[R] ANOVA "Types" and Regression models: the same?

[Liaw, Andy]

If memory serves, Bill Venables said in the paper cited several times
here, that there's only one type of sums of squares.  So there's only
one type of "ANOVA" (if I understand what you mean by ANOVA).

Just forget about the different types of tests, and simply ask yourself
this (hopefully simple and straight forward) question: Which pair of
models when compared will answer the question you have at hand?  It's
not sufficient to just ask: "Is factor X significant?"  It depends on
what else is in the model you're entertaining.

I think it's high time to retire the archaic concept of the different
types of sums of squares.  IMHO they are the biggest red herrings in
Statistics.

Best,
Andy

From: Ravi Kulkarni
> 
> Hello,
>   I think I am beginning to understand what is involved in 
> the so-called
> "Type-I, II, ..." ANOVAS (thanks to all the replies I got for 
> yesterday's
> post). I have a question that will help me (and others?) understand it
> better (or remove a misunderstanding):
>   I know that ANOVA is really a special case of regression where the
> predictor variable is categorical. I know that there can be 
> various types of
> regression models commonly called "stepwise, add, remove...", 
> where one
> controls which predictors are added to the regression model 
> and in what
> order.
>   Is this what the various "Types" of ANOVA correspond to? I 
> mean that I
> think of my ANOVA as a regression model (a General Linear 
> Model) and the
> various ways of entering predictors as the various ANOVA "Types".
>   Hope that makes sense...
> 
>   Ravi Kulkarni
> -- 
> View this message in context: 
> http://n4.nabble.com/ANOVA-Types-and-Regression-models-the-sam
> e-tp1574654p1574654.html
> Sent from the R help mailing list archive at Nabble.com.
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 
Notice:  This e-mail message, together with any attachme...{{dropped:10}}