Averaging Across ERPsets (Creating Grand Averages)

The ERPLAB > Average Across ERPsets routine is used to average together the data from multiple ERPsets.  That is, each bin from one ERPset is averaged with the corresponding bin from one or more other ERPsets.  The most common use for this is to create a grand average across subjects.  It is also used if the EEG data for a given subject were divided among different EEG files (e.g., because they were collected in different sessions) and then averaged separately into separate ERPsets.  This routine can be used only if all of the ERPsets being averaged together contain the same number of bins, channels, and sample points.

The GUI is shown in the screen shot below.  You can either specify a set of ERPsets that have already been loaded into ERPLAB and are listed in the ERPsets menu or a set of ERPsets stored in files on disk.  You can save the set of filenames in a list with the Save List option, and you can load in a set of filenames with the Load List option.

Ordinarily, each ERPset being averaged together receives equally weighting in the average that is created by this routine.  That is, the ERP waveforms in the separate ERPsets are simply summed together and then divided by the number of ERPsets.  However, there is an option for averaging in a manner that is weighted by the number of trials that contributed to each average (a “weighted average”).  Imagine, for example, that you were averaging together two sessions from a given subject, with each session stored in a separate ERPset, and bin 1 contained 10 trials in the first session and 90 trials in the second session.  If you enable weighted averaging, the average of bin 1 would be calculated as 10 times the waveform from session 1 plus 90 times the waveform from session 2, and then divided by the total number of trials (10+90).  This gives each trial equal weight in the final average.

<<Previous Section           Next Section >>


Back to Table of Contents