Departmental Seminar Series: Prof. Alan Pickering
Bayes’ theorem “is to probability what Pythagoras theorem is to geometry” (Jeffreys, 1973). In this talk, using a variety of research examples, I will illustrate the 3 distinct ways in which this simple but powerful theorem, first published in the 1760s, is used in psychological research. The goal is to try to demystify these processes and show that each use of Bayesian methods is actually relatively straightforward and powerful.
The first use is as the basis for a statistical analysis of data which contrasts with traditional null hypothesis significance tests that are taught in undergraduate psychology courses. These analyses are easy to do, and to understand. They cast helpful new light on the evidential value of psychologists’ experimental findings. The second use is as a means for statistically comparing the fit of a set of models to our data, via so-called Bayesian model selection. This method for deciding which model fits best is an alternative to the use of goodness of fit indices, or to likelihood ratio tests in the case of nested models. The third use is as a basis for formal models of learning and decision-making, employing a process called Bayesian updating or belief propagation. This simple iterative procedure formally captures the evolution of responses, over the trials of a task for example. Some theorists argue that the human brain has evolved to do Bayesian inference implying that these Bayesian models of cognition reflect neural reality. I will present a simple example to illustrate how such Bayesian models work and suggest how real neurons might encode the statistical and distributional information that they require.
Alan Pickering is a Professor of Psychology and Dean of the Graduate School at Goldsmiths. After studying Natural Sciences as an undergraduate, Alan did his PhD in the cognitive neuropsychology of memory at the University of Manchester. During a subsequent post-doc at Kings in London, Alan switched to also carry out research on human personality and psychopathology, but always his work had an emphasis on neural, formal and statistical models of behaviour. After 11 years as lecturer and senior lecturer in psychology at St George’s Medical School, Alan joined Goldsmiths in 2001. His current research is primarily focused on learning in rewarding contexts and the role that dopaminergic neurotransmission may play in reinforcing such learning. He also works on investigating the reinforcement sensitivity theory of personality, in which between-individuals variations in the functioning of dopaminergically-mediated reward pathways are proposed to contribute to variations in extraversion.
Dates & times
If you are attending an event and need the College to help with any mobility requirements you may have, please contact the event organiser in advance to ensure we can accommodate your needs.