[MR] Pollings in general

[Richard Fitzgilbert]

Greetings from Richard (Well Jeff Really),

Ok, I'll confess.  In my mundane life I help provide all kinds of infrastructure
and analysis for the design, execution and evaluation of surveys.  The nature of
non-response bias is a subject of considerable conjecture and very little
consensus.  The price of investigating non-response is high.  Quality
non-response bias research requires in-person interviews of the non-responders.
Usually, the non-responders are given a modest remuneration to entice their
participation.  With the intensive efforts involved as well as the cost of the
remuneration it's not surprising that it's expensive research.  Another reason
that very little of this kind of research is done is the surprisingly equivocal
results that have been the results of some of these studies.  The only
conclusion that each of the non-response projects I've studied can agree upon,
is that the don't know, don't care, don't want to talk about it, and similar
"uninteresting" categories are much larger for the non-responders than the
responders.  Isn't that astonishing!

OK, on the other side of the coin, statistical treatment of small sample surveys
can determine the reliability of the results drawn from the sample.  It's
counter-intuitive, in fact, how little it takes to reliably represent a large
population accurately.  I always think in terms of people not easier things like
fruit flies or sub-atomic particles.  "Accurately," of course, is open to
interpretation.  My recommendation (Jeff here not Richard [everyone knows
Richard can be something of a hard @$$]) is that we simply follow the good plans
that have been put forward to once again jump through all those hoops (oops that
was Richard again).  That's assuming, of course, that the current Queen wants to
continue this process.  Richard would suggest to Her Majesty that she consult
with Her Highness on this issue.

Yours In Service To Research Designed to Help Improve the Human Condition,
Jeff


-----Original Message-----
From: atlantia-bounces at atlantia.sca.org
[mailto:atlantia-bounces at atlantia.sca.org] On Behalf Of Michelle L. McDaniel
Sent: Monday, July 07, 2003 11:07 AM
To: atlantia at atlantia.sca.org
Subject: [MR] Pollings in general


Hey all!

In a previous life, I worked as a librarian in a library that specialized in
program evaluation literature. One of the things we had a lot of 
information on was how to conduct surveys and how to interpret the results.


Result rates for surveys vary widely. The cost of preparing and delivering the
survey directly affects the rate of successful response. (This is why spam email
and junk snail mail work, where even a one percent reply is considered good
response, so we keep getting inundated.) Also the perceived importance of the
topic affects the number of responses. The removal of the ermine spots was
terribly important to HRM Niobe. When fully explained there was a greater
impetus for the change. The change from the shell to the crown does not have the
same impact as the removal of the ermine spots. (There may well be discussion of
the reasons for the elimination of the ermine spots in the archives of the Merry
Rose. I could not find any specific posts as the search engine for  the list
before July 1999 is not working and the page for the list between July 1999 and
2001 isn't working either.) 

The lack of response for the polling of Her Majesty's arms is just that, a lack
of response. It does not mean tacit approval or disapproval. (There is a whole
pile of research on non-responders) Check literature on response rates for
discussion of the topic. 

Two interesting web site on the topic are:
http://www.dssresearch.com/library/general/mailresp.asp which discusses why
response rates to mailed surveys are low and how to improve response rates and
http://www.ascusc.org/jcmc/vol6/issue2/sheehan.html which covers e-mail response
rates and why they have been dropping since 1986.

If the heralds, Their Majesties, and Their Highnesses decide that they need to
re-poll the populace, I would suggest that the polling be conducted in both the
on-line and in-person petition formats. Just because recent surveys show that a
slim majority of Americans have access to the Internet, does not mean that
Atlantians, can conduct all business via the Internet. (Just because many
Scadians are extremely computer literate, not all are. Let's not exclude those
who aren't by only having an on-line polling.) (For more info on improving
response rates, see the above links.)

Anyways, if we want better response rates to pollings, those doing them should
look at how the professionals do them, the response rates they aim for, and the
ways they use to increase response. The research already done by those
professionals may provide useful information to counter the decision made by
Laurel.

Yours in service,

**aelfwynn of whitby** (who's really glad she no longer works in program
evaluation, it was boring ;-) )

========================================================================
                   The Merry Rose Tavern at Cheapside
    List Info: http://merryrose.atlantia.sca.org/
  Submissions: Atlantia at atlantia.sca.org
Subscriptions: http://seahorse.atlantia.sca.org/mailman/listinfo/atlantia