ABSTRACT
This chapter describes two experiments in which performance data were collected from humans as a benchmark for comparing the ability of four different modeling teams to replicate and predict the observed data. Our goal was to stress and extend existing modeling architectures by collecting a rich set of data that would require models to successfully integrate and coordinate memory, learning, multitasking, cognitive, perceptual, and motor components. The first experiment focused on multiple task management and attention sharing. The second experiment expanded on the first, embedding a category learning paradigm in a multitasking paradigm.
