Epistemology Think Tank: New Measures for Models
Save the Date: May 8-9, 2012
- We are assembling a group of 25 of the best people we can think of to tackle a central question regarding modeling and its analysis.
Over two days around a central table, with subsidiary groups as needed, our aim is to work collectively on a crucial topic regarding modeling and simulation. Invitations will be extended to active modelers in various fields, policy makers working with models, thinkers with a track record of innovation, and philosophers of science specializing in modeling and simulation to attend.
Topic: New Measures for Models
Modelers sometimes remark that their models are misunderstood, misinterpreted, or misapplied; that explanatory models or warning projections are treated as point predictions, for example. Policy makers and other model users sometimes remark that they don't know what a model is really telling them, and don't know when and whether they should trust it.
- Is it possible to outline (a) a taxonomy of types of models appropriate for different purposes, and (b) some multi-dimensional measure of confidence or credibility for models within a given type?
The hope would be that models and simulations could then come labeled appropriately for particular applications, with needed scientific qualifiers in place. Policy makers might then better be able to read and apply simulations appropriately.
- We need to know: Where does the problem stand? What steps might be made immediately? What further work needs to be done?
We have reserved six slots for members of the national MIDAS network.
Presented by the University of Pittsburgh MIDAS National Center of Excellence and the Center for Philosophy of Science