Statistics time window based on ERP peaks

1

Hi

Is it possible to adapt the script for the permutation tests so that the time window of interest is selected based on individual peaks rather than one time window across all participants (to adjust for individual variation in latencies)? Possibly by using the process "Extract>Find maximum in time"? For example if I want to assess the P3 component between two groups then the time window would be the latency of the peak +/- 100 ms for each individual rather than the same time points (i.e. 250-500 ms) for all individuals.

1 Like
2

No, this is not possible the averaging and testing processing would only compare strictly identical time points.

One solution for you could be to:

  1. locate the peak in each average,
  2. add an event at the corresponding latency,
  3. epoch again the data based on this event (Import MEG/EEG).

This way, the peak would correspond to t=0s everywhere, and all the files would have the same number of time points before and after this peak.

3

Thanks Francois