I am really interested in the approach of fusing together EEG + fMRI to explore dynamics via RSA. However, the dataset I have is for the same task, but collected non-simultaneously. Would this be substantially less meaningful to relate the RDMs for EEG and fMRI, when the recordings were taken during separate sessions?
Not at all, you can still compare the RDMs.
Fantastic - great that creating a RDM for the EEG data at each time-point and relating this to fMRI can still be valuable. Do you think variance in the exact time the task started between the different EEG trials will matter a lot? Each trial was 2 seconds, however the task is inner-speech generation, so there will likely be some variance in the timecourse
I am not an expert for these kinds of analyses, but I suspect that yes – that’ll have an impact and you’ll need to think how to best align your trials. Sorry but I can’t really give a lot of guidance beyond the very basic remark that you can compare RDMs between fMRI and EEG data from the same task if they were recorded in separate sessions.