3.3: Exercise- Examining the Single-Participant ERPsets

Applied Event-Related Potential Data Analysis 2e · LibreTexts

0%

Contents

    Text

    Size
    100%
    Theme
    Typeface
    Line length
    Spacing

    Read aloud

    How

    Speed: 1×
    Light up
    Highlight

    Symbols: Mulberry Symbols, © 2018–2026 Steve Lee, CC BY-SA 4.0 (licence, the set), the site's own copies.

    From LibreTexts, under CC BY-NC 4.0, as the source states it; its equations are written in TeX, which the PDF shows typeset. This page at LibreTexts.

    3.3: Exercise- Examining the Single-Participant ERPsets

    Now let’s take a look at the averaged ERP data from each participant. Select all 10 of the ERPsets in the ERPsets panel. Although all 10 are selected, only the first one should be displayed in the plotting region. You can choose which one is displayed using the Prev. and Next buttons (or the text box between them) in the upper right corner of the plotting region.

    We mainly care about the ERPs elicited by the target words, so select Bins 3 and 4 in the Bin & Channel Selection panel (and click Apply). And let's use the nice channel layout that we used at the end of Chapter 2: Go to Plot Settings, select Custom, click Import, load GridLocations_Custom.tsv, and click Apply. It should look something like Screenshot 3.2.

    Screenshot 3.2
    ERPset 1

    If you look at the CPz electrode site, you should see that the unrelated targets elicited a more negative voltage around 400 ms than the related targets. This is the same effect we saw for Subject 6 in Chapter 2, but we're now looking at Subject 1's waveforms, which are a bit noisier. If you click on the Next button in the upper right corner of the plotting region, you can see the data from Subject 6. There is clearly less high-frequency noise in Subject 6 than in Subject 1.

    It's also a good idea to look at the number of trials included in each average when you're viewing the ERP waveforms. To do this, open the ERP & Bin Information panel. The bottom of the panel contains a table showing the number of accepted and rejected trials in each bin. You can see that more trials were rejected for Subject 1 than for Subject 6. [You can get more detailed information by clicking the Bin Info Table button.]

    Using the Next button, take a look at the rest of the ERPsets (both the waveforms and the number of accepted/rejected trials). Although all participants showed the typical N400 effect (more negative voltage for unrelated targets than for related targets at CPz and the surrounding electrodes), the waveforms differ quite a bit across participants. The noise also varies quite a bit. For example, Subject 22 has high-frequency noise at the most anterior electrode sites, and Subject 38 has clear alpha-band oscillations remaining in the waveforms at the posterior sites. The number of rejected trials also differs quite a bit across participants. This kind of variability is perfectly normal. In fact, these data are from a fairly homogeneous group of highly compliant young adults, so the data are cleaner and less variable than what is seen in many studies.

    Viewing the Data Quality Metrics

    Now let's look at the data quality by going to the View Data Quality Metrics panel (with all 10 subjects still selected in the ERPsets panel). Click the Show in a table button, which will open a new window letting you see the aSME values. By default, the values from the first ERPset will be shown. Select the 6_N400 ERPset using the Selected ERPset menu at the top of the window. The aSME values are shown for Bin 1, but you can change to a different bin using the Selected Bin menu. Turn on the Color Heatmap option. The table should now look exactly like what you saw for this participant in Section 10 of Chapter 2, with the worst (largest) aSME values in the later time windows in the F4 channel.

    Now select the 1_N400 ERPset using the Selected ERPset menu at the top of the window. The worst data quality for this participant is in the C5 channel. If you look at the other bins for this participant (using the Selected Bin menu), you'll see that C5 has the highest aSME values for all bins, and that it's worse for bins 3 and 4 (the targets) than for bins 1 and 2 (the primes).

    Now look at the data quality for each of the remaining ERPsets. You should see that channel with the worst data quality varies quite a bit from participant to participant. This is normal; unless you have a hardware problem, the noisiest channel will be highly idiosyncratic (especially for the low-frequency noise that actually matters most for data quality when ERPs are quantified as the mean voltage over a time range). You'll also see that the highest values vary somewhat across participants, but none of the SME values are huge. I selected these 10 participants for this chapter because they all have nice, clean data.

    ERPLAB automatically computes some additional data quality metrics. For example, if you select Baseline Measure – SD from the Selected DQ Measure menu, you can see a measure of noise during the prestimulus baseline period (the standard deviation of the voltage from point to point during this period). More details can be found in the ERPLAB manual page on data quality metrics. You may now close the data quality viewing window.

    Viewing the ERP and ALLERP Data Structures

    You can do quite a lot in ERPLAB Studio without knowing anything about Matlab programming. However, even if you don't care anything about programming, it's worth knowing a little bit about the data structures that are used to store EEGsets and ERPsets.

    If you go to the main Matlab window, you should see a pane labeled Workspace (probably along the right side of the Matlab window). That pane shows all of the variables that are currently available to Matlab. The most important variables are

    • EEG, which holds the currently selected EEGset
    • ALLEEG, which holds all of the EEGsets that are currently loaded
    • ERP, which holds the currently selected ERPset
    • ALLERP, which holds all of the ERPsets that are currently loaded

    Here we will take a look at the ERP variable. Go to the Matlab command window and type ERP (followed by Return or Enter). You should see something like the information shown in Screenshot 3.3.

    Screenshot 3.3
    ERP Structure

    Typing the name of a variable in the command window tells Matlab to print out the contents of that variable. ERP is a complicated variable with many different fields, including the erpname field (the name of the ERPset), the nchan field (which stores the number of channels), the chanlocs field (which stores the names and 3-D locations of the electrodes), and the bindata field (which stores the actual ERP data, in binary format).

    You can also see the contents of a variable by double-clicking the name of the variable in the Workspace pane. Try this by double-clicking the ERP variable in the Workspace. This causes the variable to be shown in a separate pane. You can then double-click on the fields of the ERP variable to see those fields in more detail. For example, try double-clicking the chanlocs field to see what information it holds.

    Now look at the ERP.ntrials field. This allows you to see the number of accepted trials, the number of rejected trials, and the number of invalid trials for each bin. You should see that there were 47 accepted trials and 13 rejected trials for Bin 1.

    Looking at the ERP and EEG variables can be helpful if things aren't working correctly and you want to understand more precisely what is happening.


    This page titled 3.3: Exercise- Examining the Single-Participant ERPsets was last modified on Thu, 01 Oct 2026 18:52:35 GMT and is shared under a CC BY-NC 4.0 license and was authored, remixed, and/or curated by Steven J Luck directly on the LibreTexts platform.