On 14th January 2020, TAPMI had the honour of hosting Dr. Michelle McDowell as a part of the 4th Winter School. Dr. McDowell is a research scientist at the Max Planck Institute for Human Development and the Harding Center for Risk Literacy. She shared her insights on the risk literacy challenges that we are currently facing. Risk is often required to be communicated in statistical measures but should be done keeping in mind the audience to whom it is being communicated.
Dr. McDowell stressed the importance of being transparent while sharing information especially in the case of reporting scientific data. She argued for using natural frequencies rather than conditional probabilities and discussed three ways to train people to use the priors. She took the example of the medical field and explained to the audience how the ability to correctly analyze, interpret and communicate statistical data is important, especially for the physicians.
A workshop was conducted to help the audience better understand the new ideas in risk literacy, where Dr. McDowell guided the participants on design considerations while presenting data. Participants learned how to use their creative and technical skills to present data in a transparent manner. The session ended with enthusiastic participants presenting their own ideas.
TAPMI also invited Prof. Shyam Sunder, from Yale University, who addressed the audience on “What we know about the human attitude towards risks”. Prof. Sunder is the James L. Frank Professor of Accounting, Economics, and Finance at the Yale School of Management and Professor in the Department of Economics. He started the session by first introducing two operationalized definitions of risks talked about the general concept of measuring risk and its limitations. Raising a subtle but deeper philosophical issue associated risk and uncertainty, Prof. Sunder explained that both of the concepts use general probability theory as a measure. However, in the real world, most, if not all, events cannot be assigned probability values.
He proposed an alternative of moving away from dispersion-based measures of risk and lean towards loss/harm conceptualizations of risk. “We should seek an explanation in observable opportunity sets, instead of unobservable Bernoulli functions”, he concluded.