Hi Martin,
There are two strategies you can use. One is to create a boxcar regressor
for each condition that starts at stimulus onset and ends at the response.
Thus, the duration of each trial is different, depends on the RT, and is
explicitly modeled. Then you can do a contrast between the two regressors
to determine whether A activates more per unit time than B.
Alternatively, you can create a regressor that models all trials using the
individual RTs to create each epoch. Then create a second regressor that
only models condition A. Now you don't have to do an A-B contrast anymore.
The [0 1] contrast will identify all voxels that are active during
condition A but not B, and have a duration equal to the RT.
>If I understand correctly there are 2 possibilities that seem applicable
>here:
>
>1) use RT as a covariate. This strategy seems intuitive but my EVs are
>defined using RT data (i.e. in each trial I model the event time-locked
>to the response: each event is modelled as the 4 volumes leading up to
>the subject's response, where the RT is collected). Does this prevent me
>from using this strategy? I just have the intuition that it will end up
>soaking most of the action in both my reasoning tasks. Is this an
>incorrect intuition?
The problem with this strategy is that it assumes that all trials take the
same amount of time to complete, i.e. 4 volumes worth, which cannot be true.
>2) I understand you suggest, as an alternative, modulating the height of
>each event. This also seems a good possibility, but I will need guidance
>on some basic points:
>
> How do people typically transform arbitrary scales (e.g. here the
>RT, or any other measures of difficulty) into heights?
> What is a "reasonable" interval for event heights; so that I can
>guide the transformation of RT milliseconds into event height?
You point out the main criticism of this method i.e. that the transformation
of duration into intensity is arbitrary and uninterpretable. Best to keep
the temporal relationship between behavior and your regressors one-to-one.
cheers,
jack
|