The Centre for Cognitive Neuroscience and Cognitive Systems at Kent has a large body of research focussed on the role of emotions in human cognition, in general, and in human decision making, in particular. One focus of this work is on understanding the behaviour of patients with damage in ventral areas of the frontal lobes on a spectrum of decision making tasks, especially the Iowa Gambling task. In particular, we have developed a reinforcement learning model, which does a good job of explaining patient and normal behaviour on a spectrum of relevant decision making tasks. This model informs many of the theories that have been proposed, such as the somatic markers theory, myopia for future consequences and risk seeking.
In a broader context, our work fits with dual process theories, which propose that two systems interact when humans make decisions. These systems are a (fast and dirty) low level statistical system and a higher level reflective / "rational" system, which probably has the character of classic symbolic models of cognition.
The proposed PhD activities could build from this body of research in a number of ways. Firstly, a number of empirical questions have arisen from our modelling that could be evaluated via classic behavioural experimentation on normals and patient groups. Secondly, we are very keen to undertake electrophysiological (EEG and ERP) recording to inform our theoretical understanding. Finally, there is still a good deal more room for connectionist modelling in this context.