[R] ANOVA "Types" and Regression models: the same?
Frank E Harrell Jr
f.harrell at Vanderbilt.Edu
Tue Mar 2 15:01:18 CET 2010
Sorry there were 2 typos in my note:
> John Fox's posting provided a lot of information. Briefly, the "Types"
> refer to whether effects are adjusted for all other effects in the model
> (Types II, III, IV) or no (Type I or sequential tests only adjust for
no -> not
> EARLIER terms in the model). See John's posting for definitions of
> Types II and III and for reasons to almost always use II or III.
II or III -> II over III
Sorry about that
More information about the R-help