Posted by sean drummond on October 29, 2002 at 13:44:03:
I know this is an odd question, but I am wondering if 3dRegAna can deal with a within subjects design (even though linear regessions typically cannot).
The reason I ask is that some of the examples in the manual appear to do that. For example, Example 1 has several data sets from Fred, but I don't know if this was just convienent to type or if it was meant to imply that Fred was scanned several times.
My particular application would be:
I have a task with 3 levels of difficulty and would like to examine the Level x Performance interaction. If not 3dRegAna, any suggestions for doing this? 3dAnova can't handle this well becuase performance is not a categorical variable.
Thanks.
sean