Dear SPM'ers,
We performed an experiment with the aim of investigating the
difference in brain activity as a result of two types of visual
feedback during a hand-motor task. We scanned the two conditions in
different runs, in a blocked design (task conditions intermitted with
rest-periods). The motor task was identical in the two conditions; the
only difference was the type of visual feedback the participants got.
The problem was that we could not change this type of feedback during
a run, which had the result that subjects also got different visual
feedback during the rest blocks of the two different runs. We mainly
wanted to investigate the effect of the different visual feedback on
the performance of the motor task, but as this feedback was also
different during the rest blocks of the two conditions, we might have
lost some activation.
A reviewer has now suggested that in order to deal with this, we
should do a for a two factorial analysis ((activity vs. rest) x
(feedback condition 1 vs. feedback condition 2)). However, we feel we
can not model the rest periods as well as the activity periods, as we
would me more or less be modeling the same thing twice, and thereby
overfitting our model.
Anyone any thoughts on this, or suggestions for an alternative type of analysis?
Thanks,
Marian
|