lme() direction

Mike Lawrence <mike <at> thatmike.com> writes:

And if I decided to ignore the "type" variable altogether and simply
use the continuous "valence" variable, this is what I'd use?

summary(lme(
	fixed = rt~valence*color
	, data = a
	,random = ~1|id
))

I also have continuous luminance measurements of each color that vary
from participant to participant (we used different monitors). If I
were interested in luminance *instead* of color, would the following
be appropriate, or do I need to do something special to account for
the fact that each participant has different values of luminance?

summary(lme(
	fixed = rt~valence*luminance
	, data = a
	,random = ~1|id
))

Both might be appropriate, but be sure to understand the implications.
Both valence and luminance now are to be interpreted as slopes.
Since slope-interactions are a bit awkward to interpret, I would 
prefer to start with

fixed = rt~valence+luminance
fixed = rt~valence*luminance-valence:luminance

Both mean the same, the latter is ridiculous here, but may be useful
when you have more terms to remove higher ones. Also look at the 
meaning of ^2, and the difference of I()^2.

You might have a look at stepwise procedures in stepAIC, even if
in my field there are good reasons to avoid this type of model
selection.

It can be tricky to explain slope interactions in papers, but it's
probably easier in psychology where people are ready to accept
models than in medicine, were everything beyond a t-test is frowned
upon by reviewers.

Dieter