Re: st: RE: Bargraphs, order, SD

[Ronan Conroy <rconroy@rcsi.ie>]

On 11 Meith 2007, at 12:00, Nick Cox wrote:

> Biologists in particular seems overly fond of just
> showing means (+/- sd, se, or constant * se) in what
> Stata user Paul Seed has called detonator plots.
>
> A dotplot with means and sds added (see -dotplot-)
> shows far more information.
It's a lazy habit, and one which has been opposed by the guidelines  
of the International Committee of Medical Journal Editors since their  
initial publication in 1979. The guidelines state:

Describe statistical methods with enough detail to enable a  
knowledgeable reader with access to the original data to verify the  
reported results. When possible, quantify findings and present them  
with appropriate indicators of measurement error or uncertainty (such  
as confidence intervals).

Note that the appropriate indicator of measurement error is the  
confidence interval. There is also disconcerting research which shows  
the extent to which error bars are misunderstood by readers.

http://psyphz.psych.wisc.edu/~shackman/belia_PsychMeth2005.pdf

Does anyone know of a serious reason for putting error bars in  
graphs? The one I always get is "but the journal will expect it" or  
"but everyone does it", neither of which is a defense.

=============
Ronan Conroy
rconroy@rcsi.ie
Royal College of Surgeons in Ireland
+353 (0)1 402 2431
+353 (0)87 799 97 95
http://www.flickr.com/photos/ronanconroy/


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/