Skip to main content

View Diary: WOW! Simple wording change dramatically increases voter turnout (196 comments)

Comment Preferences

  •  You've scrambled up a bunch of diferent issues. (0+ / 0-)

    One issue is that if you're dealing with unusual populations, you can't be sure about how general the results will be. That had nothing directly to do with sample size. 15000 doesn't cure it, although it opens up some possibilities for guesstimating it.

    The other issue is simply estimating the effect in the type of population you've looked at. It's a simple calculation to get the statistical error values and the p-values. You can estimate it by hand, or read the paper to get the more accurately calculated stats.

    This was not a poll. it was a randomized controlled study.

    Randomly remembering context-free numbers from some class doesn't help. It makes me glad my wife emphasizes more context and core concepts to her 1800 stats students.

    Michael Weissman UID 197542

    by docmidwest on Wed Jul 20, 2011 at 06:59:28 AM PDT

    [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site