Potential PhD Projects

Imperfect Decisions

We face many competing demands on our attention in our daily life and regularly have to divide our time and attention between different tasks. In the face of this competition between multiple tasks and goals, we need to allocate our cognitive resources in the best possible way given the time we  have, our abilities and strengths, and the difficulty of the tasks involved. We recently observed a striking failure to make rational decisions about how to allocate cognitive resources under even quite simple task constraints (Clarke and Hunt, Psychological Science, 2016). We’ve replicated this failure using a range of different task contexts (e.g. eye movements, memory, even throwing beanbags into hula hoops).

We’re interested in understanding why this breakdown occurs, and whether we can improve human decisions in this context by examining various potential mediating factors and explanations such as strategy, expertise, insight and reward.

Prediction in Vision

We typically think of the brain in terms of stimulus and response, but under many circumstances, the brain does not passively wait for stimulation. Active prediction about highly probable events can give the brain a head-start on processing input before it even occurs.

A clear example comes from eye movements: when an eye movement is about to be executed, we begin to process information in the location it will be on the retina, rather than (or in addition to) where it currently is (e,g, Hunt and Cavanagh, 2009; 2011). Details about the flexibility and extent of these predictions remain to be explored.

Drivers of fixation behaviour

A classic question in psychology is the degree to which our attention, eye movements, and (as a consequence of these) our experience of the world around us are shaped by the physical properties of the input, our top-down goals, and midlevel effects of sequence and context.

These questions are typically addressed using carefully-controlled tasks and stimuli. But do the conclusions drawn from these constrained conditions scale up to unconstrained, natural situations? We are developing tools and methods to address this question.