Modeling of control over task-switching and cross-task interference supports a two-dimensional model of cognitive stability and flexibility

Poster Session A - Saturday, April 13, 2024, 2:30 – 4:30 pm EDT, Sheraton Hall ABC

Raphael Geddert¹ (raphael.geddert@duke.edu), John Pearson¹, Tobias Egner¹; ¹Duke University

Reading a book in a busy coffee shop requires the ability to focus on the task at hand while ignoring task-irrelevant distractions, referred to as cognitive stability. Setting aside the book to answer a phone call requires the ability to switch tasks, referred to as cognitive flexibility. Theories of cognitive control have conceptualized stability and flexibility as opposing ends of a one-dimensional stability-flexibility continuum, necessitating a “stability-flexibility trade-off”: increasing stability (prioritizing task focus) reciprocally reduces flexibility (a readiness to switch task sets), and vice versa. Recent evidence, however, has supported a two-dimensional stability-flexibility relationship, whereby stability and flexibility can both be maintained at high levels simultaneously when necessary. Here, we adjudicate between the one- and two-dimensional accounts by fitting competing models to behavioral data from a cued task switch experiment that manipulated the proportion of task switch trials (driving contextual adjustments in flexibility) and cross-task congruency effects (driving contextual adjustments in stability). The two-dimensional stability-flexibility model had a superior model fit, and it was also the only model capable of reproducing key behavioral phenomena in the original data set. We next considered the contribution of various drift diffusion model parameters, including drift rate, boundary separation, and non-decision time, to switch costs and congruency effects, and show that we can dissociate flexibility and stability-related components in the evidence accumulation process. Finally, we replicate the superior fit of the two-dimensional stability-flexibility model in additional samples, demonstrating the robustness of our results across data sets.

Topic Area: EXECUTIVE PROCESSES: Monitoring & inhibitory control