My research interests are in developing and evaluating mathematical and algorithmic psychological models. I have worked in areas ranging from stimulus representation, to categorization and memory, to decision-making and problem-solving.

Recurrent themes in this modeling are the benefits of thinking about both the basic and applied aspects of psychological phenomena simultaneously, with the conviction that knowing about one facilitates progress on the other, and the need to make inferences about theories and models in rigorous, systematic and flexible ways.

Both of these themes are probably sometimes strategic mistakes. I once puzzled an external accredition reviewer by discussing "applied research", because she was under the impression my work was of "good quality". She seemed to think the two things were mutually exclusive. And, anytime you use unfamiliar approaches to evaluating data or models, you run the risk of having your work dismissed as "just methodological".

But, I don't think there's much point generating psychological knowledge that isn't useful. And there's  no point building theories and collecting data if you don't have decent methods for bringing them into contact with each other. So, the current plan is to perservere.