Cognitive control is critical in adaptively guiding cognition, shaping an optimal balance between acting efficiently with existing rules, and remaining flexible enough to respond to new information. Proactive cognitive control utilizes context information to bias processing (in a top-down manner) and prepare for an upcoming event, presumably through dynamic use of highly related executive processes, including working memory maintenance, attention, and goal switching. Counter to the frequent investigation of proactive control as a unitary construct, we hypothesize that proactive control varies with the temporal delay over which goals need to be held in mind, and can be at least partly dissociated into separable neural processes. In the present experiment, we contrasted processing of cue context rules during short and long anticipated cue-probe delays. We hypothesize that rapid cue-probe delays would rely more on task-switching (rapid goal updating) processes, whereas longer cue-probe delays would rely more on active working memory maintenance of cued goal information. EEG was used to dissociate the neural processes of post-cue task-switching and working memory maintenance utilized for each delay type. Integrating neural and behavioral data, we aim to provide a more nuanced account of proactive control, clarifying the distinct proactive control processes instantiated for different temporal delays. Methods: Healthy undergraduate students (n=35) completed a variant of the AX-Continuous Performance Task (AX-CPT), assessing context processing in cognitive control, with short and long cue-probe delays. EEG data was recorded and ERP and time-frequency methods were used to analyze processing of cue information over short and long delays. Results: Accuracy was selectively impaired for rare probes following common cues (AY') during short delay blocks. RT did not vary across cue/probe pairs, but as expected, showed a positive correlation with delay length. Cue-locked N1, N2, P2, and P3 ERP components revealed interactions between cue type and delay length. P1, N1, N2, and P2 components revealed interactions between delay length and task-switch vs. task-repeat trials. Mid-frontal theta phase consistency was positively correlated with proactive/reactive ratio for Short delay, suggesting elevated task-switching proactive control (TSPC). Conversely, delta power over the time-course of the P3 was dynamically elevated for cues preceding a long delay; delta power correlated with proactive/reactive ratio for only this Long delay condition, indicating use of WM-maintenance proactive control (WMMPC), and thereby revealing a double-dissociation between type of control and delay length. Conclusions: The present study demonstrates a double-dissociation in proactive control processes invoked by known short and long delays, with distinct cue-locked ERP and oscillatory components revealing context updating and maintenance differences between delay conditions. These findings prompt consideration that AX-CPT delay length biases different use of task switching, working memory, and thereby proactive control type, and should no longer be conceived as a unitary task across timing parameters.
Level of Degree
Cavanagh, James F.
First Committee Member (Chair)
Second Committee Member
Cognitive Neuroscience, EEG, Cognitive Control
Janowich, Jacqueline R.. "A Time to Plan, A Time to Remember: An EEG Double Dissociation of Cognitive Control Instantiation Due to Temporal Delay." (2016). http://digitalrepository.unm.edu/psy_etds/66