Hi Ged.
Thanks for the response. This does help quite a bit for getting correlation coefficients and will help tremendously, thanks. However, I would also like to be able to graph the results as well. This would require the adjusted values I believe. Therefore, it would be nice to be able to output the adjusted eigenvariate. This would also be useful if you would want to create a bar graph for example comparing the means of two groups controlled for the covariates.
Thanks.
-John
--- Ged Ridgway wrote:
John D. West wrote:
--- Start of quoted text:
I am currently running a multiple regression in SPM5 to find the a
correlation between my image data and another value and have also
included 2 covariates of no interest. Once I get the results I can
extract the 1st eigenvariate for a cluster by using the VOI tool,
but this (as far as I can tell) caluculates the 1st eigenvariate of
the data not adjusted for the 2 covariates. Therefore, if I use
the matlab corrcoef function I get an r and p value that is not
representitive of the SPM results. Can anyone tell me how I can
get the adjusted eigenvariate or suggest another way of getting the
Pearson correlation coefficient for the 1st eigenvariate for a
particular cluster?
--- end of quoted text ---
Hi John,
I'm not sure how to get the adjusted eigenvariate, but can you use Matlab's partialcorr instead of corrcoef? I think if you pass in your two nuisance covariates as the Z matrix then you should get the partial (Pearson's) correlation coefficient.
Hope that helps,
Ged.
--- end of quote ---
|