2.11: Review of Processing Steps

You did a lot of exercises in this chapter, and they’re probably beginning to blur together in your mind. So let’s review the steps:

• Load the EEG data using the EEGsets panel in the EEG tab
  • Look at the dataset to verify that the EEG and event codes looked okay
• Filter out the low-frequency drifts in the EEG using the Filtering panel in the EEG tab
• Add an EventList to the EEGset using the EventList panel in the EEG tab
  • Verify that we had the correct number of occurrences of each event code
• Use BINLISTER to assign events to bins using the Assign Events to Bins panel in the EEG tab
  • Verify that we had the correct number of trials in each bin
• Extract fixed-length epochs from the continuous EEG and perform baseline correction using the Extract Bin-Based Epochs panel in the EEG tab
• Test each epoch for artifacts using a simple voltage threshold using the Artifact Detection panel in the EEG tab
  • Verify that the appropriate epochs were marked and see how many artifacts were detected
• Average together the artifact-free epochs, separately for each bin using the Compute Averaged ERPs panel in the EEG tab
  • In the summary window that appears, verify that we had the correct number of trials in each bin and view the data quality summary
• View the averaged ERP waveforms in the ERP tab
  • Create a custom channel layout using the Plot Settings panel in the ERP tab
  • Verify that each bin has the correct number of trials by looking in the ERP & Bin Information panel 
• Examine the data quality of the averaged ERPs using the View Data Quality Metrics panel in the ERP tab

Although this may seem like a lot, it’s the minimal set of steps for going from raw EEG to an averaged ERP in most studies. There are many additional steps that are often included in the EEG preprocessing pipeline, and there are also several steps after averaging. Moreover, each step has many options. 

That’s why I decided to write this book! I wanted to help you and other ERP researchers fully understand how to process EEG/ERP data using optimal methods that will lead to robust and important scientific discoveries.

---