User's Guide‎ > ‎Analysis‎ > ‎

4 - Computation of Grand Averages

Computation of Grand Averages (for one condition and across subjects) 

-          Tools >>> Averaging Files

-          Input are all (eventually) interpolated *.eph (*To_160.eph or *To128.eph or so) of your subjects of one condition in a STRICT order (because this ordering determines everything you do afterwards in terms of paired statistics, fitting etc.)

o        E.g. S1C1, S2C1, S3C1…..

-          Note that there could be now no triggers in the signal anymore >>> in this case, the program does not automatically “see” the difference between the baseline and post-stimulus period

-          Thus, set “Time Interval” to “Fixed Time Frame” (number corresponds to the baseline (or origin, i.e. the stimulus onset) you chose for the single subject average, e.g. 50 TF)

-          Duration: set as e.g. for single_avr

-          Optionally: Apply baseline again for between-subject pre-stimulus baseline correction (from “origin - pre-stimulus period” to “origin”, e.g. –50 to 0 TF). Applying a pre-stim base line correction should be justified by your hypothesis and/or design*

-          Set “Average Reference” as all the following analyses will be done on average referenced data

-          Define output folder name for this condition’s averaging

-          Save output with average and epochs (and standard error), as the epochs of the grand average couls serve as inputs in the following statistics

o        Click “Save average” to create one (if chosen, baseline-corrected) grand.eph (which you need e.g. for segmentation and ERP visualization)

o        Click “Save epochs” to create new (eventually baseline-corrected) *.eph for each single subject again (i.e. the epochs entering the grand_eph, which you need for statistics, fitting and inverse solutions)




- Baseline correction may induce artificial alterations of the electric field topographies and/or GFPs. I you apply a baseline correction your automatic assumption is that the pre-stimulus period is identical for each subject and/or condition.  Yet, this might not be the case for your data (for example, pre-stimulus periods may be different in patients as compared to controls).