×I4. How to build and how to kill trust in models, and why it matters

Models are often built and applied to support decision-making, for example to inform policy, guide investment decisions and quantify risk. However, all models are subject to limitations and uncertainty, and modellers have variable control in how their models are applied and model results interpreted. Modellers must therefore walk a delicate line between clearly communicating what their model can and what it cannot do: building trust in their model, but not too much trust.

This session welcomes submissions on issues surrounding model trust: how to build trust in models, how trust in models is lost, and the ways in which trust in models (or the lack of it) can affect model development and application. Submissions may include approaches to improve communication, reflections on case-studies with positive or negative engagement with models, examples of best practice in communication, participatory modelling, novel approaches to engagement, human perspectives on modelling or any other content relevant to communication and trust in models. Topics and presentation modes which engage the audience in discussion are particularly encouraged.

Key topics: Trust, Communication, Engagement, Model application