On Thu, Mar 4, 2010 at 9:03 AM, S Ellison <s.elli...@lgc.co.uk> wrote:
>
>
>>>> "John Fox" <j...@mcmaster.ca> 02/03/2010 02:19 >>>
>>There's also a serious question about whether one would
>>be interested in main effects defined as averages over the level of
> the
>>other factor when interactions are present.
>
> My personal take on this particular chestnut is that I often want to
> ask something about the relative size of the effects. If the so-called
> "main effect(s)" is/are very much larger than the interactions, one may
> be able to make generalisations which have practical use.
Sure. But if there is an interaction, main effect generalizations are
going to be less precise than generalizations based on simple effects.
I agree that there are some situations in which it makes sense to
ignore a small but significant interaction, but I think this should be
a rare exception to the rule "don't interpret main effects in the
presence of an interaction".
-Ista
If the effects
> are much of a size, there's nothing much to be gained by asking about
> "main effects".
>
> Mind you, that's probably a crit of significance testing as the be-all
> and end-all, rather than a problem with type I-III. Asking 'how big is
> it?' is a step beyond 'is it there?'.
>
> Steve E
>
> *******************************************************************
> This email and any attachments are confidential. Any u...{{dropped:18}}
______________________________________________
R-help@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.
