Re: One or Two tails: II ?
- From: "Kevin E. Thorpe" <kevin.thorpe@xxxxxxxxxxx>
- Date: 5 May 2006 05:35:31 -0700
Thom wrote:
*** Marvin Zelen dismisses one-sided tests in another way--he finds them unethical! Hisargument is as simple as it is elegant. Put in terms of comparing a new treatment to
standard, anyone who insists on a one- tailed test is saying the new treatment can't do worse
than the standard. If the new treament has any effect, it can only do better. The method of
analysis should be part of a study's design. If investigators are prepared to justify the use
of a one-tailed test at the start of the study, then it is unethical not to give the new
treatment to everyone!
I agree with Zelen's argument in that context, but there are other
limited contexts which make some directional tests sensible and
ethical. For example in a replication and extension one might introduce
a manipulation/intervention expected to influence effect A but not
effect B. If effect B is well established in the literature one might
use the more powerful one sided test to establish that your
participants demonstrated effect B. For effect A one would use a two
sided test.
Thom
One other situation where I feel you can reasonable justify a
one-sided test is where you would only administer a treatment
if it was superior. I'm thinking here of a surgical intervention
compared to standard medical care. In order to justify exposing
a patient to the risks associated with a surgery, it _better_ be
superior to the medical care they would receive, especially
when the surgery will be in addition to the standard medical
care.
That's how the argument goes and I'm sure others could
produce sound arguments in favour of two-sided tests in
this situation.
--
Kevin E. Thorpe
Assistant Professor, Department of Public Health Sciences
Faculty of Medicine, University of Toronto
.
- Follow-Ups:
- Re: One or Two tails: II ?
- From: Anon.
- Re: One or Two tails: II ?
- References:
- One or Two tails: II ?
- From: \"Luis A. Afonso\"
- Re: One or Two tails: II ?
- From: \"Luis A. Afonso\"
- Re: One or Two tails: II ?
- From: Thom
- One or Two tails: II ?
- Prev by Date: Re: Maximizing sum of squares
- Next by Date: Re: One or Two tails: II ?
- Previous by thread: Re: One or Two tails: II ?
- Next by thread: Re: One or Two tails: II ?
- Index(es):
Relevant Pages
|