Electrophysiological underpinnings of individual differences in verbal working memory

Abstract

Given the importance of working memory (WM) for everyday cognitive functioning, understanding the architecture of WM is crucial for understanding human behavior. Only a limited progress has been made to explore neural underpinnings of individual capacity limits despite 25 years of extensive psychophysiological WM research. Particularly, most psychophysiological studies target only the short-term memory (STM) construct, while STM is only a part of the working memory responsible for the storage of sensory information. Much less effort has been devoted to study brain mechanisms supporting the executive component of WM – the part that allows the manipulation of information. Therefore, I pursued two main goals in the current work: (1) to study electrophysiological correlates of sensory storage and central executive components of WM and (2) to assess the contribution they make to individual differences in WM performance. To reach these goals, a large number of participants (N = 156) were tested in WM tasks of average to high complexity with and without manipulations, while EEG was recorded. Increasing the complexity of the tasks allowed to increase the variance in order to avoid ceiling effect in performance and to better distinguish low and high performers. Introduction of the task with mental manipulations allowed to assess the contribution of executive WM components separately from its storage components. As expected, WM performance decreased with increasing memory load and was worse in the more difficult manipulation task than in the easier retention task. In accordance with an a priori hypothesis, manipulations in WM enhanced theta activity; hence, confirming the role of middle frontal theta in the executive functions and its relationship with the central executive component of WM. On the other hand, the hypothesis concerning the role of alpha in filtering out distractors was not confirmed. Moreover, alpha activity was not related to individual WM performance. Instead, lower beta activity in the frequency range of 16-22 Hz emerged as a potential candidate to affect the individual WM capacity; the mechanism of this effect remains unknown. Supposedly, beta oscillations can influence WM performance through regulation of the ability to execute mental manipulations. Finally, the results of the present study confirm the hypothesis that executive components of WM, rather than its components related to sensory storage, play the decisive role in individual WM capacity limits.