BGonline.org Forums

Nice Rule-of-thumb

Posted By: Chuck Bower
Date: Saturday, 12 June 2010, at 2:06 a.m.

In Response To: OLM Th 06/10/10 ROLLOUT (Jason Lee)

Assuming Gaussian stats, if two confidence intervals have the same width and their endpoints coincide, the j.s.d. is 2.8.

And even if you don't care about that, isn't this result pretty surprising?

Messages In This Thread

• OLM Th 06/10/10
  Jason Lee -- Thursday, 10 June 2010, at 5:16 p.m.
  • OLM Th 06/10/10
    Casper van der Tak -- Thursday, 10 June 2010, at 5:18 p.m.
    • OLM Th 06/10/10
      Daniel Murphy -- Thursday, 10 June 2010, at 5:24 p.m.
    • OLM Th 06/10/10
      Stuart Katz -- Thursday, 10 June 2010, at 5:34 p.m.
  • OLM Th 06/10/10
    Jason Lee -- Thursday, 10 June 2010, at 5:32 p.m.
  • OLM Th 06/10/10
    neilkaz -- Thursday, 10 June 2010, at 5:49 p.m.
  • OLM Th 06/10/10
    Neil Robins -- Thursday, 10 June 2010, at 6:33 p.m.
  • OLM Th 06/10/10
    Fabrice Liardet -- Thursday, 10 June 2010, at 8:04 p.m.
  • OLM Th 06/10/10
    misja -- Thursday, 10 June 2010, at 8:27 p.m.
  • OLM Th 06/10/10
    Timothy Chow -- Thursday, 10 June 2010, at 10:22 p.m.
  • OLM Th 06/10/10 ROLLOUT
    Jason Lee -- Friday, 11 June 2010, at 3:33 p.m.
    • Nice Rule-of-thumb
      Chuck Bower -- Saturday, 12 June 2010, at 2:06 a.m.
      • Nice Rule-of-thumb
        Grigorij Perel -- Saturday, 12 June 2010, at 9:22 a.m.