Xref: utzoo news.groups:15570 news.admin:7942 Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!tut.cis.ohio-state.edu!ucbvax!bloom-beacon!eru!luth!sunic!dkuug!freja!stodol From: stodol@freja.diku.dk (David Stodolsky) Newsgroups: news.groups,news.admin Subject: Re: Recent Voting Stats Summary: No votes meaningless Keywords: votes 2/3 rule stv newsgroup creation no Message-ID: <5026@freja.diku.dk> Date: 9 Dec 89 23:01:00 GMT References: <854@pmafire.UUCP> <855@pmafire.UUCP> <856@pmafire.UUCP> Organization: DIKU, U of Copenhagen, DK Lines: 28 For vote stats to be meaningful, an adequate sample must be available. Any voting method will pass some proposals that are bad and not pass some that are good. The recently presented vote stats contained a single proposal that did not pass. This certainly gives no basis for any statistical conclusion. And what conclusion one can draw, will depend in part upon whether one feels that the group that did not pass was a good proposal or not. This in turn will be determined by what period is covered by the stats. Thus, the conclusion is totally irrelevant to the guidelines used. And no conclusion about the current guidelines can be supported by such stats. This is not to indicate that the author of the stats rigged the results. One simply is in error to try to draw such a conclusion on the basis of such a small sample, especially if it is so unbalance as to have a single instance of one of the categories. Statistically speaking, there is no way to determine how large the measurement error is, in such a situation. However, the question of whether "no" votes are independent of the proposal could be answered, especially if one analyzes the data before it is summarized. If "no" votes are for the most part transmitted by the same people (supported by casual observation), then this would be further evidence for this hypothesis. If "no" votes are independent of the proposal, the same results could be achieve by eliminating "no" votes and changing the current guidelines to, for instance, "Yes > 130". -- David S. Stodolsky, PhD Routing: <@uunet.uu.net:stodol@diku.dk> Department of Psychology Internet: Copenhagen Univ., Njalsg. 88 Voice + 45 31 58 48 86 DK-2300 Copenhagen S, Denmark Fax. + 45 31 54 32 11