Statistical test to compare TGMs of two groups

External Email - Use Caution

Dear MNE users,

I would like to ask for your help regarding the following:

I have two groups of participants (e.g. males/females) for which I have produced separate temporal generalisation matrices classifying different stimulus properties (e.g. perception of happy vs neutral faces). I would like to use a test to find out whether there are significant differences on the TGMs of the two groups, but I wasn?t able to find an example on that. Any suggestions?

Many thanks :slight_smile:

Ioanna Taouki

External Email - Use Caution

hi Ioanna,

I would use permutations testing.

if your chance level is 0.5 for all subjects (balanced classes or using ROC
AUC) you can
do a test rejecting H0 that is that the values have a mean of 0.5.