Dear SPMers,
We would like to know if there is a limit to the number of conditions that
can be entered in an event-related analysis. In the archives, we have found
a post of someone who claimed to have used 50 conditions, but most experienced
users here tell us that they never had more than nine conditions, and that they
would recommend not using more than that.
We would like to enter 15-20 conditions into one model, 12 of which would be
conditions of interest. Does that still make sense and, if so, how many trials
per would be necessary to meaningfully estimate such a model?
Thanks!
Björn Schott
--
Bjoern H. Schott, MD
Dept. of Neurology II
University of Magdeburg, Germany
email: [log in to unmask]
www: http://neuro2.med.uni-magdeburg.de/~bschott
|