John,
One thing that Danny's questionaire didn't address was "How many additional
folks might consider putting their names in the hat if they knew the
candidates. In past years, when I have gotten a request from NOMCOM to review
the padded list, I've thought to myself 'If I knew only these folks were
running, I would have considered ...' I wonder if other people have thought the
same.
John L.
_____ Original message _____
Subject: Re: Complaining about ADs to Nomcom (Re: Voting (again))
Author: "John C Klensin" <john-ietf(_at_)jck(_dot_)com>
Date: 16th May 2005 9:58:21 PM
In the light of this and Dave's comments, and since I used to
teach people how to design survey questions so that the
questions were as non-reactive as possible and the answers could
be interpreted. There is nothing inherently wrong with a
self-report question. We ask them all the time and normally
expect truthful answers. The tricky part is understanding which
questions people may not want to answer truthfully, the reasons
why, and, if the person who is reluctant to answer provides
_some_ answer, how either that or a pattern of non-response is
likely to bias the results.
For example, if one asks a large sample of 10-year-olds how old
they are, the answers will, predictably, be mostly truthful:
there are few incentives to lie and mistakes will tend to be
nearly randomly distributed (slightly fatter tail to the
"younger" side because of forgetting birthdays). If one asks
the same question of 60 year olds, the answer pattern would
probably be different, and it is important, if one is trying to
interpret validity, to understand those differences and their
likely impact, rather than assuming either that all population
groups are the same or that all self-report answers are invalid.
Coming back to the question at hand, if the nomcom asks people
whether they would have accepted nominations if their names
would become public, why would someone lie? And, if they did,
then which way would the report be biased. I would think that
people who are inclined to give incorrect answers would be more
inclined to answer "no problem" given the community's biases
about openness and unwillingness to admit that they require
secrecy. Maybe I'm wrong about that, but, if I'm not, the
results Danny reported would, if anything, underestimate the
number of people who would not be willing to be considered if
their names were public.
We now return you to the regularly-scheduled religious arguments
on the subject.
john
--On Monday, May 16, 2005 10:52 +0200 Brian E Carpenter
<brc(_at_)zurich(_dot_)ibm(_dot_)com> wrote:
You've seen Danny's message with the results of asking the
question in a straightforward way - 20% of IESG nominees
say they would not have volunteered. Unlike Dave, I am
willing to believe them.
fwiw I responded "Yes" to Danny's question, but not
without careful thought and some hesitation.
Brian
Dave Crocker wrote:
Seems fairly easy to judge the validity of that argument to
me. ASk the nomcom to ask volunteers whether they would
have volunteered if their name was gonig to be made public.
Collect statistics.
Sam,
Sorry, no.
As I posted earlier, that sort of methodology relies on what
survey researchers call "self-report".
It is very good for assessing attitudes and very bad for
assessing actual behavior.
For example, what you are likely to get are responses that
indicate whether the people would like to have nominations
be public.
It does not guarantee -- and well might not even correlate
with -- whether they really would run or not run, depending
on the public-ness of the nomination.
It is one thing to ask simple questions about simple issues.
As soon as we get into something more "political" the
psychodynamics get messy.
d/
---
Dave Crocker
Brandenburg InternetWorking
+1.408.246.8253
dcrocker a t ...
WE'VE MOVED to: www.bbiw.net
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf