🚀 go-pugleaf

In article <1emuddINNs9k@talon.UCS.ORST.EDU> stanley@skyking.OCE.ORST.EDU
(John Stanley) gives various objections to the validity of the straw
poll which I conducted recently.

Overall, it seems evident to me that John misunderstands the reason
for the straw poll on the r.m.synth renaming.  He seems to believe it
is being held up as a replacement for the reorganization vote.  This
is why he objects that fewer people responded to it than to the vote
and less time was allowed for response.  The following quote
exemplifies John's position.

   despite all these differences (and all of them limitations in
   allowing people to participate in your straw poll) your straw poll
   is a better indication of what people want.  Sorry. It isn't.

I have made no such claim.  I am making no attempt to replace the vote
results with the straw poll results.

The straw poll asked *.synth readers and news admins what they want
done, now that a dispute has arisen over the vote.  I believe the
results of the poll show that most *.synth readers and nearly all the
news admins want the dispute ended and the revote dropped.

Since the disputants are (ostensibly) acting either as advocates for
the readership or for the admins, I repeat my call on all involved to
drop it and concentrate on improving the guidelines to avoid this
happening again.

Now I shall address the specific objections John has to the straw
poll.

   You didn't include a number indicating total response in your
   extremely long analysis [. . .]  If you DID include totals
   somewhere, it apparently was not in the first 26 lines, and thus
   those results are meaninless.

I included the total response in the section titled RESPONSES, lines
15--27.  The total number of respondents is 109.  Your unfamiliarity
with the crosstabulation leads me to suspect that you did not take
much care to understand the results.

   The vote had, as I recall, slightly more than 270 voters. Even if
   not, the result would have had to had 100 MORE yes than no votes
   (and thus more than 100 votes total) to pass.  I think more people
   participated in the vote than in your straw poll.

Only a part of the voting population were *.synth readers.  You are
trying to compare 270 votes on widely posted renaming proposal
affecting multiple groups with 109 responses to a poll posted posted
to and relevant only to *.synth readers and news admins.  This is like
comparing apples with oranges.

In my opinion, the question is not whether the sample is as large as
the vote, but whether it informs us of the opinions of the *.synth
readership.  If you can show that the ratio of voters to poll
respondents somehow correlates with the representativeness of the
sample, then this objection is of interest.

Your objection quoted my claim:  "The dispute brought more people into
the process than the vote did."  Perhaps my meaning was unclear.  This
was not a comment on the straw poll.  It referred explicitly to "the
dispute."  I never claimed to have gotten a larger response than the
vote.  The numbers are a matter of public record.  I specifically did
/not/ claim, "The /poll had more respondents/ than the vote did."  My
point, instead, was that the dispute got people involved who were not
involved in the vote.  What I implied was that the straw poll sampled
both people who voted and people who did not vote.  The comments of
the respondents bear me out on this.

It was claimed that some who did not respond to the call for votes
were misled by the subject line of the article, and would have voted
had this not happened.  Some of the straw poll respondents indicated
that this happened to them.  They cared about the outcome but the
process let them down.  This implies that the poll may be more
representative of the *.synth readers than the vote.  However, I am
not prepared to speculate about this.  I simply think the poll
supports the position that the publicity for multiple group votes
ought to be written with more consideration for the readership of the
affected groups.

You had several other questions about the validity of the straw poll
results.  I will answer them as well as I can.

   The vote was open for at least 21 days, and your straw poll was
   open for only about a week.

There is nothing about netnews propagation which will skew the results
of a poll towards one side or another if the returns are cut off at a
particular time.  I received responses from all over the world, and
doubt that the polls closed too soon to get a representative sample.
As above, if you can show that the ratio of vote period to poll period
somehow correlates with the representativeness of the sample, then
this objection is of interest.

On the other hand, the dispute ought to be resolved as quickly as
possible, and dragging a straw poll out for weeks is counter-
productive.

If you still think that the poll closing date was of importance,
perhaps it will be more convincing to you to learn that a longer
polling period would have made very little difference in the number of
responses.  Here is a tabulation of the number of responses I have
received to date, broken down by date of response in the header.

	Sun Nov 15	(poll opens) 0
	Mon Nov 16	69
	Tue Nov 17	26
	Wed Nov 18	10
	Thu Nov 19	5 (poll closes)
	Fri Nov 20	6
	Mon Nov 23	1

I waited to close the poll until the numbers showed that the number of
responses I would discard was no longer large enough to balance the
value of quick results.  The late respondents now number exactly 8.
(One Thursday response arrived Friday.)  The responses of these 8 are:
ya, ya, nc, ya, yb, nc, nb, yb.  These responses do not invalidate the
analysis I posted earlier.

   You, by your own admission, gave those who read news.admin.* only 4
   days to respond.

I gave them neither more nor less time than I gave *.synth readers.  I
do not understand why you single out news.admin and news.admin.policy.

   You delayed notification of those whose admins had implemented the
   name change.

This question was addressed in the results posting.  Since the poll
did appear in the new group, and the responses have died out, I
conclude that there is no significant effect on the results.

Regards, Ed
-- 
Ed McGuire                   1603 LBJ Freeway, Suite 780
Systems Administrator/       Dallas, Texas 75234
 Member of Technical Staff   214/620-2100, FAX 214/484-8110
Intellection, Inc.           <ed@intellection.com>