Paul Poulain wrote:
Ben Ide a écrit :
Horowhenua Library Trust and a foundation specific to Koha win so long as we don't count the "Non Votes" -- the set of non completed results -- or we consider the polling a simple horse race with the winner of the first category receiving 1st place, the winner of the second category receiving 2nd place, etc. [...] Shouldn't we consider only the 1st choice ? (as, if it's possible to build a foundation using the 1st choice, then the 2nd choice won't be useful ). We need only one winner ;-)
Well, we should have declared the voting system beforehand. Sorry for being more than a bit overloaded at the moment and only mentioning that now. When I tried to look at the poll beforehand, I couldn't log in and didn't get time to resolve that before the vote started, so I didn't realise that it was using the same approach as the initial survey. I feel that the best approach now would be to run the complete ballots through the various systems, see how they come out and probably almost all will give the same results and it won't matter. I hope. I have a program here called voteengine by Blake Cretney. It's PD and I'm not sure where it was from (sourceforge?). It wants an input file of one ballot per line as space-seperated codes. Currently it looks like: E F B I L A L B C A K E which means voter 1 voted E as rank 1, F as 2, B, I, L, then A, while voter 2 voted L, then B, then ... If anyone can upload txt files of complete votes in that form, I'll run it through the engine. If not, I'll do it when I find time. Hope that helps, -- MJ Ray (slef) Webmaster and LMS developer at | software www.software.coop http://mjr.towers.org.uk | .... co IMO only: see http://mjr.towers.org.uk/email.html | .... op