In my research, I use experimental and computational methods to understand the psychology of mood, learning, and decision making. The primary tool that I use in this research is reinforcement learning modelling of choice behaviour, and where relevant I also use computational methods for analysis of neural data (e.g., multivariate pattern analysis of EEG and fMRI data). Several of my current research projects are detailed below.
One primary thread of my postdoctoral work with Yael Niv at Princeton University was to build computational models of human mood. In this theoretical work, we focused on how working with concepts from reinforcement learning might help us to build a computational theory of human mood.
- Bennett, D., Davidson, G., & Niv, Y. (in press). A model of mood as integrated advantage. Manuscript accepted for publication in Psychological Review. [pdf]
In my empirical experimental work, I’m very interested in the question of how human mood states interact with cognitive processes like attention, learning, and decision making. For instance, how does being in a bad mood change the way we learn about the world around us? Does being in a good mood make us more likely to engage in risky behaviour? Do effects of this kind differ across people in a way that might explain why some people are prone to mood disorders?
- Bennett, D., & Niv, Y. (2020). Opening Burton’s clock: Psychiatric insights from computational cognitive models. Chapter in The Cognitive Neurosciences (6th edition; M. Gazzaniga, ed.). [pdf]
- Bennett, D., Niv, Y. & Langdon, A. J. (in press). Value-free reinforcement learning: Policy optimization as a minimal model of operant behavior. Manuscript accepted for publication in Current Opinion in Behavioral Sciences. [pdf]
In my PhD, I studied the psychological processes underlying information seeking under uncertainty. This question is related to the problem of reconnaissance: is it better for an individual to act on the basis of present beliefs, or incur a cost to acquire additional information which may aid their decision?
- Bennett, D., Sutcliffe, K., Tan, N. P., Smillie, L. D., & Bode, S. (in press). Anxious and obsessive-compulsive traits are independently associated with valuation of non-instrumental information. Manuscript accepted at Journal of Experimental Psychology: General. [pdf]
- Bennett, D., Bode, S., Brydevall, M., Warren, H., & Murawski, C. (2016). Intrinsic valuation of information in decision making under uncertainty. PLOS Computational Biology, 12(7), e1005020. [pdf]