Time index for input data

[Trang0412]

Hello,
I am now trying to use mTRF toolbox for auditory related EEG signal data.
I'm quite confused about the time index of input.
How can I identify the time index of EEG as input for mTRcrossval or mTRFtrain?
For example, we got the EEG data recorded from -500ms to 2500 after the stimuli onset. In this case, which ranges of data should I put as input for the model? Or just put the signal recorded from 0ms after stimuli onset time?
If in this case, then we can not get results at -200ms time lag, is it right?

[mickcrosse]

Hi,
From your description, I'm assuming that your EEG trials are 3 seconds long, correct? And your stimuli are less than that? The time-lags you input should reflect a window of temporal delays between your stimulus and EEG response - think of the timecourse of an ERP for example. So a safe bet would be to use something like tmin=0 and tmax=300 (milliseconds). Make sure that the onset of your stimulus corresponds to the start of the EEG data though (so I think you would remove the first 500 ms of EEG?). Does that make sense?
Mick