Comparing ERPs for multiple sessions with a single participant

đź’¬ Support & Discussions
,
1

Hi everyone,

I am running an ERP experiment where I will be conducting multiple sessions across different days for a participant (different conditions/ERPs for each session). How can I best address potential noise issues between the different sessions (for example, different electrode skin contact etc.). I will also be computing difference waves for the different ERPs so that will also be an issue if I don’t address it properly.

So far based on reading some threads, I understand that when performing ICA correction, it must be done for each separate session, and not on concatenated data. Is there anything else that I should be doing and watching out for?

Kind regards,
Nathan

2

You could try to calculate a metric for ERP data quality for each of the sessions and check whether there are significant differences or not. See for example: A New Metric for Quantifying ERP Data Quality — ERP Info

In any case, you’ll have to discuss this issue as a potential source for differences between signals.

1 Like
3

Thanks! That’s a great idea!

1 Like