Hi,
Relative NEWBIE question.
I am involved in analysing data using SPM. The two experiments involved
visual stimuli.
1)
The first experiment has two conditions, A and B, each A and B contain 3
scans each. There are 12 A's and 12 B's. The ordering is fully random. I
presume that I just need to specify to SPM this order of EPOCHS (e.g. 1
2 1 2 2 1 1 ... etc). In general A is the experiment and B is the
control. I am specifying 2 contrasts: 1,-1 and -1,1. Does this seem
correct?
2)
The second experiment has 4 conditions, A, B, C, D. Condition A and C
are experiment and B and D are controls. In general B is the control for
A and D is the control for C. However a much more sophisticated analysis
will take place later. Again the order of A,B,C, and D are pretty random
through out the scan. Again I can specifiy the epoch as (1 4 3 1 2 ...
etc) ( we have a total of ~200 A,B,C and D's together) and each
EPOCH again has 3 scans each. For the contrasts I am in the dark how to
specifiy. I would like to view A, B, C, and D as lying along basis
vectors of activation, yet how do I tell SPM that C and D are pretty
much othogonal (activation, hypothesised that is) to A and B; and that A
and B are hypothesised to lie along the same basis vector of
activation/deactivation?
Any guidance from advanced SPM users would be highly appreciated.
Thanks alot,
Robert Welsh
--
Robert C. Welsh, PhD
Department of Radiology
University of Michigan
[log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|