Re: [S] "Sums of squares", the taxonomy issue again.
Fri, 1 May 1998 10:14:31 +1200

> Everytime I hear this back and forth arguement about the Type I through
> Type IV sums of squares
> I start asking people whether or not they have ever read Milliken and
> Johnson's book
> "The Analysis of Messy Data Volume I: Designed Experiments" (1992).
> Invariably the answer is no.
> >Since they are both students of Franklin
> Graybill ( argueably
> the father of linear models) and they have an impressive record of
> publication in the field of linear
> models, I would highly reccomend it. Especially the discussion of Type I
> through Type IV sums
> of squares, which is a constant theme throughout the text.

I did look at one of Milliken and Johnson's "Analysis of messy data"
books (maybe it was volume 2). What their description of how to obtain
correct standard errors for differences between treatment means in
split plot designs (and more complex procedures are needed for
split-split plots etc) indicated to me was the futility of using the
ANOVA/GLM method, such as in SAS when there is any error structure. In
this case it was not the data that was messy but the method of

I was reminded of the comment by one of the founders of Genstat
analysis of variance (how it should be done, and essentially how it is
done in Splus aov) when asked why most Americans don't know about and
use the Rothamsted/Genstat concept of analysis of variance by strata of
variation: "the dominant nation syndrome".

Rod Ball

Dr Roderick D. Ball Ph 64-7-3475899
Statistician, Fx 64-7-3479380
New Zealand Forest Research Institute email:
P.B. 3020, Rotorua, New Zealand

reminded me

This message was distributed by To unsubscribe
send e-mail to with the BODY of the
message: unsubscribe s-news