The goal of our research is to understand how people organize knowledge that they acquire from their diverse experiences to produce intelligent behavior. The concern is very much with how it is all put together and this has led to the focus on what are called "unified theories of cognition." A unified theory is a cognitive architecture that can perform in detail a full range of cognitive tasks. Our theory is called ACT-R and takes the form of a computer simulation that is capable of performing and learning from the same tasks that subjects in our laboratories work at. So, for instance, we have a model that plays (and learns to play) a video game just like our participants.
Our theoretical goals have led us to study a wide range of empirical question such as how strategies for problem-solving evolve, how people discover things about a new domain, how they deal with the working memory load imposed by the tasks, and how they get faster at accessing information relevant to task performance. A particular focus has been on the learning of mathematical knowledge with a recent emphasis on how people can use such knowledge in novel ways. This research has contributed to the development of cognitive tutors that are currently being used to help teach courses in schools around the country.
We have come to increasingly rely on data from neural imaging to provide converging evidence about cognitive processes. ACT-R models automatically make predictions about the data we should see from imaging modalities like fMRI and EEG. We have also been working on machine learning techniques that mine brain-imaging patterns and discover the structure of cognitive processes. We have been particularly focused on methods that will identify the sequential structure of cognitive processes. These methods have proven successful both at identifying how complex tasks are performed and at decomposing responses times into the latencies of the basic cognitive steps.