Venkatesh Rao@vgr·May 23, 2016Did they ask people who got rejected and gave up or people who chose a different medium?Quote TweetBenjamin Cook@DustyBowl·May 23, 2016"91%...say last paper improved through peer review". Peer review needs improvement but is not fundamentally broken. https://twitter.com/cshperspectives/status/734736157003685888…14
Benjamin Cook@DustyBowl·May 23, 2016Replying to @vgrlooks like they just sampled randomly from Thompson Reuters author database: http://onlinelibrary.wiley.com/doi/10.1002/asi.22798/abstract…21
Venkatesh Rao@vgr·May 23, 2016Replying to @DustyBowlThis is a bit like polling people who still subscribe to paper newspapers and concluding the business is doing fine.12
Benjamin Cook@DustyBowl·May 23, 2016Replying to @vgrNot quite-my understanding is they sampled from ANYONE who published, regardless of # times, whether they still do, etc.11
Venkatesh Rao@vgr·May 23, 2016Replying to @DustyBowlI don't think dead people and inactives would have a high resp rate. I still get occasional emails from that world that I ignore.1
Benjamin Cook@DustyBowl·May 23, 2016Replying to @vgrWell, you can't force people to respond who can't or won't. So seems like this is the best we can do.1
Venkatesh Rao@vgrReplying to @DustyBowlThere's techniques to construct proper samples, but they cost more in survey design/deployment. This q is just not worth effort.4:01 PM · May 23, 2016