[FieldTrip] Interpretation of Sensor Level Clusters and Relating to Cortical Sources

[Paul Dhami]

Dear FieldTrip community,

I have two time-frequency datasets, and I am interested in testing whether
there is any significance difference between them, and if so, finding the
potential cortical sources of such differences.

At the sensor level, I have used the cluster-based permutation method to
establish that there is indeed a significant difference between the two
datasets.

My question is if I can use the features of the cluster to predefine my
source analysis, or would that involve the incorrect interpretation of the
cluster method?

For example, let's say the largest cluster was found to exist in the range
of 100 - 300 ms in the frequency band alpha (8 - 12 Hz). Would it then make
sense to limit my source analysis to this time and frequency window? Is
this an incorrect approach? Or correct but perhaps too conservative?

Any insight would be much appreciated.

Best,
Paul
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20200413/e1f7faf0/attachment.htm>