Is there any reason not to use 3dDeconvolve to estimate responses associated with blocked designs? Most descriptions of blocked design analyses I've seen involve using waver to create a reference wave form. I understand that it's not necessary with blocked designs because the responses don't need to be deconvolved, but I'm not sure how to create the same kinds of stats and graphs (e.g., average response for each condition, % signal change, ROI stats) with blocked data. So while it's not necessary to estimate IRFs with blocked data, is it inappropriate to do so?