Dear SPMers,
I'm looking for a simple explanation of covariate interactions
in terms of their regression slopes. Say you have two groups
scanned over two conditions and have two covariates (one per
group). So, if we choose condition x covariate interactions we
end up with a design matrix with the first two columns
corresponding to the two conditions in group 1, the second two
columns corresponding to the two conditions in group 2, and the
next four columns containing the regressors.
it is my understanding that contrast [0 0 0 0 1 -1 0 0] would
show me those areas where the regression slope of group 1
condition 1 is "more positive" than the slope of group 1
condition 2. is this correct?
if so, then contrast [0 0 0 0 -1 1 0 0] would show me those
areas where the regression slope of group 1 condition 1 is "more
negative" or "less positive" than the regression slope of group
1 condition 2?
In other words do these contrasts define relative slopes of the
regressors or absolute slopes( where 1 is a positive slope and
-1 is a negative slope)?
What would contrast [0 0 0 0 1 -1 -1 1] suggest?
Thanks for your help.
Nathan White
=====
Nathan S. White
[log in to unmask]
__________________________________________________
Do You Yahoo!?
Get personalized email addresses from Yahoo! Mail
http://personal.mail.yahoo.com/
|