Hi All,
I want to run a simple correlation analysis on my block design experiment.
The quandary I'm having is that if I am regressing one timeseries to another, and both timeseries have similar 'block-y' shapes, they are obliged to correlate to each other even if residuals from the regression are quite different.
I was thinking of using 3ddeconvolve and specifying my block design as a base regressor, but at the moment my seed has the block design too!
What should I do if I want to specify the convolved block design as a regressor of no interest?
-Prantik
