[ncsg-principal_reps] NCSG Checkin

[Dan Krimm]

On Thu, July 21, 2011 12:44 pm, Avri Doria wrote:

>              (
>               personal aside, I find it remarkable that the
>               two ratios came out so close given proportional
>               voting based on organizational size or individual status -
got to be amused by the little things in this job
>               or you will go completely bonkers!
>               )


This sort of thing is my occupation these days (policy research,
statistical analysis), so permit me to engage this tangent.

What this indicates is that the probability of response (or non-response)
is not correlated with (i.e., appears to be independent of) the vote-count
per respondent.  (I checked it per respondent type, and it is comparably
close across types:  large = 7/21 = 33.333%, small = 22/64 = 34.375%,
individual = 55/166 = 33.133%)

Why this should be remarkable or not is an open question.  :-)

Why would you hypothesize that they would be different?  Did you think
institutional members would be systematically different (on average) from
individual members, in this regard?  Do you have a theory of response that
predicts this?

Dan


-- 
Any opinions expressed in this message are those of the author alone and
do not necessarily reflect any position of the author's employer.