3 Responses to Potpourri

  1. Statistical Significance August 13, 2012 at 4:20 pm #

    I think I’d be willing to bet money that:

    “an unusually large number of psychology findings are reported as “just significant” in statistical terms.” — “- there was a glaring bump in the distribution between .045 and .050.” — “an uncanny abundance of reported results just sneaked into the region of statistical significance.”

    Would be the same in Political Science. This isn’t surprising. The moment that .05 becomes “statistically significant” and that means something is the moment models get manipulated to hit the mark.

  2. Eric L. August 13, 2012 at 5:35 pm #

    Stat. Sig,

    You are correct. If you try to make the bet, don’t let your mark read the work of Gerber, Malhotra, et al. on the political science literature:

    http://www.qjps.com/prod.aspx?product=QJPS&doi=100.00008024
    http://apr.sagepub.com/content/38/4/591.short

    as well as for sociology:

    http://smr.sagepub.com/content/37/1/3.short

  3. Eric L. August 13, 2012 at 5:38 pm #

    p.s. And now that I have looked at the psych paper, it’s too bad they didn’t cite Gerber et al.