Finally, our large datasets allow us to quantitatively investigate the effect of data quality on some of the phenotypes that were collected in the population‐based and clinically ascertained samples. Meta‐data about the cleaning process—for example, data recording length, number of channels lost, or the number of epochs rejected after visual cleaning—could all be used to predict, for example, the age of the subject, or any psychiatric or behavioral outcome. As such variables of recording and processing quality may be associated with phenotypes, this information could be invaluable to the whole field of EEG and possibly result in specific thresholds for acceptable data.
