Explicit representations of uncertainty drive attentional adjustments (360G-Wellcome-206480_Z_17_Z)

The aim of the proposed project is to understand how confidence serves cognitive control, more specifically how explicit representations of uncertainty guide attention. My own research findings and those of others suggest that humans possess an impressive ability to track and report uncertainty in the form of confidence judgements. Confidence has been found to co-vary with people’s objective performance and could thus serve as an internal teaching signal for situations when there is no external feedback that could guide behaviour. In the first part of the fellowship, I plan to study how different sources of uncertainty (internal and external) affect confidence. I predict that low confidence due to internal uncertainty should lead to an increased attentional focus (exploitation), whereas low confidence due to external uncertainty should lead to a widened attentional focus (exploration). In the second part of the fellowship, I would apply my questions to a special clinical population with impaired uncertainty processing: Patients diagnosed with obsessive-compulsive disorder (OCD) have often been found to be underconfident. I expect to find that this is due to an oversensitivity to internal uncertainty and predict to find more exploitation in OCD patients, given their inflated sense of uncertainty.