🚀 go-pugleaf

>Well, it seems to me that when you use an ANOVA, you engage in classic
>ANOVA interaction thinking, otherwise you would use another technique.

I wish that was true.  Unfortunately, I don't think it is.  I have seen far
too many graduate programs whose idea of teaching statistical analysis and
experimental design is to roll out, in order, Chi Squares, T-tests, and
ANOVA.  Occasionally folks graduate to MANOVA.  Rarely is much more than
handwaving done over regression.  Most people don't engage in class ANOVA
interaction thinking because they have selected it.  They select it because
its all they know.

>The whole point to the ANOVA is the conditionality of the statement.  If
>you have an interaction, you can't make an unconditional statement.

I don't agree.  Interactions are not subsidiary effects.  They are effects
in and of themselves which, when shaped correctly, can be interpreted
independently of the main effects.  If all one does with an interaction is
to say that you can no longer make an unconditional statement about
the main effects, one probably hasn't done a complete analysis.

Davis

Snailmail..........................Davis A. Foulger
Internet: FOLGER@WATSON.IBM.COM    IBM T.J. Watson Research Center
Prodigy:  XFRR20A                  P O Box 218, Yorktown Ht, NY 10598