Find out about some of the previous events held by Durham Energy Institute:
Let's not confirm our models...
This talk will discuss scientific models and their evaluation from a conceptual
point of view.
A view of models as representational tools will be introduced. Basic model evaluation, which aims to learn about model-target relationships, will be distinguished from applied model evaluation, which aims to learn whether a model is adequate for purposes of interest. It will be argued that, while confirmation and other familiar concepts can play a role in model evaluation, what are to be confirmed, falsified and so on are not models themselves but rather hypotheses about models. Reasons why this apparently minor shift -- from confirming models to confirming hypotheses about models -- makes a difference in practice will be discussed.
Contact email@example.com for more information about this event.