Known Unknowns and Unknown Unknowns: Coping with Different States of Uncertainty in a Changing World

People are required to learn a vast number of relationships between things in the world. Sometimes these relationships are very consistent: when it is dark and we press a light switch, the light almost always turns on. But sometimes these relationships are uncertain: a police officer might be faced with a crowd of people, and this situation may or may not suddenly become more volatile and violent. In these latter situations, people are faced with trying to resolve the uncertainty, ideally before any important event occurs. In order to make accurate predictions about what is going to happen, and act upon it, people need to use the information that is available in the environment, as well as their knowledge gained from past experience . The police officer, for example, may need to pay attention to the ages of the people, their gender, the number of people, and the time of the day. Some of these "cues" might be particularly relevant, and learning which ones are important will help them to make better predictions about what it likely to happen. The officer's knowledge about these cues will prove crucial when deciding how to act in this complex environment.

This project aims to provide a thorough experimental examination of the way in which the cognitive system copes with different uncertain situations. The starting point for this work is to conceptualise different uncertain situations as either "expected" or "unexpected". Sometimes we can provide a good estimate of the uncertainty we might face (expected uncertainty), but in other cases, the uncertainty occurs out of the blue (unexpected uncertainty). At the moment, psychologists have a very poor understanding of the way in which these very different types of uncertainty affect the basic processes of learning and attention.

In this research we will conduct three series of experiments to address this knowledge gap. In Series 1 we will investigate how stimulus processing is affected by uncertainty: do people change the way they look at and attend to stimuli; do they start to neglect stimuli over time; do they fail to commit those stimuli to memory? This will tell us about how the relationship between perceptual and cognitive processing changes in uncertain situations.

In Series 2 we will look at how stimuli become bound together in memory during complex uncertain situations. Is it the case that different parts of the environment are treated individually or are they considered more holistically as a consequence of uncertainty? The data will have important implications for understanding how uncertainty changes the content of memory for the world around us.

In Series 3 we will look at how relationships between events are learnt during uncertainty: do people need more information in order to become sure about relationships that exist; do sudden changes in the environment make us more aware of future events; does uncertainty affect everything we learn about, or just those stimuli that have been experienced as uncertain?

Our current theories of how uncertainty affects learning and attention have been shown to be inadequate to explain the existing published data. The current project will help us to develop a comprehensive set of effects that our models need to explain. The final aim of the project is to use our findings to update existing theories of behaviour. By doing so, we will provide a formal, mathematical theory of how uncertainty affects learning and attention.