Why do children perseverate when they seem to know better: graded working memory, or directed inhibition?

Children sometimes have trouble switching from one task to another, despite demonstrating an awareness of current task demands. This behavior could reflect problems either directly inhibiting previously relevant information or sufficiently activating graded working me mory representations forthe current task. We tested competing predictions from each account, using a computerized card-sorting task in which we assessed children's task switching abilities and their response speed to simple questions about current task demands. All children answered these questions correctly, but children who successfully switched tasks responded more quickly to questions than did children who perseverated on previous tasks, even after factoring out processing speed and age. This reaction time difference supports graded working memory accounts, with stronger representations of current task demands aiding both task-switching and responses to questions. This result poses a challenge for directed inhibition accounts, because nothing needs to be inhibited to answer simple questions that lack conflicting information.