Contextual recruitment of cognitive control in preadolescent children and young adults.

• Children aged 9–12 years were able to demonstrate contextual control recruitment. • Adjusting control resources in a trial-by-trial manner turned out to challenge children. • Young adults showed reliable adjustments of control for divers contextual requirements. The ability to use contextual cues to adjust cognitive control according to situational demands is a hallmark of flexible and adaptive behavior. We investigated the development of three different types of contextual control recruitment in children (9- and 12-year-olds) and young adults. First, we implemented a list-wide proportion congruence manipulation in which conflict trials were frequently/infrequently presented within a list of trials. Second, we implemented a location-specific proportion congruence manipulation in which conflict trials were frequently/infrequently presented at one of two locations. Both types of contextual control recruitment are based on the formation of high-level associations between context features (lists and locations) and the respective cognitive control set. Contextual recruitment of control is observed in reduced interference at contexts with high conflict frequencies. Finally, we investigated a trial-by-trial, conflict-triggered recruitment of cognitive control. Here, the experience of a conflict in the previous trial is expected to reduce subsequent conflict. In all three forms of control recruitment, distinct contextual cues reveal information about the required extent of cognitive control. Young adults showed reliable adjustments of control for all types of contextual cues. Children were able to demonstrate contextual control recruitment based on stable context–control associations (lists and locations). However, using single conflict signals turned out to challenge children in that they were able to adapt control resources only for error reduction, not for reaction times. Altogether, the results indicated that children can learn and use high-level associations between context and control sets. Implications regarding proactive and reactive mechanisms of cognitive control are discussed. [ABSTRACT FROM AUTHOR]