Notice: On March 31, it was **announced** that Statalist is moving from an email list to a **forum**. The old list will shut down on April 23, and its replacement, **statalist.org** is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
Marcello Pagano <pagano@hsph.harvard.edu> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
Re: st: RE: RE: RE: one-tailed tests |

Date |
Thu, 08 Jul 2010 18:47:38 -0400 |

Can't remember if he ever published it. m.p. On 7/8/2010 6:24 PM, Lachenbruch, Peter wrote:

Once upon a time I suggested we do the two-tailed test with 0.01 level at the 'uninteresting' hypothesis and 0.04 at the one that would lead to approval. It got nowhere. Tony Peter A. Lachenbruch Department of Public Health Oregon State University Corvallis, OR 97330 Phone: 541-737-3832 FAX: 541-737-4001 -----Original Message----- From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Airey, David C Sent: Thursday, July 08, 2010 2:25 PM To: statalist@hsphsun2.harvard.edu Subject: st: RE: RE: one-tailed tests . Exactly. We might expect/predict/hope a drug to perform better than a placebo or other gold standard, but we need to be prepared for the drug to do harm. You might want to follow up both possibilities. A directional hypothesis does nothing to prepare you for the other direction. Also, experiments may not be simple, but may allow for complex outcomes (interactions) that may not be easily predicted, but you may wish to follow up either direction for the outcome. While we are on this topic, there is discussion on the net about differences between a directional hypothesis and the use of one or two tailed tests. They don't seem to be the same thing, exactly. Here's a fun article on one-tailed tests. <http://www.bio.sdsu.edu/pub/stuart/2009MisprescriptionOneTailed.pdf>In some cases, one must make a decision that can be defended legally. I am thinking about drug approval. In this case, the FDA wants a two-tailed test (I didn't agree wholly while I was there as I think the theory should tell you the direction), so if it's harmful, they can be very strict on future trials. Tony* * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

* * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**Follow-Ups**:**RE: st: RE: RE: RE: one-tailed tests***From:*"Lachenbruch, Peter" <Peter.Lachenbruch@oregonstate.edu>

**st: Popularity of R, SAS, SPSS, Stata...***From:*Jose Benuzillo <Jose.Benuzillo@hsc.utah.edu>

**References**:**st: RE: RE: one-tailed tests***From:*"Airey, David C" <david.airey@Vanderbilt.Edu>

**st: RE: RE: RE: one-tailed tests***From:*"Lachenbruch, Peter" <Peter.Lachenbruch@oregonstate.edu>

- Prev by Date:
**Re: st: RE: one-tailed tests** - Next by Date:
**st: Popularity of R, SAS, SPSS, Stata...** - Previous by thread:
**st: RE: RE: RE: one-tailed tests** - Next by thread:
**st: Popularity of R, SAS, SPSS, Stata...** - Index(es):