Traditionally, scalp EEG has been thought of in terms of waves or oscillations of various frequencies. Recently it has become apparent that the EEG contains not only oscillations but also non-wave noise elements. These non-oscillations may not relate to functional brain activity, and require identification and removal before the true impacts of brain oscillations can be assessed. To this end we developed the PaWNextra algorithm, which provides valid estimates of Pink noise (PN) and White noise (WN) from an individual’s observed EEG spectrum.

Over the years we have focused our brain dynamics work on the auditory equiprobable Go/NoGo task. This is a simple two-choice reaction-time (RT) task suitable for use in a variety of populations of different ability levels. Using an optimised temporal Principal Components Analysis (t-PCA) approach, we have linked ERP components from this task to the underlying perceptual and cognitive processing stages in a Processing Schema.

To explore the component structure of the resting EEG spectrum, we extended our PCA approach to the frequency domain (f-PCA), a data-driven approach that avoids the arbitrary frequency ranges in the traditional EEG band literature. This has led to consistent spectral decompositions useful in clarifying links between different alpha components and autonomic arousal.

These methodological elements are brought together here in a study exploring EEG-ERP-behaviour linkages in young adults during an equiprobable auditory Go/NoGo task. University students (N = 47) carried out this task while continuous EEG was recorded. Exploratory multiple regression analyses indicated that prestimulus alpha and beta oscillations predicted N2c, P3b and SW1 components associated with the imperative Go response, while PN facilitated N1b and N1c, early markers of processing and identification of the NoGo stimulus. There were no direct effects of prestimulus EEG measures on behaviour,but the EEG-affected Go N2c and P3b components impacted Go RT performance measures.