On Mon, 2 Feb 2004, Ernest Prabhakar wrote: > > We can quibble about the exact terminology to use (and knowing this > list, I'm sure we will :-) but I think it is important to recognize > that there *is* an appropriate place for worry about 'sincerity', even > if its difficult to measure. >
Here's one approach: start with preferences that are assumed to be sincere. Then add in some (more or less accurate) statistical information about the set of preferences. Now construct ballots that would be likely reactions to the statistical information (or disinformation as the case may be). Finally, compare the original preferences with those expressed on the ballots. Forest ---- Election-methods mailing list - see http://electorama.com/em for list info
