On Sat, Aug 28, 2010 at 10:17:20AM -0700, Dave CROCKER wrote: > We really need to get these surveys produced by someone with training in > survey design. Or stop using the surveys. It isn't clear to me that the surveys are a good idea at all, not just because of sampling biases and survey design issues, but because they may lead to apparent statements of preference that don't conform to the general principles that ought to undergird IAOC decisions. Surveys are a good way to make a decision look democratic, but there are well-known and serious problems with collective preference expression. (If you don't believe this, you might start by looking up Kenneth Arrow, whose exploration of this issue is the most famous. There's a large body of work around this general topic, however.) If we have general principles for venue selection (for instance), then asking people to complete a survey is not the right thing to do: either a given site can be shown to conform to the venue selection principles, or it can't. If it does conform, then a survey might actually reveal a collective "preference" to go elsewhere, but that would not be a rational preference. (I wonder in fact whether the decision to go to Québec will turn out this way.) A -- Andrew Sullivan ajs@xxxxxxxxxxxx Shinkuro, Inc. _______________________________________________ Ietf mailing list Ietf@xxxxxxxx https://www.ietf.org/mailman/listinfo/ietf