> Can I replace the blocks I do not want to consider for analysis with 'NaN' s or
> something that does not affect the analysis?
Model those blocks of no interest as a separate regressor for their average effect (just as you would do in the typical individual subject regression analysis).
Gang
