[Coral-List] Statistics.
John Ware
jware at erols.com
Tue Jun 1 14:00:32 EDT 2010
Dear List Members,
While I am sure we all appreciate Richard Dunne's recent and learned
comments on statistical testing, let us not forget that what most
coral-reef scientists are trying to find out is something of real
significance.
Too often, we seem to replace "statistical significance" with
"significance". Clearly there can be statistically significant results
that have no real significance and one can obtain non-statistically
significant results when there is a very real significance of the test
variable (but the effect is not detected due to poor design or a limited
sample).
John
--
*************************************************************
* *
* John R. Ware, PhD *
* President *
* SeaServices, Inc. *
* 19572 Club House Road *
* Montgomery Village, MD, 20886, USA *
* 301 987-8507 *
* jware at erols.com *
* http://www.seaservices.org *
* fax: 301 987-8531 *
* *
* Past *
* Treasurer and Member of the Council: *
* International Society for Reef Studies *
* _ *
* | *
* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|~~~~~~~~~~~~~~~~ *
* _|_ *
* | _ | *
* _______________________________| |________ *
* |\/__ Untainted by Technology \ *
* |/\____________________________________________/ *
*************************************************************
If you are a coral-reef scientist and you are not a member
of the International Society for Reef Studies, then
shame on you.
Become a member of the International Society for Reef Studies
http://www.coralreefs.org
More information about the Coral-List
mailing list