Dear SPM experts,
I have a problem to model incorrect trial. I followed past discussion 'regressor versus condition for incorrect responses' in this forum. In the end, people suggested that the incorrect trials should be modeled into one regressor so that there won't be more unexplained errors in the model. Moreover, they suggested that when there were no incorrect trials in one given session. Just take one last TR to pretend as an incorrect trial in the regressor. in this way, there won't be a problem of unequal regressors across sessions. However, I wonder how this would influence the results to have a 'non incorrect trials' in the regressors for incorrect trials. Also, I wonder whether there are other legitimate way to deal with unequal incorrect trials nowadays. Thanks!
Veda
|