Task‐switching costs have distinct phase‐locked and nonphase‐locked EEG power effects. (29th January 2020)

• Record Type: Journal Article
• Authors: McKewen, Montana; Cooper, Patrick S.; Wong, Aaron S. W.; Michie, Patricia T.; Sauseng, Paul; Karayanidis, Frini
• Is Part Of: Psychophysiology. Volume 57:Number 5(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-01-29
• Subjects: EEG -- nonphase‐locked power -- phase‐locked power -- task‐switching -- theta; Psychophysiology -- Periodicals; 612.8
• DOI: 10.1111/psyp.13533
• Languages: English
• ISSNs: 0048-5772

Abstract

Abstract: Event‐related potentials (ERPs) and total time–frequency power analyses have shown that performance costs during task switching are related to differential preparation to switch tasks (switch cost) and repeat the same task (mixing cost) during both proactive control (cue‐to‐target interval; CTI) and reactive control (post‐target). The time–frequency EEG signal is comprised of both phase‐locked activity (associated with stimulus‐specific processes) and nonphase‐locked activity (represents processes thought to persist over longer timeframes and do not contribute to the average ERP). In the present study, we used a cued task‐switching paradigm to examine whether phase‐locked and nonphase‐locked power are differentially modulated by switch and mixing effects in intervals associated with the need for proactive control (CTI) and reactive control (post‐target interval). Phase‐locked activity was observed in the theta and alpha bands, closely resembled that seen for total power, and was consistent with switch and mixing ERP positivities. Nonphase‐locked analyses showed theta and alpha power effects for both switch and mixing effects early in the CTI and as well as more sustained alpha and beta activity around cue onset, and extending from mid‐CTI into the post‐target interval. Nonphase‐locked activity in pretarget alpha and posttarget theta power were both correlated with response time mixing cost. These findings provide novel insight into phase‐locked and nonphase‐locked activity associated with switch and mixing costs that are not evident with ERP or total time–frequency analyses. Abstract : Time–frequency power is comprised of both phase‐locked and nonphase‐locked activity. Nonphase‐locked activity consists of signal that is removed from the ERP by the averaging process. Using a cued trials task‐switching paradigm, we examined whether phase‐locked and nonphase‐locked power were differentially modulated by task switch and mixing effects in intervals associated with the need for proactive control (cue‐target interval) and reactive control (post‐target interval). Switch and mixing effects were present in both phase‐locked and nonphase‐locked theta/alpha power. Nonphase‐locked power also showed sustained alpha and beta switch and mixing effects. Importantly, only the nonphase‐locked theta/alpha power was associated with individual variability in response time, suggesting that this measure captures unique activity not represented in either phase‐locked or the ERP signal.