Hmm, Ville has identified a design flaw in the experimental protocol. Not to mention that the current system lends itself to ballot stuffing by partisan supporters of sample A.

To remedy these shortcomings, there are now two new rules:
1. You must vote with invisible ink. That way your vote will not influence subsequent voters, and all I'll have to do is develop the tally with iodine vapor.
2. After voting, you must dip your right index finger in indelible ink. Or, in the absence of indelible ink, you may rub yourself with the ink things from the printer. One person, one vote.

Ivan





On Tue, Sep 8, 2009 at 8:52 PM, Ville Bergholm <ville.bergholm@iki.fi> wrote:
On Tue, Sep 8, 2009 at 4:51 PM, Ivan Kassal<kassal@fas.harvard.edu> wrote:
>
> Dear group,
>
> In a stroke of genius, Alejandro pointed out to me that instead of
> tabulating the coffee scores, I can get you to tabulate the coffee scores,
> thus reducing my workload by 100%. So, there's now a piece of paper by the
> coffee machine where you can register your ratings. We're still on sample A.

Seeing other people's answers will affect the validity of the results
since there may be a social pressure to conform to the opinion of the
majority.
In other words, the only reliable method is to use a locked ballot
box. If no sample gets a clear majority of high ratings the ballots
are to be mixed with wet straw and burned in the kitchen microwave.
The black smoke that issues from the oven will then notify the
participants that the experiment needs to be repeated.


Ville

--
       Ville Bergholm
       ville.bergholm@iki.fi
       http://iki.fi/ville.bergholm/
       US mobile  +1-617-3869378
       EU mobile  +358-44-5722802
_____________________________________________
Aspuru-list mailing list
Aspuru-list@lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/aspuru-list