jerry ... how can you avoid them? generally, software won't let you and, if you are doing work for clients ... they won't think it is legit without them

also, if you fail to instruct students about them ... others will claim that you are not providing adequate instruction

so, i would say that we are in a real bind ...

try getting an empirical sort of article published ... or even considered ... without them

what if you wanted to use confidence intervals in your paper ... and decided that NO null hypotheses were necessary to make your points ... the editors would NOT let your article into their journal

there is a dominant ... clear ... editorial and publishing bias ... that dictates that you MUST talk about statistical significance ... you really have no choice IF you want to publish in refereed sources

that is NOT proof that p values are useful or valuable ...

i can say that when i look at a paper ... it is not the statistics that my attention is drawn to ... it is/are the method or methods they have used that produce the data/results

rarely can "bad" statistics kill you BUT, far too often, using bad methods for collecting data or ... losing control over your (experimental) conditions will do it faster than you can shout ... p value



At 04:20 PM 3/24/2003, Jerry Dallal wrote:
In some recent threads, many people have been critical of P values.
While I don't base decisions solely on P values, I find them
useful.  I use P values in my work.

So, I ask those critical of P values, "Do you use them in your
work?"  I'm not asking whether you are aware of them, but whether
you generate them and report them as part of your assessment of
data.  I do.  Do any of those who are critical of P values avoid
using them altogether? If so, what do you do instead?  This is not a
question about what we might like to do or what would be preferable
from a theoretical viewpoint.  I'm curious to hear what people
actually do when analyzing data.

Thanks!
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

_________________________________________________________
dennis roberts, educational psychology, penn state university
208 cedar, AC 8148632401, mailto:[EMAIL PROTECTED]
http://roberts.ed.psu.edu/users/droberts/drober~1.htm


.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to