On Mon, 24 May 2004 15:23:25 +0200, "Jeroen Jansen"
<[EMAIL PROTECTED]> wrote:

> Hi All,
> 
> Currently I am looking for references to articles (and books) about ANOVA
> models for experimental designs. I would like to find a nice discussion
> about the pipeline: experimental question --> experimental design --> ANOVA
> model that is appropriate to answer the question.
> Does anyone have any suggestions ?
> 
> Thanks
> 
> Jeroen Jansen
> Biosystems Data Analysis group
> University of Amsterdam

It sounds to me as if you are asking for something vastly
over-simplified.  Perhaps I am too influenced by the 
metaphor of a 'pipeline'  since that is far from the way 
that it seems to me.  Elementary textbooks on *statistics*, 
however, tend to follow that model, since they are directed
to the goal of showing how to do the analysis.

In my head, the problem might be more like mapping a 
road trip between known locations, when you don't know
beforehand  which roads are good and which bridges may
be down -- You usually have to look closely at the 
question (what already has been done? criticisms?) and 
the data that are possible for answering it (what sources?
reliability? validity?).

A textbooks on experimental design will illustrate, as a focus
or not, many of the standard problems in a given area.  
Psychological and engineering research can have 'error terms'
that are at extremes in magnitude and meaning; and a textbook
in neither area pays attention to time series to the extent that 
econometricians do.

 

-- 
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
.
.
=================================================================
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