v3@ucsu.Colorado.EDU>,
: : penrose@ucsu.Colorado.EDU writes:
: : [cnc115@cybernetics.net (Pete Lienemann) wrote]
: : |> >Once again we see, "Pick a conclusion then acquire/manipulate the data to
: : |> >support it."
: : |>
: : |>
: : |> Unfortunately this is and has been the basis of what we consider "science."
: : No, it isn't. You are wrong.
: While in general, science proceeds by following the data in an unbiased
: fashion, it unfortunately does not work so well always. For instance, I
: am working some meta-analysis of studies of chlamydia. The research
: team that I belong too recently worked on a similar meta-analysi of
: treatments for a certain medical condition. The physicians in the study
: specifically excluded people who might be sensitive to side effects of
: the medications. They then conclude that there are no side effect
: problems.
What you have described may not be bad science, if the conclusion
completely stated is that there are no side effect problems when people
who can be identified in advance as sensitive to side effects are
excluded. This could be valuable knowledge. I would like to believe
that this more elaborate conclusion is what was meant.
David Wasserman (davidwss@freenet.edmonton.ab.ca)
Curmudgeon-At-Large (DWasserman@edc.gov.ab.ca)
"If you can't do something well, learn to enjoy doing it badly."
|