Achievement

Trainee Kibbe (Psychology) used her IGERT training in computational and statistical modeling to study visual search and memory in human subjects. The task required search through arrays of multi-featured hidden objects (revealed by mouse-clicks or eye fixations) for 3 belonging to the same category. A finite state Markov model applied to the sequences of observed movements showed that subjects revisited previously seen items when a memory buffer with a capacity of about 4 objects was filled, but the tradeoffs between exploration and memory could be influenced by both motor costs or category complexity. This suggests sophisticated rules of resource management govern human search in order to optimize performance (Kibbe & Kowler, 2011, Journal of Vision,11(3): 14, 1-21)