Tracing Problem Solving in Real Time: fMRI Analysis of the Subject-paced Tower of Hanoi

• Published in: Journal of cognitive neuroscience Vol. 17; no. 8; pp. 1261 - 1274
• Main Authors: Anderson, John R., Albert, Mark V., Fincham, Jon M.
• Format: Journal Article
• Language: English
• Published: One Rogers Street, Cambridge, MA 02142-1209, USA MIT Press 01.08.2005; MIT Press Journals, The
• Subjects: Adolescent; Adult; Algorithms; Biological and medical sciences; Cognition & reasoning; Cognition. Intelligence; Echo-Planar Imaging; Female; Fingers - innervation; Fingers - physiology; Fundamental and applied biological sciences. Psychology; Humans; Learning - physiology; Magnetic Resonance Imaging; Male; Neurology; Neuropsychological Tests; NMR; Nuclear magnetic resonance; Oxygen - blood; Parietal Lobe - physiology; Problem solving; Problem Solving - physiology; Psychology. Psychoanalysis. Psychiatry; Psychology. Psychophysiology; Psychomotor Performance - physiology; Real time; Reasoning. Problem solving; Time Factors; Human; Central nervous system; Problem solving; Cognition; Nuclear magnetic resonance imaging; Encephalon; Functional imaging
• ISSN: 0898-929X; 1530-8898
• DOI: 10.1162/0898929055002427

Previous research has found three brain regions for tracking components of the ACT-R cognitive architecture: a posterior parietal region that tracks changes in problem representation, a prefrontal region that tracks retrieval of task-relevant information, and a motor region that tracks the programming of manual responses. This prior research has used relatively simple tasks to incorporate a slow event-related procedure, allowing the blood oxygen level-dependent (BOLD) response to go back to baseline after each trial. The research described here attempts to extend these methods to tracking problem solving in a complex task, the Tower of Hanoi, which involves many complex steps of cognition and motor actions in rapid succession. By tracking the activation patterns in these regions, it is possible to predict with intermediate accuracy when participants are planning a future sequence of moves. The article describes a cognitive model in the ACT-R architecture that is capable of explaining both the latency data in move generation and the BOLD responses in these three regions.