[Sigia-l] ballot usability redux

[Christopher Fahey [askrom]]

> > I was responding to the notion of making *every* voter's ballot 
> > unique, to introduce statistical randomness.
> 
> Do you need to make *every* ballot unique to introduce "statistical 
> randomness" ?

It seems sound to me from a purely statistical point of view. If every
voter sees a totally different ballot sequence, then the effects of each
specific ballot sequence (for example, the risk that voters will tend to
prefer candidates that appear on the first page) will be dissolved among
the mass of different ballots.


> > since you are not parsing the actual 193 names anyway, you'd be 
> > looking for secondary attributes, such as party affiliation. 

AFAIK the ballot will *not* indicate party affiliation. Just names. In
fact, it seems to me that there is no way for a voter to know that the
candidate named "Michael Jackson" (there is one) isn't the world-famous
Michael Jackson (he isn't). This is bad news for Democrats who are
hoping that people know enough to scan through the ballot for the one
marked Democrat. 

-Cf

[christopher eli fahey]
art: http://www.graphpaper.com
sci: http://www.askrom.com
biz: http://www.behaviordesign.com