[EM] Re: Participation Criterion

[Joe Weinstein]

Re:  Participation Criterion

Forest Simmons suggests (message 12578) that a 1% margin of error “cannot be 
improved upon in typical public polls.”  As a practicing statistician, I 
agree.  Of course, no one of the specific factors Forest noted - respondents 
lying, misunderstanding, or changing their mind - will necessarily operate, 
in any given poll or survey, to bias the results.  It might be that in a 
given poll all these factors are relatively independent of how the 
respondent will finally vote, so fairly much cancel out.  But nonetheless, 
any survey runs a danger of biased results, from factors - others, if not 
these - which are unforeseen or uncorrectable in advance and maybe not 
recognizable even after the fact.   (For instance, even trained interviewers 
can eventually get tired and careless, so that their manner of questioning 
may elicit biased responses.)

I appreciate both Forest’s endeavor - to try to overcome unfairness to 
voters when the election method fails the Participation Criterion - and 
Richard’s opinion - that maybe it would be better either to insist on the 
Criterion (that’s my bias!) or to decide simply to live with its uncorrected 
failures.

Joe Weinstein
Long Beach CA USA

_________________________________________________________________
Add MSN 8 Internet Software to your current Internet access and enjoy 
patented spam control and more.  Get two months FREE!     
http://join.msn.com/?page=dept/byoa