On 17 Mar 2004 at 8:19, Phillip Good wrote:

> The burden of proof remains on you [EMAIL PROTECTED] --

Really? You are attacking a workhorse of statistics, but apparently 
you know something none other of us knows 1) - all opf this is 
misled, 2) you have a better method. 

Until you tell us about your better method I stay with models 
and with George Box' "All models are wrong, but some are usefull"

> unless what
> you intended to say was:
> 
> As we know,
> There are known knowns.
> There are things we know we know.
> We also know
> There are known unknowns.
> That is to say
> We know there are some things
> We do not know.
> But there are also unknown unknowns,
> The ones we don't know
> We don't know.
> 

You said that, not me.  

Kjetil


> [EMAIL PROTECTED] wrote:
> On 16 Mar 2004 at 12:26, Phillip Good wrote:
> 
> > I was unaware that maximum likelihood had any desirable properties
> > except in the case of normally-distributed random variables where
> > the max likelihood approach leads to estimators that are desirable
> > for entirely different reasons.
> > 
> 
> Could you please explain what in your opinion is wrong with likelihood
> methods, which in effect makes up the workhorse of todays applied
> statistics, not only for normal models, but for instance in
> generalized linear models and a lot of others?
> 
> What is your opinion on the likelihood principle, as referenced in a
> text I referenced in another letter today?
> 
> Kjetil Halvorsen
> 
> > Phillip Good
> > 
> > Paul Allison wrote:
> > On April 23-24, 2004, I will be offering a two-day course in 
> > Philadelphia on Missing Data .
> > 
> > After reviewing the strengths and weaknesses of conventional
> > methods, the course will focus two newer methods, maximum likelihood
> > and multiple imputation, that have much better statistical
> > properties. These new methods have been around for at least a
> > decade, but have only become practical in the last few years with
> > the introduction of widely available and user friendly software.
> > What's remarkable is that these methods depend on less demanding
> > assumptions than those required for conventional methods. At
> > present, maximum likelihood is best suited for linear models or
> > log-linear models for contingency tables. Multiple imputation, on
> > the other hand, can be used for virtually any statistical problem.
> > 
> > Multiple imputation will be illustrated with the new MI procedure in
> > SAS. Maximum likelihood will be implemented with structural equation
> > modeling software (either Amos or LISREL).
> > 
> > The text for the course will be my "Missing Data" published by Sage
> > in 2001.
> > 
> > For complete details, go to www.ssc.upenn.edu/~allison
> > 
> > .
> > .
> > =================================================================
> > Instructions for joining and leaving this list, remarks about the
> > problem of INAPPROPRIATE MESSAGES, and archives are available at: .
> > http://jse.stat.ncsu.edu/ .
> > =================================================================
> > 
> > Phillip Good
> > http.ms//www.statistician.usa
> > "Never trust anything that can think for itself if you can't see
> > where it keeps its brain." JKR
> > 
> > Do you Yahoo!?
> > Yahoo! Mail - More reliable, more storage, less spam
> 
> 
> .
> .
> =================================================================
> Instructions for joining and leaving this list, remarks about the
> problem of INAPPROPRIATE MESSAGES, and archives are available at: .
> http://jse.stat.ncsu.edu/ .
> =================================================================
> 
> Phillip Good
> http.ms//www.statistician.usa
> "Never trust anything that can think for itself if you can't see where
> it keeps its brain."  JKR
> 
> Do you Yahoo!?
> Yahoo! Mail - More reliable, more storage, less spam


.
.
=================================================================
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