A tale of two polls

Poll one wasn’t condusive to the right headlines and the journal adopted a policy of “that EOS should not accept summaries of opinion polls”.

Opinion poll two produced unambiguously the right answer, was published in EOS, and is splashed about.

Comments

  1. #1 Magnus W
    2009/01/20

    I guess it is to much to ask… to think that journalist could try to check the actual papers on the issue…?

  2. #2 JCH
    2009/01/20

    So 648 climatologists were prohibited from participating?

  3. #3 bigcitylib
    2009/01/22

    The Annan/Brown/Pielke survey had issues. And as far as I can tell, once you account for the issues, it gave the same answer as the Boran survey.

    [Everything has issues. But we discussed those at the time. The point now is the blatant double standards on the part of EOS, which is not just picking papers because it happens to like the interpreation of the results, but also lying to people about its policy to try to justify that rejection -W]

  4. #4 Raymond Arritt
    2009/01/23

    Those of us in the field already know the answer to the basic questions of these polls (most scientists accept that the climate has warmed, that humans have something to do with it, blah blah). I’ve probably said this before, but the more interesting question would be to look at who doesn’t respond to these polls and why — i.e., the effect of selection bias and whether the sample is truly representative of the population.

    I’ve declined to respond to several of these polls because I thought the questions were too ambiguous. Extrapolating from there, might we hypothesize that at least some who don’t respond are more concerned with nuance and ambiguity than those who don’t? Might we also ask whether those who have strong opinions on one side or the other are more likely to respond? And of course there are those who are just too busy. But even the “too busy” non-responders are systematically different from the responders.

    It would be interesting if one of the pollmakers could send a very brief note to the non-responders asking for just a word or two on why they didn’t respond. Granted many of the non-responders to the original poll probably wouldn’t respond to the followup. But perhaps enough of them could respond to produce some interesting results.

  5. #5 Michael Tobis
    2009/01/26

    Of course, those responding to the nonrespondents query would not be representative of the nonresponding nonrespondents…

    Consensus emerges in science by a process that is strikingly informal compared to how it is usually described. Perhaps it needs to be be codified somehow? Or perhaps that’s intractable.

    This idea of doing a poll seems somehow inappropriate; not everyone’s opinion in science carries comparable weight on comparable questions.