[291] in Discussion of MIT-community interests
Re: The numbers (complete version)
daemon@ATHENA.MIT.EDU (Richard Tibbetts)
Sat Apr 28 15:41:07 2001
Date: Sat, 28 Apr 2001 15:40:34 -0400
From: Richard Tibbetts <tibbetts@MIT.EDU>
To: "Jimmy_B,MajMoola,MechWarrior,etc._Chien-ta Wu" <jimmbswu@MIT.EDU>
Cc: Zhelinrentice L Scott <zlscott@MIT.EDU>, mit-talk@MIT.EDU
Message-ID: <20010428154033.H29362@multics.mit.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <200104281849.OAA08006@MECHWARRIOR.MIT.EDU>; from jimmbswu@MIT.EDU on Sat, Apr 28, 2001 at 02:49:24PM -0400
On Sat, Apr 28, 2001 at 02:49:24PM -0400, Jimmy_B,MajMoola,MechWarrior,etc._Chien-ta Wu wrote:
> One problem I had with that statistic was that it does not break down
> the class by ethnicity/race *and* gender. This being MIT, I wish they
> will make the raw statistics available for analysis. (ie, the real
> spreadsheet they used, with names deleted.)
This had always annoyed me as well. I asked about it during the
advising forum, and found out an interesting aspect of the problem.
According to people from the advising office, they have to report on
their work to the MIT human experimentation advisory board. Apparently
this group takes a dim view of publishing raw data sets, since it
tends to violate the privacy of the participants (for example, if
there is only one black female at MIT from Colorado, telling you the
GPA of students by gender, race and home state would be kind of bad).
You might argue that they could publish more information, but they
can't publish the raw data. And creating more aggregate information
takes time and effort.
tibbetts
