[R] Type IV sum of squares

(Ted Harding) Ted.Harding at nessie.mcc.ac.uk
Sat Mar 8 10:23:20 CET 2003

On 08-Mar-03 Bill.Venables at cmis.csiro.au wrote:
> I'm really puzzled to know what has caused the sudden rush on these
> kinds of questions.
>       [WNV]  My friendly advice is forget all that malarkey, and even
> forget about putting it all in one big analysis of variance table. 
> Calmly decide which null hypotheses you want to test within which outer
> hypotheses, fit each null and each outer model, separately if necessary,
> test them using anova( ) and report your results.

Well, those are welcome commonsense words, to one who never understood
the "malarkey" and indeed found "Type I", "Type II", "Type III" to be
totally enigmatic terms for what seemed to be pointless distinctions.

>       Hard as it may be to accept, that whole business of Type ? sums of
> squares is a gigantic red herring and a distinction without a
> difference. It would truly have been better if the concepts were never
> invented and certainly not inflicted on a naive and unsuspecting whole
> statistical generation.  This is one situation where curiously the
> reality is much simpler than it might seem from some of the propaganda.

I'm curious as to how it came about -- that current of history (which
seems to be SAS-related) seems to have passed me by, alive and conscious
though I believe I was at the time ...

What really bothers me about it (and other things) these days is that
a Procedure With A Name becomes a fetish[*] to be worshipped by rituals
performed with a computer. The result of "propaganda"? I'm thankful that,
with R, it's quite hard to be irrational.

[*] Chambers' definition is to the point: "fetish: An object believed
    to procure for its owner the services of a spirit lodged within it:
    something regarded with irrational reverence."

>       Trust me: I'm a statistician...

But of which Type ... ?

Best wishes,

E-Mail: (Ted Harding) <Ted.Harding at nessie.mcc.ac.uk>
Fax-to-email: +44 (0)870 167 1972
Date: 08-Mar-03                                       Time: 09:23:20
------------------------------ XFMail ------------------------------

More information about the R-help mailing list