Escape from Model Land – How Mathematical Models Can Lead Us Astray and What We Can Do About it

In the third episode of MINDS CAST, we had the privilege of engaging with Erica Thompson, the author of “Escape from Model Land” This session delved into the intricacies of using mathematical models in decision-making, highlighting both the potential benefits and the pitfalls of relying heavily on these models. Here, we summarize the key insights from this enlightening discussion, including Thompson’s perspectives on the role of models in understanding complex issues and the importance of critical engagement with these tools.

About the Presenter:

Dr. Erica Thompson, Associate Professor of Modelling for Decision Making at UCL’s Department of Science, Technology, Engineering, and Public Policy, brings a wealth of expertise to the topic. Her research focuses on the application of mathematical and computational models in decision-making, with a particular emphasis on developing techniques for assessing and interpreting models. As a Fellow of the London Mathematical Laboratory and a Visiting Senior Fellow at the LSE Data Science Institute, her work also delves into the philosophical aspects of model outputs and their integration with expert judgment in decision-making processes.

Webinar Overview:

1. The Essence of Models in Decision-Making: Thompson opened the discussion by exploring the role of mathematical models in aiding our understanding of complex issues like economic regulation, climate change, and pandemics. She emphasized that while models are crucial tools, they do not produce perfect predictions of the future.
2. Challenges in Mathematical and Social Contexts: The conversation highlighted the mathematical and social challenges in using models for decision-making. Drawing on examples from climate change and COVID-19, Thompson underscored the importance of recognizing the limitations and assumptions underlying these models.
3. Best Practices for Responsible Modeling: The session also touched upon best practices for responsible modeling, especially in politically charged contexts. Thompson stressed the need for transparency, accountability, and the acknowledgment of value judgments embedded within models.
4. Navigating Model Land: The concept of ‘Model Land’ – a metaphorical space where models exist in their idealized form – was a key focus. Thompson discussed the risks of becoming overly reliant on models and strategies for effectively navigating out of Model Land while still leveraging the benefits they offer.

Key Points from the Discussion:

• Model Diversity and Assumption Transparency: Thompson advocated for the diversity of models and perspectives in modeling processes. She emphasized the need for models to make their assumptions and biases explicit, allowing for more critical engagement and evaluation of their outputs.
• The Role of Experts in Modeling: The discussion also addressed the evolving concept of expertise in modeling. Thompson highlighted the importance of involving a range of experts in the modeling process to ensure diverse perspectives and assumptions are considered.
• Models as Conviction Narratives: Thompson described models as ‘conviction narratives,’ tools that provide a structured way of thinking about the future. She stressed the importance of understanding models as narratives that require translation and interpretation to be effectively used in decision-making.
• Future of Modeling: Looking ahead, Thompson envisioned a modeling landscape that is more integrally connected with social and political contexts. She called for models that are transparent about their value judgments and capable of incorporating diverse perspectives to enhance trust and utility in decision-making.

Erica Thompson’s insights in the MINDS CAST episode provide a valuable perspective on the complexities of mathematical modeling. Her emphasis on critical engagement, diversity of perspectives, and the need for transparency in modeling practices offers a roadmap for navigating the challenges and potentials of models in decision-making. As we continue to rely on models to understand and shape our world, Thompson’s call for a more nuanced and responsible approach to modeling is both timely and essential.

Watch the recording below

About the Speaker