Genetic studies typically investigate how individual differences in phenotypes are affected by genetic variants—specifically for our consortium, individual differences in EEG parameters. Therefore, the quality of EEG feature extraction needs to be assessed against the background of the variability of the EEG features in the population being measured. This means that apparatus, data quality control, and sampling/cohort characteristics must not greatly affect the individual participants' rank ordering on the EEG features extracted and should largely capture the same variation. Some aspects of recording are not likely to affect the variability of the EEG features, that is, when they cause a fixed bias. For example, the recording filter settings, with their mostly linear effects on oscillation power and when applied constantly across individuals, will not affect the relative score between individual participants on EEG power. Other aspects of recording, on the other hand, may greatly affect the rank ordering of individual data. For example, if a subset of participants were to fall asleep during the resting recordings, this would greatly affect their average power of oscillatory activity (Niedermeyer, 1999b). To avoid
