George Siemens points to an article on Task-Switching, Emotional Motivation, and Reward from Drs. Fernette and Brock Eide. They in turn are writing about a study on functional magnetic resonance imaging (fMRI) that studied people who are asked to switch tasks. The study leans heavily on the analysis of fMRI, which is not an area of my expertise. From the Fernette & Eide post:
Task-switching is a common cognitive task associated with attention and cognitive control. In this very interesting study, researchers examined what brain areas were associated with good task switching ability. Surprisingly, the most highly correlated area for efficient task switching was ... a region implicated in emotional and social motivation and reward.
It may be a stretch to associate this work with multi-tasking. However, the results of the study do hint at the oft-mentioned cognitive costs of multi-tasking: it takes time to switch from one task to another simply because you need to shift your attention between the items. And some people may be better at doing this than others - they are more able to shift their attention completely to the new task in a reasonable amount of time. I wonder if the people who jump from activity to activity are those that can't keep their attention locked on the thing in front of them.
By the way, theory of constraints generally considers multi-tasking a bad thing. The mere fact of doing something else guarantees a delay on the primary task or project. Multi-tasking is defined as switching from one task to another to be seen as making progress on both. It is not talking on the phone while reading email.