Hi there,

Just want to get an opinion on this.

I have been using GAM functions on data from an event related design with stimulus durations of 3 seconds.

First, is 3 seconds too long for use of the GAM function? I know the documentation mentions short presentation times as being between 0-2 seconds. If it is too long, which function might be preferable?

Second, I realize I modeled fixation crosses as one of the trial types in the gltsym. These however are NOT presented for 3 seconds - they are only on screen for about .5-1 seconds between each stimulus presentation. I realize now that this violates the recommendation of having conditions with the same stimulus time. Would you say this might be issue?
If not, great! But if so, would you recommend leaving fixation crosses out of the model to serve as implicit baselines? (That way all modeled trials have the same stimulus length..) Or rather, would it be better to use a different basis function while keeping fixation crosses included in the gltsym?