>Gene,
>
>You have made extended comments about
>the IWF report "Confession without Guilt?"
>(at http://www.iwf.org/news/mitfinal.pdf
>about women biologists at MIT.
>
>Some background information:
>
>------------
>The IWF is the second in a series criticizing
>the MIT report on the Status of Women.
>The original MIT report may be downloaded
>from http://mindit.netmind.com/proxy/http://web.mit.edu/fnl/
>
>An earlier IWF report by Judith Kleinfeld revealed many
>of the serious shortcomings of the
>MIT report. Kleinfeld's report is at
>http://www.uaf.edu/northern/mitstudy/#note9back
>
>I would urge all students to download and read
>all three of these papers, so they can get a better
>feel for who, MIT or its critics,
>is more objective and scientific in
>their treatment of a very touchy subject.
>-----------------
>
>Gene, although you apparently intended
>them as honest criticism, your
>comments revealed a deep, fundamental
>confusion about statistical inference.
>I expect any trained statistician would recognize
>the error in your argument immediately,
>but some students may have been misled.
>
>So let me try to alleviate some
>of the confusion your posts may have
>generated, and try to help you see
>the error of your ways. And,
>might I suggest, strongly,
>that you show this post to
>your ECOS611 class? This might
>make an excellent discussion piece
>for them.
<Snip>
I'll share Dr. Steiger's comments with my statistics class. Dr. Steiger asks
the readers of this newsgroup to read the whole set of documents on the MIT
gender-bias issue. I must confess to not having done so. I read his IWF
report, co-authored with Dr. Hausman, only after the Boston Globe described his
conclusions under the heading "MIT bias claims debunked." My post was in
reference to the claims made in his IWF article, not to the other documents
that preceded it. His article can be downloaded at:
http://www.iwf.org/news/mitfinal.pdf
Dr. Steiger accuses me of improperly using statistical tests to make
inferences to a larger population. I didn't do that. Drs. Steiger and Hausman
in their IWF report claim to have found "striking," "compelling," and
"dramatic" differences in productivity between senior male and female Biology
Faculty at MIT. I read their report and didn't see much evidence for gender
differences at all. Most of the apparent gender differences in their graphs
disappeared when the data were plotted on a logarithmic scale:
http://www.es.umb.edu/edg/ECOS611/iwflnfigs.pdf
Dr. Steiger's post states, "There were HUGE differences in the citation rates
of senior men and women. The mean number of citations was, as I recall, roughly
7000 for the men and 1400 for the senior women." The actual data were 7032 for
the men and 1539 for the women (with sample sizes of 6 and 5 respectively).
The geometric means were 4800 and 1400. A Mann-Whitney U test indicates that
12.6% of the permutations of these 11 data would produce differences in
citation number as extreme or more extreme than those reported. Do these 11
data offer compelling or dramatic evidence for gender differences in
productivity? Not to my way of thinking. Was I making inferences to a larger
population? I didn't intend to. I was just trying to assess Steiger &
Hausman's claim of HUGE gender-based differences in productivity.
Dr. Steiger's post recommends that I should eschew using any formal statistical
tests on the IWF report data. What is the alternative? The approach used in
the IWF report is to point out an individual datum or a few data points, and to
make claims about "striking," "compelling" and "dramatic" differences between
the sexes. I regard Dr. Steiger's evaluation of these data as being highly
subjective. I have no idea what objective criteria he used to reach his
conclusions of "compelling," "striking" and "dramatic" differences between male
and female faculty.
Eugene D. Gallagher
ECOS
P.S. Here are answers to your questions:
>1. Suppose the IWF report had gathered data on ALL female scientists at MIT,
> compared them to a group of males matched for seniority, and had shown huge
> differences in performance. Would you still be doing a significance test?
Of course not. I posted a note called "Florida votes and statistical errors,"
on 11/30/00 on this newsgroup on this very topic. DejaNews is dead, but you
can find this post on Google.
> 2. Suppose Mary has 1051 citations in the last 10 years, and has earned
> 4 million in grants, and Fred, in the same department, and about the same
> age and seniority, has 12000+ citations and has earned 23 million in
> grants, and Fred finds out that Mary is making the same salary he is.
> Should Fred consult you about doing a statistical "significance" test
> before asking for a raise?
No. What does this hypothetical have to do with your claim that "dramatic"
gender differences exist in the productivity of MIT biology faculty?
>3. Suppose you go to a nursing school, and find 5 female faculty with 7,000
> citations apiece, and, in the same department, 5 female faculty of the same
> seniority with 1400 citations apiece. Would you expect their salaries and
> departmental influence to be the same?
I have no idea. It is you who claim that salary differences and departmental
influence are related to publication and citation patterns. I made no such
claim.
I didn't find the basketball analogy amusing or appropriate. You resort to
ad hominem attacks -- "Gene Gallagher, intervenes. Gene considers himself
an expert statistician," -- rather than defending your conclusions in the IWF
report. By the way, I've never claimed to be a statistician, expert or
otherwise. I am an oceanographer.
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
