Comparison of the Working Memory Load in N-Back and Working Memory Span Tasks by Means of EEG Frequency Band Power and P300 Amplitude.

Christian Scharinger,Alexander Soutschek,Peter Gerjets,Torsten Schubert

doi:10.3389/fnhum.2017.00006

Abstract

According to theoretical accounts, both, N-back and complex span tasks mainly require working memory (WM) processing. In contrast, simple span tasks conceptually mainly require WM storage. Thus, conceptually, an N-back task and a complex span task share more commonalities as compared to a simple span task. In the current study, we compared an N-back task, a complex operation span task (Ospan), and a simple digit span task (Dspan) by means of typical WM load-related measures of the Electroencephalogram (EEG) like the parietal alpha and beta frequency band power, the frontal theta frequency band power, and the P300 amplitude, to examine whether these tasks would show commonalities or differences in WM processing-load. We expected that increasing WM-load would generally lead to a decreased alpha and beta frequency band power, an increased theta frequency band power, and a decreased P300 amplitude. Yet, based on the conceptual considerations, we hypothesized that the outcomes of these measures would be more comparable between the N-back and the Ospan as compared to the Dspan. Our hypotheses were partly confirmed. The N-back and the Ospan showed timely more prolonged alpha frequency band power effects as compared to the Dspan. This might indicate higher demands on WM processing in the former two tasks. The theta frequency band power and the P300 amplitude were most pronounced in the N-back task as compared to both span tasks. This might indicate specific demands on cognitive control in the N-back task. Additionally, we observed that behavioral performance measures correlated with changes in EEG alpha power of the N-back and the Ospan, yet not of the Dspan. Taken together, the hypothesized conceptual commonalities between the N-back task and the Ospan (and, for the Dspan, differences) were only partly confirmed by the electrophysiological WM load-related measures, indicating a potential need for reconsidering the theoretical accounts on WM tasks and the value of a closer link to electrophysiological research herein.

Highlights

Each cognitive task of our daily live like checking a receipt by means of mental arithmetic induces load on working memory (WM)
The current study addressed this issue by investigating typical EEG measures for WM-load like the EEG theta (4–6 Hz), alpha (8–13 Hz), and beta (14–24 Hz) frequency band power (Gevins and Smith, 2000; Pesonen et al, 2007; Palomäki et al, 2012; Scharinger et al, 2015), and the P300 amplitude (Wickens et al, 1983; Watter et al, 2001; Scharinger et al, 2015) for an operation span task (Ospan), an N-back, and a digit span task (Dspan) task
The current study compared typical EEG correlates of WM processing-load like the EEG theta, alpha, and beta frequency band power and the P300 amplitude between a complex Ospan task, a simple Dspan task, and an N-back task

Summary

Introduction

Each cognitive task of our daily live like checking a receipt by means of mental arithmetic induces load on working memory (WM). WM might be fractionated into shortterm memory (STM) storage components and a processing component, the central executive (Engle, 2002; Baddeley, 2003, 2007). The central executive may be fractionated further into executive functions (EF) like updating, inhibition, and shifting: the updating of information temporarily memorized for processing, the shifting of the attentional focus between different task demands, and the inhibition (or interference control) of information not (or no longer) relevant for the current processing step (Miyake et al, 2000; Baddeley, 2007; Bledowski et al, 2010). Complex span tasks as well as N-back tasks are conceptualized to induce WM processing-load (see “Conceptual Task Analyses of WM Span Tasks and N-Back Tasks” for detailed task analyses)

Methods

Results

Discussion

Conclusion