Step-by-Step Protocols for Collecting Clean EEG Data

Standard Version: Farrens, J. L., Simmons, A. M., Luck, S. J., & Kappenman, E. S. (2019). Electroencephalogram (EEG) Recording Protocol for Cognitive and Affective Human Neuroscience Research. Protocol Exchange. https://doi.org/10.21203/rs.2.18328/v2 [PDF]

COVID-19 Version: Simmons, A. M., & Luck, S. J. (2020). Protocol for Reducing COVID-19 Transmission Risk in EEG Research. Protocol Exchange. https://doi.org/10.21203/rs.3.pex-974/v2

We have published an in-depth description of our EEG recording procedures, which provides an extremely detailed, step-by-step account of how we currently record EEG data in our laboratories (along with a modified version to minimize risk of COVID-19 transmission). Although this level of detail important for collecting clean data, it would be unrealistic to include 20+ pages of recording details in the Method section of a journal article. By publishing this protocol and then citing it in our papers, other researchers will know exactly how we recorded our data, which will enhance reproducibility. In addition, this protocol provides a forum for sharing the tips and tricks we have developed for collecting clean EEG data, which may help you improve your data quality. We encourage other researchers to either follow and cite our protocol or publish and cite their own protocols. If you’d like more information about why we think this is important, read on…

If you’ve ever recorded or processed raw EEG data, you know how noisy the data can be. The neural signals we want to record are contaminated by a variety of biological and non-biological noise sources, including muscle activity (EMG), heartbeats (EKG), skin potentials, movement artifacts, and induced electrical noise from the environment. To maximize the likelihood of finding the neural effects of interest, researchers need to do everything they can to reduce the noise during a recording session. Postprocessing techniques such as filters can help, but they can’t eliminate all the noise, and they often have a cost (e.g., temporal distortion). This is why one of our ERP Boot Camp mottos is “There is no substitute for clean data!”

When you run an EEG recording session, there are a million little details that together impact the quality of the data. Every lab has its own approach, but our field has no widely used mechanism for sharing these methodological details. Recording methods are usually described in journal articles with only a brief mention of the recording parameters (e.g., filter settings and sampling rate) and no information about the millions of other details that impact data quality. And really, who would want to slog through a Method section that described how the electrodes were oriented in their holders or listed the specific make and model of chair that was used to ensure that the participant remained comfortable?

However, these details really matter. For example, we have shown that the temperature of the recording environment can have a substantial impact on statistical power (Kappenman & Luck, 2010), but we have never seen an EEG/ERP paper from another lab that mentioned the temperature of the recording room.

A mechanism now exists for reporting all of these details. Specifically, one can now easily publish a formal protocol that is permanent and citable (and even contributes to citation counts on Google Scholar, if you care about that sort of thing). There are a variety of ways to publish a protocol, but we chose Nature’s Protocol Exchange web site. It’s free, appears to be robust, and automatically generates a DOI. It does not involve peer review (because you are merely listing your procedures, not drawing any formal conclusions), but it does involve a quick administrative review (to make sure that inappropriate materials are not posted). The protocol is published with a Creative Commons license, so Nature does not own it, and anyone can use it. Other sites may be even better, but Protocol Exchange fits our current needs reasonably well (although we would have appreciated a little more control over the formatting).

We encourage other researchers to read our protocol and use it to get ideas for their own recording methods. Our methods may not be ideal for some kinds of research, and other labs may have equivalent or even better methods.

More than anything, we hope that our protocol inspires other researchers to start publishing their own detailed protocols. This sharing of information will help everyone collect the cleanest possible data, improving the quality of published research in our field. Detailed protocols will also increase reproducibility of research methods and perhaps even replicability of research findings. Note that our protocol is longer and more detailed that what most labs would publish (because we included advice about things like fixing broken electrodes).

You can find the standard version of the protocol at protocolexchange.researchsquare.com/article/663a5a19-c74e-4c7d-b3fc-9c5188332b46/v2 or at doi.org/10.21203/rs.2.18328/v2. You can download a nicely formatted PDF of the protocol here.

You can find the COVID-19 version at https://protocolexchange.researchsquare.com/article/pex-974/v2 or at https://doi.org/10.21203/rs.3.pex-974/v2