Cognitive control is flexibly adapted to task requirements in healthy individuals. Medio-frontal negativities in the event-related potential of the electroencephalogram can serve as indicators of cognitive control. With increasing conflict frequency, stimulus-locked control, as indexed by the N2, is increased and response-locked control, as indexed by the correct-related negativity, is reduced. On the behavioral level, this shift is associated with improved conflict resolution as evident in reduced response times and error rates in incompatible trials and a reduced response time congruency effect. Cognitive control adaptation might be implemented through experience-based task sets specifying advantageous processing strategies. Here, we investigated whether the cognitive control task set will be sustained when coupled with a contextual cue, even when the initial task requirements are no longer present. A flanker task with two conflict frequency conditions (25% incompatible and 75% incompatible trials) was presented. In the training phase, the conflict frequency conditions were paired with a distinct context cue (i.e., background color). In the test phase, the previously associated cues were again presented, but conflict frequency was identical in both conditions (50% incompatible trials). Although typical cognitive control adaptation was observed in the training phase on the behavioral and event-related potentials level, this pattern was not sustained in the test phase. Thus, the present study provides further evidence that cognitive control is flexibly adapted to task requirements even in the presence of misleading cues.
