Permutation abs mean test; average selected signals

Hello Brainstorm,
I have a question about how the "average selected signals" option works in the permutation absolute mean test.

Suppose I have 13 cortical auditory evoked potentials (CAEPs) in group A and 21 CAEPs in group B.
I would like to compare the activity of a cluster of channels within each CAEP between the two groups. I am guessing there are two approaches to specifying this cluster of channels. Either I specify the relevant channel names in the sensor types or names, OR I can first make a cluster and extract the relevant channel data using extract cluster time series.

My first question is properly understanding what "average selected signals" means.
Suppose I have the below CAEP in panel A with the relevant channels selected. If I extract these channels I would get panel B and if I then "average selected signals" will I get panel C? (in this case this was made using the "mean" cluster function which averages across the channels).

If panel C is indeed how "average selected signals" work, I have the second question. Let's say that instead of making a cluster and extracting cluster time series I specified the channel names like the below screenshot. If I select "average selected signals" will it first average these specified signals within each CAEP and compare that between the two groups?

I ask these questions because I wasn't sure if by selecting "average selected signals" Brainstorm would average across channels meaning that for channels with opposite polarities they would cancel each other out.

Yes, this is how it works. The signals are first averaged across sensors, (for each file) and then the permutation test is carried out with those averages. Indeed, you will see that the channel in the resulting stat file is called AVG

Understood thank you for the answer!