Comparing Groups of Data
Just a couple questions on comparing groups:
If I wanted to compare the data across two independent groups, should I preprocess one group of data separately from the other group? Or should I preprocess them all together to maintain consistency?
Also, in a previous lecture it was mentioned that to compare the data and run statistics (currently), we would extract the values into a group results file. How would you decide which values to compare?
If they have the same probe geometry, that is the same SD file, then you can process them either together or separately. If the probe geometry is different then you must analyze them separately. Processing them together does help ensure that you use the same processing options for both.
When comparing results across conditions or groups, we typically compare the mean amplitude of the HRF averaged over time scale of a few seconds around the peak of the HRF response. This is typically the last few seconds of the duration of the trial. You can get Homer to export this information for you from the menu "file" -> "Export" -> "Subjects HRF Mean"
I hope this helps