Michael Granaas wrote:
>
> On Thu, 16 Nov 2000, Thom Baguley wrote:
> >
> > I was also struck by the poll result in Florida newspaper reporting that most
> > Floridians thought the election was fair. Most was less than 2/3 in that case.
> > I would be horrified if 1/3 of the electorate in an English election thought
> > the election might be unfair. (I'd hope for 99% plus percent thinking it fair).
> >
> > Thom
>
> As someone living in the U.S. I'm surprised that over 50% thought the
> election was fair. From the media we are getting a picture that
> indicates that all Gore supporters thought that the election was unfair
> and the Bush supporters thought that it was fair.
"It all depends what you mean by" fair - and that may be more a problem
for the psychologists, philosophers, and anthropologists than the
statisticians.
For instance, there is a tendancy to view the US VietNam draft lottery
as "unfair" in retrospect because birth dates at certain times of year
were favored by a flawed randomization process. However, it seems to be
agreed that this was done without malicious intent.
How would you rate the "fairness" of the following schemes?
-order of callup determined by day number: eg, Jan 1st then Jan. 2nd
then... then Dec. 31st
-same but starting at a randomly selected day (Apr. 23 followed by ...
followed by April 22nd
-same as first, but with one toss of a coin deciding once and for all
whether to go forward or backward through the year?
-day numbers ordered by some "more uniform" scheme (say, every 17th day
- note 17 and 366 are coprime - after a randomly selected starting
point)
-day numbers randomly selected by a well-tested pseudorandom number
generator (note that these are in fact deterministic?)
-individual day numbers selected completely at random, as attempted in
real life?
- a number from 1 to 365 assigned at random to each male American
without reference to birth date?
Would your answer to any of the above (except the last) change if the
scheme had been decided in (say) 1945, before the birth of the young men
involved, so that their birth dates were decided after the ranking not
before? Is it "fairer" because the draft lottery position is randomized
individually for each man, or less fair because he is "doomed from
birth" instead of spending many years knowing he was going to be called
up first?
What if it had been decided and kept secret?
Would your answer change if your belief in the justness of compulsory
military service in the VietNam war were different from whatever it is?
-Robert Dawson
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
