Is Cognitive Control Reliable? When means are not enough

Most theories of cognitive control assume goal-directed behavior takes the form of performance monitor-executive function-action loop. Recent theories focus on how a single performance monitoring mechanism recruits executive function - dubbed single-process accounts. Namely, the conflict-monitoring hypothesis proposes that a single performance monitoring mechanism, housed in the anterior cingulate cortex, recruits executive functions for top-down control. This top-down control manifests as trial-to-trial micro adjustments to the speed and accuracy of responses. If these effects are produced by a single performance monitoring mechanism, then the size of these sequential trial-to-trial effects should be correlated across tasks. To this end, we conducted a large-scale (N=125) individual differences experiment to examine whether two sequential effects - the Gratton effect and error-related slowing effect - are correlated across a Simon, Flanker, and Stroop task. We find weak correlations for these effects across tasks which is inconsistent with single-process accounts.