Understanding strategic adaptation in multitask settings

How do people interleave their attention when performing multiple tasks, such as dialing a phone number while driving, or checking e-mail while writing a paper? To investigate these issues a variety of modeling frameworks have been used, for example EPIC (Meyer & Kieras, 1997), SOAR (Lallement & John, 1998), ACT-R Threaded Cognition (Salvucci & Taatgen, 2008) and Cognitively Bounded Rational Analysis models (Howes, Lewis, & Vera, 2009). The majority of these frameworks focus on understanding how multiple tasks interfere with each other, for example as a result of having limited resources (e.g., two eyes, two hands) to dedicate to each task. Within the cognitive modeling community, relatively less attention is given to understanding how more top-down aspects, such as instructions and priorities, interact with these architectural aspects. However, some exploration has been done elsewhere. For example, it has been demonstrated that people adapt their performance to instructions to spend more time on a task (e.g., Gopher, 1993), or to changes in payment associated with performance (e.g.,Wang, Proctor, & Pick, 2007). In situations like these, the adaptation process can be understood as making trade-offs between performance on each of the tasks (e.g., Navon & Gopher, 1979; Norman & Bobrow, 1975). In my doctoral dissertation work I try to understand this flexible adaptation of dual-task performance, where people interleave attention in different ways despite being exposed to the same stimuli. As a modeling approach, I use Cognitively Bounded Rational Analysis Models (Howes, et al., 2009). However, I also have an interest in informing and using other architectural frameworks.