Vanity action and randomness

[dave_hoaglin at abtassoc.com]

Rob, WS1A, reminds us that "random" does not mean "randomized."
Unfortunately, saying "random" often creates an expectation of
"randomized."

Either way, it is important to recognize that the randomness is in the
process, NOT in the outcome.  If the applications were processed in a
truly randomized order (many would call it a random order!), all of
the possible permutations would be equally likely.  The actual
permutation is one of these.  Thus, one cannot determine, simply from
analyzing the outcome, whether something went wrong in the processing.
One can, of course, find patterns in the outcome that suggest
problems.  A number of people have been doing just this and proposing
explanations.  The key step, which we are not yet able to take, is to
relate these patterns to a detailed description of the actual process.
Such an account of what the FCC actually did could settle most of the
questions.  It might even be sensible to publish the detailed record
of the order in which applications were processed.  If the order was
determined by a computer program, such a record might be available.

Some of you may remember the draft lottery of the late 1960s.  The
outcome caused quite a stir, because of a definite trend in the
relation between birthdays and lottery numbers: men who were born
later in the year were more likely to have received low numbers.  An
inquiry produced a detailed description of the process, which involved
placing the capsules (containing the days of the year) in a tray and
carrying the tray up and down a flight of stairs -- in a thoroughly
inadequate attempt to "randomize" the capsules.  The details of the
process gave a good explanation of why the order had turned out as it
did.  (All this was in an article in Science magazine,
about 1970.  I don't have the reference handy.)  Subsequently a panel
of experts recommended new procedures, involving actual randomization
(via random numbers, as I recall), which was used for the draft
lottery in the following years.  That process could also have produced
the order seen in the first lottery.  If the randomized process had
been in use from the start, the unusual pattern would not have led to
the discovery of any problems with the lottery.  Bottom line: We can't
tell from the outcome alone; we have to study the process.

By the way, as I recall, despite its flaws, that initial draft lottery
was not overturned.  Our current flap is minor in comparison.

Dave, K1HT (ex-K8JLF)    dave_hoaglin at abtassoc.com

(I submitted my application by Express Mail and received my third
choice.)


>From kg6lf at c-zone.net (Jerry Boyd)  Sat Nov 16 05:36:19 1996
From: kg6lf at c-zone.net (Jerry Boyd) (Jerry Boyd)
Date: Fri, 15 Nov 1996 21:36:19 -0800
Subject: Hardcopy licenses
Message-ID: <328D52D3.2758 at c-zone.net>

Since someone asked if anyone had received the actual hard copy of a 
vanity call in the mail from the FCC, the answer is yes.  The Club Call 
(N6UI) for which I am trustee arrived today.
Jerry
KG6LF (maybe going to change this one too!)